FUNCTIONAL OUTCOME AFTER A SPINAL FRACTURE · 3.7% of total health care costs. Spinal fractures (including spinal cord injury) rank 7th (3.8%) in total trauma costs, with a mean cost

$Page 1: FUNCTIONAL OUTCOME AFTER A SPINAL FRACTURE · 3.7% of total health care costs. Spinal fractures (including spinal cord injury) rank 7th (3.8%) in total trauma costs, with a mean cost$

FUNCTIONAL OUTCOME AFTER A SPINAL FRACTURE

R.B. Post


For the printing and distribution of this thesis financial support by Bauerfeind Benelux BV, Biomet Nederland BV, DePuy Spine, Maatschap chirurgie/orthopedie Scheper Ziekenhuis Emmen, Medtronic Nederland BV, Stichting Anna fonds and Stichting Beatrixoord Noord‐Nederland is gratefully acknowledged. Thesis university of Groningen, the Netherlands ‐ with references ‐ with summary in Dutch. ISBN: 978‐90‐367‐3535‐3 (book) ISBN: 978‐90‐367‐3534‐6 (electronic version) Copyright © 2008 by R.B. Post All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording or otherwise without the prior written permission of the copyright holder. Chapter 2 to 5 reprinted with kind permission from Springer Science + Business Media. Printed by Drukkerij van Denderen, Groningen.


RIJKSUNIVERSITEIT GRONINGEN

FUNCTIONAL OUTCOME AFTER A SPINAL FRACTURE

Proefschrift

ter verkrijging van het doctoraat in de Medische Wetenschappen

aan de Rijksuniversiteit Groningen op gezag van de

Rector Magnificus, dr. F. Zwarts, in het openbaar te verdedigen op

woensdag 29 oktober 2008 om 13.15 uur

door

Richard Bernardus Post geboren op 19 augustus 1977

te Elp


Promotor: prof. dr. H.J. ten Duis Copromotores: dr. C.K. van der Sluis dr. V.J.M. Leferink Beoordelingscommissie: prof. dr. S.K. Bulstra prof. dr. J.H.B. Geertzen prof. dr. P. Patka


Paranimfen: drs. E.M. Post drs. R.J. Sol


CONTENTS

Table of contents Chapter 1 General introduction and outline of thesis 1 Chapter 2 Spinal mobility: Sagittal range of motion measured with the 31 SpinalMouse, a new non‐invasive device Chapter 3 Sagittal range of motion after a spinal fracture: does ROM 45 correlate with functional outcome? Chapter 4 Functional outcome 5 years after non‐operative treatment of 59 type A spinal fractures Chapter 5 Non‐operatively treated type A spinal fractures: mid‐term 73 versus long‐term outcome Chapter 6 Long‐term functional outcome after type A3 spinal fractures: 85 operative versus non‐operative treatment Chapter 7 General discussion and conclusions 99 Chapter 8 Summary 117 Chapter 9 Nederlandse samenvatting 125 List of abbreviations 135 Dankwoord 137 Curriculum Vitae 139


1

Chapter 1

General introduction

CHAPTER 1

2

Spinal fractures: epidemiology, costs

In the Netherlands, from January 2004 until December 2006, 6099 patients were treated in a hospital for a thoracolumbar spinal fracture without neurological deficit. These numbers include spinal fractures due to trauma, as well as osteoporosis‐induced fractures and pathological fractures [125]. During the same period, 2947 patients in the age group of 20 to 60 years were treated for a traumatic thoracolumbar spinal fracture without neurological deficit. This means an incidence of traumatic thoracolumbar fractures (without neurological deficit) of approximately 1.2 per 10,000 per year in the Netherlands [125]. A study reporting about the incidence of spinal fractures in Canada shows an incidence of 64 per 100,000. These figures include all spinal fractures, including fractures induced by osteoporosis and cervical fractures [46]. In a study from England, the annual incidence of spinal fractures between the age of 20 to 60 years was 2.5 per 10,000 for men and 1 per 10,000 for women [124]. Neurological deficits, ranging from single root lesions to complete paraplegia, were found in 22% of the cases in a cohort of 1,212 thoracolumbar spinal fracture patients [77]. A recent study reported about a cohort of 1,251 spinal fracture patients, from which 18% displayed neurological deficits [59]. Total medical costs of injuries in the Netherlands in 1999 were EUR 1.15 billion or 3.7% of total health care costs. Spinal fractures (including spinal cord injury) rank 7th (3.8%) in total trauma costs, with a mean cost of EUR 6,600 per patient [87]. Total costs of spinal accidents were found to be approximately EUR 22 million in 1997 [112].

Classification

A classification should allow the identification of any injury by means of a simple algorithm based on easily recognizable and consistent radiographic and clinical characteristics. In addition, it should provide a concise and descriptive terminology, information regarding the severity of the injury and guidance as to the choice of treatment. Finally, it should serve as an useful tool for future studies [77]. Böhler was one of the first to classify spinal fractures in 1930 [9]. Subsequently, Watson‐Jones recognized that the concept of stability and ligamentous integrity would be crucial in spinal fracture management [142]. Nicoll, who published in 1949 about spinal fractures in miners, also emphasized the concept of stability [97].

GENERAL INTRODUCTION

3

In 1963, Holdsworth presented a classification based on a two‐column theory [45]. The spine was visualized by 2 columns: the anterior column, consisting of the vertebral body and intervertebral disc, and the posterior column comprising the facet joints and the posterior ligamentous complex. After classification schemes by Kelly in 1968 [53] and Whitesides in 1977 [145], the first to present a three‐column theory was Louis in 1977 [75]. In the era of the computed tomography (CT), Denis presented in 1983 the nowadays frequently used three‐column theory [24]. The spine is divided into the anterior column (the anterior longitudinal ligament and the anterior two thirds of the vertebral body), the middle column (posterior one third of the vertebral body and the posterior longitudinal ligament) and the posterior column (all structures posterior to the posterior longitudinal ligament). In this system, spinal fractures are classified into four different types: compression fractures, burst fractures, seatbelt type injuries and fracture dislocations. Each of this type is then sub‐divided into one of three to four subtypes. According to Denis, loss of integrity in 2 out of the 3 columns will result in instability, consequently necessitating operative stabilization. Despite its widespread use, criticism on the Denis classification grew, stressing the oversimplification of the subject of instability. Attempts to modify the classification (emphasizing the presumed mechanistic properties of injury) were made by Ferguson and Allen [36]. McAfee extended Denis’ classification to further clarify stability in spinal fractures [81]. In 1994, two new classifications were presented; the load sharing classification (LSC) and the Comprehensive Classification (CC) [77, 82]. The LSC, developed by McCormack et al., rates the injury by giving points to 1) the amount of damaged vertebral body (comminution), 2) the spread of the fragments in the fracture site and 3) the amount of kyphosis correction necessary to restore the normal sagittal alignment [82]. This classification associates the vertebral body fracture‐anatomy with mechanical stability (the more points, the less load transfer capacity) and attempts to give direction to treatment. In addition to the Denis classification and the CC, the LSC is more and more used in literature [1, 102, 122]. Influenced by the increasing accessibility of CT and the need for a more sensitive classification, Magerl et al. presented the Comprehensive Classification in 1994, based on the AO fracture classification format [77]. It is based upon the patho‐morphological characteristics of the fracture, resulting in a progressive scale of growing morphological injury. The system distinguishes 3 main fracture types, following the suspected mechanism of injury:

CHAPTER 1

4

• type A fracture (compression of the vertebral body, no posterior lesions) • type B fracture (distraction, transverse disruption of 1 or 2 columns) • type C fracture (rotation, two‐column injury with rotational displacement)

Each of this fracture types is divided into three subgroups which are divided into a following subgroup, known from regular AO arrangement. In this classification, stability reduces by increasing classification, so a type C fracture is less stable than a type A fracture (see Table 1 and Figure 1). In this thesis, the Comprehensive Classification is used.

Table 1 Comprehensive Classification

A1.1 Endplate impaction

A1.2 Wedge impaction A1 Impaction fracture

A1.3 Vertebral body collapse

A2.1 Sagittal split fracture

A2.2 Coronal split fracture A2 Split fracture

A2.3 Pincer fracture

A3.1 Incomplete burst fracture

A3.2 Burst‐split fracture

A Compression injury

A3 Burst fracture

A3.3 Complete burst fracture

B1.1 With disc disruption B1 Posterior ligamentary lesion

B1.2 With type A fracture

B2.1 Transverse bicolumn

B2.2 With disc disruption B2 Posterior osseous lesion

B2.3 With type A fracture

B3.1 With subluxation

B3.2 With spondylolysis

B Distraction injury

B3 Anterior disc rupture

B3.3 With posterior dislocation C1.1 Rotational wedge fracture

C1.2 Rotational split fracture C1 Type A with rotation C1.3 Rotational burst fracture

C2.1 B1 lesion with rotation

C2.2 B2 lesion with rotation C2 Type B with rotation C2.3 B3 lesion with rotation

C3.1 Slice fracture

C Rotation injury

C3 Rotational shear injury C3.2 Oblique fracture


5

A1 Impaction A2 Split A3 Burst

B1 Posterior B2 Osseous distraction B3 Posterior distraction ligamentous disruption injury with anterior disruption

C1 Rotation with C2 Rotation with C3 Rotation with A fracture B fracture shear

Fig. 1 Comprehensive Classification: Type A fractures (compression), type B fractures (distraction) and type C fractures (rotation)

At present, the Comprehensive Classification as well as the Denis classification are the most commonly used schemes in classifying spinal fractures [106]. However, some concerns are present when studying both schemes. Reliability and repeatability of both systems have shown to be moderate [7, 63, 147]. Furthermore, both systems lack an important issue: they do not completely consider the integrity of the posterior ligamentous complex (PLC). This complex is believed to be of great

CHAPTER 1

6

importance in maintaining spinal stability [19, 99, 100, 152]. Hence, disruption of this structure might result in spinal instability and may lead to severe pain if not managed properly [67]. Even so, a CT‐scan does not provide direct information on the soft tissues, so the role of the PLC is not entirely acknowledged in the (CT‐based) CC and Denis classification. Lesions to the PLC can only be assumed on CT‐scans when interspinous widening is present. Detecting PLC injury on plain X‐rays or CT‐scans has shown not be accurate. For example, Leferink et al. showed that 30% of type B fractures (PLC lesion present) are misdiagnosed and are classified as being type A fractures (PLC intact) when only plain X‐rays and CT‐scans are used [71]. Whereas the CT‐scan can not directly detect injuries to the PLC, images made by using Magnetic Resonance Imaging (MRI) can visualize damage to the soft tissues, including the PLC. Lee et al. demonstrated the accuracy of the MRI detecting PLC injury to be 97%, with a negative predictive value of 100% [66]. Recognizing the importance of the PLC (and intervertebral disc) in spinal stability, the use of MRI will most likely play an important role in new classification systems in the near future [100]. Recently, Vaccaro et al., acknowledging the role of the PLC, proposed a new classification and severity score, the ThoracoLumbar Injury Severity Score (TLISS) [134]. It is based upon 3 categories with points assigned to each specific variable in a category; 1) the mechanism of injury (1 to 4 points), 2) the integrity of the posterior ligamentous complex (0 to 3 points) and 3) the patient’s neurological status (0 to 3 points). Points are summed, 3 points or less would implicate non‐operative treatment, 5 points or more indicate operative treatment should be preferred. Four points is an intermediate score leading to management either way [134]. The system demonstrated good reliability in terms of intra‐observer and inter‐observer agreement [106]. Lately, its concept has been modified by placing more emphasis on the morphology, resulting in the ThoracoLumbar Injury Classification and Severity Score (TLICS) [67, 132]. In the future, this scheme might possibly replace the commonly used classification schemes.

Treatment

The treatment goal in spinal fractures is to obtain early patient mobilization and a painless, balanced, stable vertebral column with maximum spine mobility and optimal neurological function [32]. In the light of the ICF (see page 12) this would mean a patient with no loss of body function, who can undertake all activities in the context of his or her culture [150].


7

Until the 1970’s non‐operative treatment was the paradigm in curing spinal fractures. Hippocrates was one of the first to treat spinal fractures [89]. Hippocrates, and later on Oribasius, treated patients by distraction, reduction and rest on a scamnum (see Figure 2). The word “scamnum” originates from Latin denoting “low bench” [89]. Since that time, many variations in non‐operative treatment have been used.

Fig. 2 Distraction and reduction on a scamnum

Non‐operative treatment can consist of bed rest, postural reduction, direct mobilization, ambulatory bracing (for example with a reclination brace, see Figure 3), and combinations of these. An early goal of non‐operative treatment is a mobile patient with or without brace. The means used as how to achieve this rather vary in literature and seem to be to some extent empirically based. Mumford et al. claimed good results after one month of bedrest followed by 3 months of bracing [94]. Shen advocated direct mobilization with or without a Jewett brace in three‐column “burst” fractures [120]. Closed reduction (on a Cotrel frame by axial traction and anterior shear) and casting for 3 months were described by Tropiano et al. [130]. Kinoshita et al. proposed 3 months of bedrest followed by a brace [54]. Others describe more or less equal treatment strategies, ranging from one week to 3 months of bedrest followed by a brace or thoracolumbosacral orthosis (TLSO) for 3 to 6 months [1, 14, 15, 38, 104, 107, 128]. Weinstein et al., as one of the most cited authors in this line of work, claimed good results after immediate mobilization with a brace or up to 3 months of bed rest [143].

CHAPTER 1

8

Fig. 3 Example of a three‐point reclination brace

With the development of operative techniques in the 1970’s, however, a second treatment modality for spinal fractures became available. Harrington instrumentation, which originally was developed for scoliosis surgery, was presented for use in spinal fractures in 1973 [28]. The Harrington system, using distraction and fixation, became the worldwide standard for operative stabilization in spinal fractures. Despite, some problems were encountered: a large part of the spine had to be immobilized (from 3 segments above the injured level to 3 segments below) to create a firm fixation. The Luque rod system, using sublaminar wires, achieved better fusion although more neurological complications occurred compared to the Harrington system [64]. Some of these problems were solved by the “Harrington‐like” Cotrel‐Dubousset instrumentation [92]. Meanwhile, Roy‐Camille et al. presented a technique consisting of posterior plates with screws positioned through the pedicles [117]. This transpedicular technique, combined with the “Harrington rod idea”, resulted (partially via Magerl’s fixateur externe) in the nowadays frequently used system according to Dick [26, 27, 78]. This technique consists of transpedicular placement of screws one level above and one level below the fractured vertebral body, which act as levers in reducing the kyphosis. These screws are connected by two short rods and so construct the “fixateur interne” according to Dick [27]. The most important advantage of this procedure is its capacity to create (and partly preserve) reduction of fractures by only immobilizing 2 segments.


9

Nowadays, posterior transpedicular fixation devices are the standard in dorsal operative approaches. Many dorsal implants are available today, all referring to the Dick internal fixator (see Figure 4) [16, 65, 116]. In this thesis, all patients who were managed operatively were treated by internal fixation, using the Universal Spine System [65].

Fig. 4 Example of an internal fixator in a model, bridging one segment

The dorsal approach is not the only possible operative procedure, though. Dunn and Kaneda presented a ventral approach in 1984 [31, 52]. This new technique was developed because of concerns about the retropulsed bony fragments which became visible on CT‐scans. The consideration was that a direct, anterior approach would offer better decompression of the spinal cord than an indirect posterior approach mainly based on ligamentotaxis [136]. Kostuik put the anterior and posterior approach together and presented the anterior Kostuik‐Harrington distraction device [60]. Presently, multiple types of anterior devices are available [138]. The anterior approach allows decompression of anterior neural compression, reconstruction of the anterior and middle columns of the thoracolumbar spine, and osteotomy through the vertebral body if needed [111]. It can be used as the first and only step (for example in high thoracic fractures) or as a second procedure when dorsal instrumentation has failed to adequately decompress the spinal canal [138]. The spinal column can be approached through thoracotomy, video‐assisted thoracoscopic surgery, and open transabdominal and retroperitoneal exposure [47].

CHAPTER 1

10

Recently, vertebroplasty and balloon kyphoplasty have become a topic of interest in the treatment of traumatic spinal fractures [137]. In vertebroplasty and balloon kyphoplasty, an inflatable balloon is brought into the fractured vertebral body percutaneously. By inflating the balloon, it restores height and corrects the kyphotic deformity. Afterwards, cement is injected into the remaining cavity. It is a commonly used technique for treating osteoporotic impression fractures [72]. However, recently it has also been used in the treatment of traumatic spinal fractures [101, 135]. The technique was found to be safe, but clinical results are still uncertain. Nowadays spinal fractures, like most other fractures, can be treated operatively or non‐operatively. Both modalities have their own advantages and disadvantages. Benefits of the operative approach are the improvement of spinal alignment, decreased deformity, early mobilization and rehabilitation (with a decrease in the complications of long bed rest) and sometimes improvement in neurological function or decreasing the possibility of neurological deterioration [40, 119, 146]. On the other hand, non‐operative treatment lacks the risks of surgery, such as deep wound infection, iatrogenic neurological injury and implant failure [107, 120, 146]. Furthermore, non‐operative treatment seems to be less expensive [44, 112, 121].

Indications

The decision to treat either operatively or non‐operatively is based on clinical (age, co‐morbidity, neurological status, other major injuries) and radiological findings. The distinction between stability and instability of the spine and the patient’s neurological status play an important role. Instability can be defined as the loss of the ability of the spine under physiological loads to maintain relationships between vertebrae so that there is no initial or additional neurological deficit, no major deformity, and no incapacitating pain [144]. In general, patients with stable fractures without gross deformities and no neurological deficits are treated non‐operatively. Patients with gross deformity and progressive neurological deficits are treated operatively. On the other hand, these are only indistinct criteria. In clinical practice, the decision on how to treat a traumatic thoracolumbar spinal fracture seems to be less simple. This is especially true for the so‐called “burst” fracture, i.e. the type A3.1, A3.2 and A3.3 fracture according to the CC [77]. This type of fracture is characterized by comminution of the vertebral body with centrifugal extrusion of fragments, whereas the posterior ligamentous complex is intact. The hallmark of this type of fracture is the extrusion


11

from bone into the spinal canal (disruption of the dorsal side of the vertebral body) (see Figure 5). The most favourable treatment for this fracture is still unknown; a large amount of literature is available concerning this “burst” fracture, reporting good results after both operative as well as non‐operative treatment [19, 22, 62, 107, 119, 122, 146].

Fig. 5 X‐ray (a) and CT‐scan (b) of a type A3.1 fracture (T12) in an 18‐year‐old male. Post‐operative status is shown in (c)

Nevertheless, when one has to decide which treatment is viable for a particular patient, which measure should one choose in determining success? Should the result of treatment be judged on radiological appearance of the vertebrae? Is the cost of treatment of any importance? Or should the result be measured in terms of patient satisfaction, pain or restrictions in daily activities? During the last decades, the concept of functional outcome has gained attention to evaluate the result of treatment [126].

Functional outcome

A precise definition of functional outcome is not easy to formulate. According to Baumberg et al., outcome is “the result of health care processes” [3]. However, this might not cover the complete meaning of functional outcome. Liebenson describes functional outcome as “the measurement of a patient’s status, either symptomatically or functionally” [74]. Outcome after trauma can be evaluated in numerous ways. One can measure survival, which is a simple, but in the field of spinal fractures less suitable approach. Usually, functional outcome is measured as a summary of numerous

CHAPTER 1

12

characteristics of daily living, like pain, return to work, ability to sport or social functioning. The International Classification of Impairments, Disabilities and Handicaps (ICIDH), published by the World Health Organization in 1980, is a model to describe the result of disease on patients’ health status [149]. In short, 4 entities are considered for any kind of disease (including trauma): pathology, impairment, disability and handicap. According to the World Health Organization, health can be defined as “a state of complete physical, mental and social well‐being and not merely the absence of disease or infirmity” [148]. In 2001, the “revised version” of the ICIDH was published, the International Classification of Functioning, Disability and Health (ICF) [150]. It consists of 3 more positively emphasized categories (body function/structure, activity, participation), all of these influenced by personal and environmental factors [150]. Significant deviations, or loss of body function and structure replace “impairment”. Activity is defined as performance of person‐level tasks or activities undertaken by a person in the context of their culture. Participation replaces “handicap” and expands the scope of disablement by classifying most areas of human life (see Figure 6) [127]. As being a more psychosocial model than the ICIDH, the ICF makes it possible to grade all the variables related to patients’ health status. Nevertheless, in reality it becomes clear that most outcome measures (including questionnaires) do not cover all the domains of the ICF [127].

Fig. 6 Health model according to the ICF

disorder / disease

environmental factors personal factors

body function & structure activity participation


13

Whereas in the beginning functional outcome was traditionally the area of rehabilitation medicine, during the last decade also other domains of medicine have paid interest in functional outcome. This includes the field of traumatology as well [19, 119, 122, 146]. In spinal fracture research, Weinstein et al. in 1988 were one of the first to study functional outcome [143]. Later on others studied outcome in different types of spinal fractures and treatments using variable outcome measures [1, 15, 94]. Kraemer et al., in 1996, even referred to the “traditional” radiological results as “surrogate outcome” [62]. Why should one measure functional outcome? Functional outcome measurements make it possible to 1) quantify clinical signs and symptoms, 2) objectify clinical symptoms, 3) make a baseline assessment, 4) evaluate the clinical course, 5) possibly predict the clinical course for the future and 6) establish a reliable basis for decision making [21]. By means of measurement instruments (including questionnaires) the afore‐mentioned data can be assembled in a uniform manner. This raises the question which instruments are available for evaluating outcome in spinal fractures.

Functional outcome measures

Measurement instruments can be divided into anthropometrical instruments (for example an inclinometer), questionnaires (to be completed by patients) and observational lists (to be completed by the examiner). Furthermore, one can test physical performance. Finally, combinations of all these entities are possible. When using a measurement instrument it should be reliable, valid, and responsive to the clinical change that occurs over time. Reliability describes how uniformly a test can be repeated when utilized on more than one occasion or by more than one rater, i.e. the consistency. Reliability can be tested as inter‐rater reliability (i.e. the reliability between more than one rater) and intra‐rater reliability (i.e. the reliability for the same rater when measuring at different occasions). Validity is the extent to which the instrument measures what it intends to measure. Responsiveness is the capacity of the measure to identify changes in patients’ health status over time. For a measurement instrument to be useful in clinical practice, it should satisfy at least the first two criteria described, and when measuring at different moments in time the last condition should be fulfilled as well.

CHAPTER 1

14

To measure outcomes in patients who sustained a spinal fracture multiple instruments are available (classified according to the ICF):

Measurements of impairments in body function and structure: Neurological status The neurological status after a spinal fracture is a gross, though useful measure. The most frequently used classification is that of Frankel, which describes spinal cord injuries according to the severity of deficit below the level of injury [39]. • Group A: complete interruption of all sensation and motor function • Group B: incomplete interruption, with some sensation but no motor function • Group C: incomplete interruption, with demonstrable voluntary motor function

but at a minimal, non‐useful level • Group D: incomplete interruption, with some voluntary motor function that is

useful to the patient • Group E: normal functioning

Physical capacity Physical performance measures have the potential to complement clinicians’ assessments and patients’ reports of outcome. Some of the measures used are:

• Range of Motion The Range of Motion (ROM) is a frequently proposed outcome measure. Concerning its use as outcome measure, literature reveals conflicting results, reporting about no to poor relationship between ROM and disability as well as significant correlation [18, 88, 95, 103, 140].

• Muscle strength One can use isokinetic or non‐dynamometric tests for assessing their correlation with subjective low back pain symptoms. For example, leg raising or repetitive arch‐up and sit‐up tests can be performed. In literature, the latter correlated significantly with pain and disability [69, 74, 110]. The Sorensen test, which is a static back‐extensor test, was found to correlate with disability in low back pain patients [6].

• Endurance tests Functional capacity (quantifying a larger component of body functioning) can be tested with lifting or carrying tests. Functional capacity, focussing on aerobic (cardiopulmonary) ability can be assessed with the use of a cycle ergometer. Cor‐relation with disability varied in literature though [37, 73, 84, 123].


15

The progressive isoinertial lifting evaluation (PILE), which we used in Chapter 4, is a psychophysical, isoinertial lifting test [79]. The patient is asked to repeatedly lift a weight from the floor to a table, this should be completed 4 times in 20 seconds. After each cycle, the load is increased [79]. Isoinertial relates to the force of a human muscle that is applied to a constant mass in motion. The psychophysical component lies in the fact that a patient can stop lifting when he finds himself at a point of discomfort or overexertion [79]. As such, this test represents a self‐selected “real world” lifting technique. The patient chooses the posture he experiences comfortable, and stops lifting when psychophysical (cognitive) factors like fatigue necessitate doing so. A weakness of the PILE (and all lifting tests) is the incapability to distinguish the “weak link” anywhere along the biomechanical chain.

Measurements of limitations in activity or participation: Return to work Return to work (RTW) is an outcome that is highly valued by patients, employers, insurance companies and society [1, 104, 119, 122, 143]. Clinicians frequently include return to work as one of the treatment goals. Although being a valuable outcome measure, RTW is affected by socio‐economic characteristics, economic incentives, job characteristics as well as employment status [43, 76, 109].

Health‐related quality of life Instruments measuring health‐related quality of life are mostly questionnaires. These questionnaires can be classified as generic (designed for broad use in a variety of patient populations) or condition‐specific (designed for use in specific patient populations). Condition‐specific instruments have several advantages. First, they target specific components of function that are most relevant to the disease or condition, furthermore they may be more responsive than generic instruments. In addition, many of these instruments can be scored quickly and the interpretation of their scores is less complex [109]. The following questionnaires have been used in spinal fracture patients:

Generic instruments • SF‐36 The Medical Outcomes Study 36‐item Short Form health survey (SF‐36) scale contains 9 scales measuring physical functioning, social functioning, role restriction due to physical problems, role restriction due to emotional problems,

CHAPTER 1

16

mental health, energy and vitality, pain, general perception of health and change in health over the past year. Scores can vary from 0 to 100, higher scores indicate better results [42, 141]. In literature, the test was found to be a reliable and valid measure [85].

• Sickness Impact Profile The Sickness Impact Profile (SIP) has been used in different (trauma) populations and is a reliable and valid instrument to measure the health‐related quality of life [5, 105]. The instrument is composed of 136 statements describing health‐related dysfunctional behaviors. The statements are grouped into 12 categories. A score can be computed for the overall instrument (SIP‐total) and for two subscales that characterize physical (SIP‐physical) and psychosocial dysfunction (SIP‐psychosocial). SIP scores from 0 to 3 are considered to reflect no disability, scores from 4 to 9 reveal mild disablement and scores from 10 to 19 illustrate moderate disability; severe disablement is reflected by SIP scores from 20 to 100 [51].

• EQ‐5D This questionnaire, formerly known as the EuroQol instrument, was published in 1990. The system consists of 5 domains: mobility, self‐care, usual activity, pain/discomfort and anxiety/depression. Each dimension has 3 levels, reflecting “no problem”, “some problem” and “extreme problem” [129]. Since 1998, a 6th dimension (cognition) has been added [61]. It has proved to be a valid and reliable instrument [17].

• Nottingham Health Profile (NHP) The NHP was originally developed to be used in epidemiological health studies. It assesses perceived or subjective health by asking for “yes” or “no” responses to 38 statements in 6 categories (energy level, emotional reactions, physical mobility, pain, social isolation and sleep). Scores, using weighted values, can range from 0 (no problems) to 100 (all items checked) for each category [83]. The NHP was found to be a valid and reliable measure [48].

Condition‐specific instruments More than 40 back pain questionnaires are available. The most frequently used are:

• Roland‐Morris Disability Questionnaire (RMDQ) The RMDQ is derived from the Sickness Impact Profile, from which 24 out of 136 items are selected. Those 24 questions are ticked dichotomously (yes/no). Each positive answer results in one point. The lowest possible score is 0 (no impairment) and the highest score is 24 (maximum impairment) [115]. The questions deal with


17

body functions (pain, sleeping and appetite) as well as activities (self care, walking, sitting, standing, lifting, work, dressing, stairs, housework and resting), but no environmental questions are included [93]. The RMDQ is one of the most frequently used questionnaires in spinal fracture populations, and showed to be a sensitive, reliable and valid instrument [93, 109, 126]. The Dutch version of the RMDQ was used in this thesis. This Dutch version also proved to be a reliable and valid measure [12, 114].

• Oswestry Disability Index The Oswestry Disability Index (ODI) is a valid and reliable questionnaire designed for determining the degree of functional limitation in patients consulting with low back pain in secondary care [20]. Ten items covering pain intensity, personal care, lifting, walking, sitting, standing, sleeping, sex life, social life, and travelling are scored [35]. However, important items considering the ability to work, need for help and items about environmental factors are not included. Nevertheless, together with the RMDQ it is the most frequently used questionnaire in low back pain and spinal fracture research [93].

• Denis outcome scale The Denis outcome scale recognizes 3 categories (pain, restriction in work and restriction in recreational activities), all on a scale of 1 to 5. One point is the most perfect situation, whereas 5 points indicate the worst possible outcome [25]. As a rather simple tool, it is popular in spinal fracture literature, although no studies concerning its psychometric characteristics are available.

• Visual Analogue Scale Spine Score The Visual Analogue Scale Spine Score (VAS) has the unique feature that it is developed to be used in spinal fracture patients. Patients are asked to rate the functional outcome in 19 items on a 10 cm visual scale. The patient’s perception of pain and restriction in activities related to back‐problems is measured. Higher scores represent better results, converted to percentages of the maximum score (0‐100). It has proved to be a reliable and valid instrument [58].

• Million Visual Analogue Scale This questionnaire was first published in 1982 for use in patients with chronic back pain. The 15 items focus on body functions (pain, sleep, stiffness and twisting), on activities (walking, sitting, standing and work) and on social life [90]. Answers are scored on a 10 cm visual analogue scale. According to the literature it is a valid and reliable instrument [93].

CHAPTER 1

18

• Waddell Disability Index (WDI) The WDI is a brief 9‐item scale focussing on disabilities (walking, sitting, standing, lifting, sex life, travelling and dressing), on body functions (pain, sleep) and on social life. Questions about work, self care and sports are not included [139]. Psychometric properties were reported to be good [20].

Literature review

Some data regarding functional outcome after a spinal fracture are available. Comparison of the results remains a difficult topic since treatment modalities, fracture classification, numbers of patients and outcome measures frequently vary between different authors. Some issues though seem to be generally accepted. There appears to be no correlation between the radiological appearance of the healed vertebral body (e.g. anterior wedge angle, vertebral height) and the functional outcome [38, 62, 94, 108, 122, 128, 143]. Furthermore, outcome in patients without neurological injury generally seems fairly good, both after operative as non‐operative treatment. Neurological deficit seems to have the greatest impact on outcome [86].

McLain studied outcome after spinal fractures treated with Cotrel‐Dubousset instrumentation [86]. Seventy percent of the subjects returned to full‐time work, 56% had no functional limitations. In a study concerning operative treatment after type A, B and C fractures (Comprehensive Classification) the RTW rate was found to be 50%, the mean Hannover spine score was 72% [56, 57]. In a meta analysis, 84% of the patients were found to have a P1 or P2 status (meaning no or minimal pain) after dorsal stabilization, 83% of the patients achieved W1 and W2 (indicating return to heavy labour or lighter labour) [25, 136]. A short time ago, Briem et al. measured outcome after operative and non‐operative treatment for type A and B fractures [10]. Results for the operative group showed a score of 72 points on the physical functioning index of the SF‐36 together with a VAS spine score of 60 points. In the non‐operatively treated group, these numbers were 75 and 67, respectively. Outcomes did not differ between these groups [10]. Reinhold et al. measured functional outcome 16 years after a non‐operatively treated type A fracture [108]. A mean VAS spine score of 58 points (indicating moderate impairment) was found. A study concerning outcome after non‐operatively treated wedge fractures (without neurological deficits) showed a score of 56 points (demonstrating rather severe impairment) on the Oswestry scale, 25%


19

of patients had changed their job [38]. Tezer et al. studied outcome after non‐operative treatment for spinal compression and “burst” fractures [128]. Pain was measured by means of Denis’ scale; the mean pain score was 1.66 (compression fractures) and 1.26 in the “burst” fractures [128]. The so‐called “burst” fracture (the type A3 fracture according to the Comprehensive Classification [77]) remains a fierce topic of debate. It is a fracture type that shows different outcomes in different treatment modalities. Operative treatment in this type of fracture shows good results. Leferink et al. reported good results after dorsal instrumentation; the mean RMDQ score was 4 and a mean VAS spine score of 79 was found [69]. In another study a score of 69 points on the SF‐36 physical functioning scale was found 4 years after dorsal instrumentation [11]. Sanderson et al. found good to excellent outcomes in 62% of patients treated operatively [118]. Recently, Defino et al. reported 66% of patients displaying P1 or P2 (indicating no or occasional pain [25]) two years after operative treatment for a type A3 fracture [22]. Non‐operative treatment in this type of fracture demonstrates good outcome as well. Mumford et al. found good to excellent outcomes in 66% of patients and the RTW rate was 81% [94]. Reid et al. reported a satisfactory pain score in all patients [107], whereas Aligizakis et al. found satisfactory results in 91% of patients [1]. Also other studies showed good results after non‐operative treatment [14, 15, 130, 143]. Studies directly comparing operative and non‐operative treatment for the type A3 “burst” fracture reveal contradictory results. Denis et al. found in a retrospective study superior outcomes after operative treatment, with a neurological deterioration in 17% of patients treated non‐operatively versus no deterioration after operative treatment [25]. These high percentages of neurological worsening though seem extraordinary. Such considerably high numbers have never been reported in other papers. Butler et al. found better outcomes (as measured by Denis’ outcome scale) for those treated non‐operatively [13]. Shen et al. reported no significant differences in RTW, SF‐36 and Oswestry scores after operative and non‐operative treatment at a 2‐year follow‐up. Operative treatment resulted in earlier pain reduction than non‐operative treatment, yet costs of operative treatment doubled that of non‐operative treatment [119]. Also other authors could not demonstrate a difference in outcome between operative and non‐operative treatment for the type A3 fracture [30, 55, 62]. Studies afore‐mentioned were all carried out in a retrospective setting, however. Recently, a literature review concerning optimal treatment in the type A3 “burst” fracture has been presented

CHAPTER 1

20

by Dai et al. [19]. According to this review, no superior treatment exists in the neurological intact type A3 “burst” fracture. A recent Cochrane review found only one adequate prospective randomized controlled trial comparing operative and non‐operative treatment [146, 151]. This study, by Wood et al., found a significant higher RMDQ score of 8.2 for those patients treated operatively versus 3.9 for those treated non‐operatively. RTW rates did not differ between the groups, SF‐36 and Oswestry scores did not differ either. They concluded that non‐operative treatment in type A3 “burst” fractures is at least as valuable as operative treatment [146]. Short after this Cochrane publication, a paper by Siebenga et al. was published comparing treatment outcomes after type A3 fractures, studied in a multi‐centre, prospective randomized setting [122]. They found better outcomes in patients treated operatively. Above‐mentioned studies report nearly all on dorsal operative procedures. Data on functional outcome after ventral operative procedures are scarce. On one hand, anterior surgery could produce a more complete and reliable decompression of the spinal canal; on the other hand it requires a more sophisticated technique and may result in serious adverse effects [33]. Okuyama et al. found good results after anterior surgery, 84% of the patients scoring P1 or P2, indicating minimal or no pain [25, 98]. Ghanayem et al. found good or excellent results in 92% of patients after anterior instrumentation [41].

The aim of this thesis is to study different aspects of functional outcome after a spinal fracture. Considering the above described, much is known on this topic, but many questions remain unsolved. For example, what is the ROM after a spinal fracture, how does it correlate with functional outcome, and how to measure the ROM? Furthermore, what is the short‐term and long‐term outcome after non‐operatively treated type A fractures without neurological deficit? Also the optimal treatment (operative versus non‐operative) in the type A3 “burst” fracture remains unknown. Together with other specific questions this thesis tries to find an answer to these issues.

Outline of the thesis

Information on epidemiology, classification, treatment, functional outcome and its measures as well as a literature review on the topic of spinal fractures is provided in Chapter 1. In measuring functional outcome, one proposed tool is the assessment of ROM. Many methods of evaluating spinal range of motion have been described. One


21

method used is radiological analysis (CT‐scans, plain‐ and biplanar radiography) [29, 49, 91]. Radiological measurement, however, carries the risk of the relatively high dose of radiation it requires, which precludes its use as a routine measurement in clinical practice. Consequently, many non‐invasive, external methods have been developed like goniometers, skin markers, inclinometers and spondylometers [68, 80, 96]. Since they are relatively easy to use and involve little clinical time, external methods are nowadays commonly used [96]. The clinical usage and validation of the SpinalMouse, a computerized external device for measuring spinal ROM is presented in Chapter 2. Inter‐rater reliability and use in clinical practice were studied. The residual range of motion after a spinal fracture is uncertain. Literature with reference to total spinal mobility is scarce, as most studies report about intersegmental ROM [23, 70, 113]. The few studies available concerning total spinal ROM after a spinal fracture reveal contradictory results. In one study sagittal spinal ROM was found to be normal after operative treatment for thoracolumbar spinal fractures [50]. Another study reported that spinal ROM did not return to normal after Harrington rod removal in patients treated operatively for a spinal fracture [29]. As the ROM after a spinal fracture is still uncertain, little is known about the influence of the resulting spinal ROM on subjective impairment. In other words, is measurement of spinal ROM a valid measure for assessing functional outcome? Previously published papers concerning this issue show different results [18, 95, 103]. Spinal range of motion after a spinal fracture is illustrated in Chapter 3. We measured thoracolumbar ROM and functional outcome in operatively and non‐operatively treated spinal fracture patients as well as in controls. The following issues were addressed: • Is there a difference in sagittal spinal ROM between operatively treated

patients, non‐operatively treated patients and controls? • Do the average VAS and RMDQ scores differ between operatively treated

patients, non‐operatively treated patients and controls? • Does sagittal spinal ROM correlate with subjective impairment, measured by

the RMDQ and VAS?

In Chapter 4 the functional outcome after non‐operative treatment of type A spinal fractures without neurological deficit is presented. Functional outcome was determined in a wide spectrum following the International Classification of Functioning, Disability and Health (ICF), measuring restrictions in body function and structure, restrictions in activities, and restrictions in participation/quality of

CHAPTER 1

22

life [131, 150]. Patients completed physical tests (dynamic lifting tests as well as an ergometry exercise test) plus questionnaires to construct a well‐based functional outcome dimension. Most of the published data on functional outcome after a spinal fracture concentrate on relatively short‐term results. Literature regarding long‐term outcome (10 years and over) is reasonably scarce [38, 108, 143]. It is known that pain may arise in the long term due to changed facet joint motion and hyperextension of adjacent spinal regions, leading to ongoing degenerative processes [99, 133]. Also fatigue pain from the soft tissues has been described as contributing to back pain in the long term [4, 130]. Chapter 5 describes the long‐term functional outcome of non‐operatively treated type A spinal fracture patients. Functional outcome approximately 10 years after trauma was measured by means of questionnaires. Long‐term outcome was compared to the mid‐term functional outcome (4 years post‐injury) in the same cohort of patients. In spite of much literature trying to find the optimal treatment (operative versus non‐operative) in the type A3 “burst” fracture still no clear answer is available regarding this topic [119, 122, 146]. Operative treatment provides the benefits of improvement in spinal alignment, decreased deformity, early mobilization and improvement in neurological functioning [2, 25, 34, 99]. Alternatively, non‐operative treatment does not carry the risks of surgery, like deep wound infection, iatrogenic neurological damage and implant failure [14, 94, 107, 120]. Some studies comparing short‐term functional outcomes are available, literature regarding long‐term outcome is less presented. Several authors fear complications in the long term though, like progressive kyphosis and pain [8, 133]. In Chapter 6 we compared long‐term (5 years) functional outcomes of operatively and non‐operatively treated patients who sustained a type A3 “burst” fracture without neurological deficit. A general discussion is provided in Chapter 7. The studies enclosed are reviewed and conclusions are drawn, some recommendations are made and future research options are discussed. Finally, a summary is presented in Chapter 8, followed by a summary in Dutch in Chapter 9.

References

1. Aligizakis A, Katonis P, Stergiopoulos K, Galanakis I, Karabekios S, Hadjipavlou A (2002) Functional outcome of burst fractures of the thoracolumbar spine managed non‐operatively, with early ambulation, evaluated using the load sharing classification. Acta Orthop Belg 68:279‐287


23

2. Andress HJ, Braun H, Helmberger T, Schurmann M, Hertlein H, Hartl WH (2002) Long‐term results after posterior fixation of thoraco‐lumbar burst fractures. Injury 33:357‐365

3. Baumberg L, Long A, Jefferson J (1995) International workshop: culture and outcomes. Leeds: European Clearing House on Health Outcomes

4. Been HD, Poolman RW, Ubags LH (2004) Clinical outcome and radiographic results after surgical treatment of post‐traumatic thoracolumbar kyphosis following simple type A fractures. Eur Spine J 13:101‐107

5. Bergner M, Bobbitt RA, Carter WB, Gilson BS (1981) The Sickness Impact Profile: development and final revision of a health status measure. Med Care 19:787‐805

6. Biering‐Sorensen F (1984) Physical measurements as risk indicators for low‐back trouble over a one‐year period. Spine 9:106‐119

7. Blauth M, Bastian L, Knop C, Lange U, Tusch G (1999) Interobserverreliabilität bei der Klassifikation von thorakolumbalen Wirbelsäulenverletzungen. Orthopäde 28:662‐681

8. Bohlman HH, Kirkpatrick JS, Delamarter RB, Leventhal M (1994) Anterior decompression for late pain and paralysis after fractures of the thoracolumbar spine. Clin Orthop Relat Res 300:24‐29

9. Böhler L (1930) Die Techniek der Knochenbruchbehandlung im Frieden und im Kriege. Verlag von Wilhelm Maudrich, Wien

10. Briem D, Behechtnejad A, Ouchmaev A, Morfeld M, Schermelleh‐Engel K, Amling M, Rueger JM (2007) Pain regulation and health‐related quality of life after thoracolumbar fractures of the spine. Eur Spine J 16:1925‐1933

11. Briem D, Lehmann W, Ruecker AH, Windolf J, Rueger JM, Linhart W (2004) Factors influencing the quality of life after burst fractures of the thoracolumbar transition. Arch Orthop Trauma Surg 124:461‐468

12. Brouwer S, Kuijer W, Dijkstra PU, Goeken LN, Groothoff JW, Geertzen JH (2004) Reliability and stability of the Roland Morris Disability Questionnaire: intra class correlation and limits of agreement. Disabil Rehabil 26:162‐165

13. Butler JS, Walsh A, OʹByrne J (2005) Functional outcome of burst fractures of the first lumbar vertebra managed surgically and conservatively. Int Orthop 29:51‐54

14. Cantor JB, Lebwohl NH, Garvey T, Eismont FJ (1993) Nonoperative management of stable thoracolumbar burst fractures with early ambulation and bracing. Spine 18:971‐976

15. Chow GH, Nelson BJ, Gebhard JS, Brugman JL, Brown CW, Donaldson DH (1996) Functional outcome of thoracolumbar burst fractures managed with hyperextension casting or bracing and early mobilization. Spine 21:2170‐2175

16. Ciappetta P, Delfini R, Costanzo G (1996) Posterolateral decompression and stabilization of thoracolumbar injuries using Diapason instrumentation. Acta Neurochir 138:314‐321

17. Coons SJ, Rao S, Keininger DL, Hays RD (2000) A comparative review of generic quality‐of‐life instruments. Pharmacoeconomics 17:13‐35

18. Cox ME, Asselin S, Gracovetsky SA, Richards MP, Newman NM, Karakusevic V, Zhong L, Fogel JN (2000) Relationship between functional evaluation measures and self‐assessment in nonacute low back pain. Spine 25:1817‐1826

19. Dai LY, Jiang SD, Wang XY, Jiang LS (2007) A review of the management of thoracolumbar burst fractures. Surg Neurol 67:221‐231

20. Davidson M, Keating JL (2002) A comparison of five low back disability questionnaires: reliability and responsiveness. Phys Ther 82:8‐24

21. Davies AR (1994) Patient defined outcomes. Qual Health Care 3 Suppl:6‐9 22. Defino HL, Canto FR (2007) Low thoracic and lumbar burst fractures: radiographic and

functional outcomes. Eur Spine J 16:1934‐1943

CHAPTER 1

24

23. Dekutoski MB, Conlan ES, Salciccioli GG (1993) Spinal mobility and deformity after Harrington rod stabilization and limited arthrodesis of thoracolumbar fractures. J Bone Joint Surg Am 75:168‐176

24. Denis F (1983) The three column spine and its significance in the classification of acute thoracolumbar spinal injuries. Spine 8:817‐831

25. Denis F, Armstrong GW, Searls K, Matta L (1984) Acute thoracolumbar burst fractures in the absence of neurologic deficit. A comparison between operative and nonoperative treatment. Clin Orthop Relat Res 189:142‐149

26. Dick W (1984) Osteosynthese schwerer Verletzungen der Brust‐ und Lendenwirbelsäule mit dem Fixateur interne. Langenbecks Arch Chir 364:343‐346

27. Dick W (1987) The “fixateur interne” as a versatile implant for spine surgery. Spine 12:882‐900

28. Dickson JH, Harrington PR, Erwin WD (1973) Harrington instrumentation in the fractured, unstable thoracic & lumbar spine. Tex Med 69:91‐98

29. Dodd CA, Fergusson CM, Pearcy MJ, Houghton GR (1986) Vertebral motion measured using biplanar radiography before and after Harrington rod removal for unstable thoracolumbar fractures of the spine. Spine 11:452‐455

30. Domenicucci M, Preite R, Ramieri A, Ciappetta P, Delfini R, Romanini L (1996) Thoracolumbar fractures without neurosurgical involvement: surgical or conservative treatment? J Neurosurg Sci 40:1‐10

31. Dunn HK (1984) Anterior stabilization of thoracolumbar injuries. Clin Orthop Relat Res 189:116‐124

32. Eskenazi M, Bendo J, Spivak J (2000) Thoracolumbar spine trauma: evaluation and management. Curr Opinion Orthop 11:176‐185

33. Esses SI, Botsford DJ, Kostuik JP (1990) Evaluation of surgical treatment for burst fractures. Spine 15:667‐673

34. Esses SI, Botsford DJ, Wright T, Bednar D, Bailey S (1991) Operative treatment of spinal fractures with the AO internal fixator. Spine 16:S146‐S150

35. Fairbank JC, Couper J, Davies JB, OʹBrien JP (1980) The Oswestry low back pain disability questionnaire. Physiotherapy 66:271‐273

36. Ferguson RL, Allen BL (1984) A mechanistic classification of thoracolumbar spine fractures. Clin Orthop Relat Res 189:77‐88

37. Filho IT, Simmonds MJ, Protas EJ, Jones S (2002) Back pain, physical function, and estimates of aerobic capacity: what are the relationships among methods and measures? Am J Phys Med Rehabil 81:913‐920

38. Folman Y, Gepstein R (2003) Late outcome of nonoperative management of thoracolumbar vertebral wedge fractures. J Orthop Trauma 17:190‐192

39. Frankel HL, Hancock DO, Hyslop G, Melzak J, Michaelis LS, Ungar GH, Vernon JD, Walsh JJ (1969) The value of postural reduction in the initial management of closed injuries of the spine with paraplegia and tetraplegia. Paraplegia 7:179‐192

40. Gertzbein SD (1992) Scoliosis Research Society. Multicenter spine fracture study. Spine 17:528‐540

41. Ghanayem AJ, Zdeblick TA (1997) Anterior instrumentation in the management of thoracolumbar burst fractures. Clin Orthop Relat Res 335:89‐100

42. Grevitt M, Khazim R, Webb J, Mulholland R, Shepperd J (1997) The short form‐36 health survey questionnaire in spine surgery. J Bone Joint Surg Br 79:48‐52


25

43. Haefeli M, Elfering A, Aebi M, Freeman BJ, Fritzell P, Guimaraes Consciencia J, Lamartina C, Mayer M, Lund T, Boos N (2008) What comprises a good outcome in spinal surgery? A preliminary survey among spine surgeons of the SSE and European spine patients. Eur Spine J 17:104‐116

44. Hitchon PW, Torner JC, Haddad SF, Follett KA (1998) Management options in thoracolumbar burst fractures. Surg Neurol 49:619‐626

45. Holdsworth FW (1963) Fractures, dislocations, and fracture‐dislocations of the spine. J Bone Joint Surg Br 45:6‐20

46. Hu R, Mustard CA, Burns C (1996) Epidemiology of incident spinal fracture in a complete population. Spine 21:492‐499

47. Ikard RW (2006) Methods and complications of anterior exposure of the thoracic and lumbar spine. Arch Surg 141:1025‐1034

48. Jenkinson C, Fitzpatrick R, Argyle M (1988) The Nottingham Health Profile: an analysis of its sensitivity in differentiating illness groups. Soc Sci Med 27:1411‐1414

49. Johnsson R, Selvik G, Stromqvist B, Sunden G (1990) Mobility of the lower lumbar spine after posterolateral fusion determined by roentgen stereophotogrammetric analysis. Spine 15:347‐350

50. Junge A, Gotzen L, von Garrel T, Ziring E, Giannadakis K (1997) Die monosegmentale Fixateur interne‐Instrumentation und Fusion in der Behandlung von Frakturen der thorakolumbalen Wirbelsäule. Indikation, Technik und Ergebnisse. Unfallchirurg 100:880‐887

51. Jurkovich G, Mock C, MacKenzie E, Burgess A, Cushing B, deLateur B, McAndrew M, Morris J, Swiontkowski M (1995) The Sickness Impact Profile as a tool to evaluate functional outcome in trauma patients. J Trauma 39:625‐631

52. Kaneda K, Abumi K, Fujiya M (1984) Burst fractures with neurologic deficits of the thoracolumbar spine. Results of anterior decompression and stabilization with anterior instrumentation. Spine 9:788‐795

53. Kelly RP, Whitesides T (1968) Treatment of lumbodorsal fracture‐dislocations. Ann Surg 167:705‐717

54. Kinoshita H, Nagata Y, Ueda H, Kishi K (1993) Conservative treatment of burst fractures of the thoracolumbar and lumbar spine. Paraplegia 31:58‐67

55. Knight RQ, Stornelli DP, Chan DP, Devanny JR, Jackson KV (1993) Comparison of operative versus nonoperative treatment of lumbar burst fractures. Clin Orthop Relat Res 293:112‐121

56. Knop C, Blauth M, Buhren V, Hax PM, Kinzl L, Mutschler W, Pommer A, Ulrich C, Wagner S, Weckbach A, Wentzensen A, Worsdorfer O (1999) Operative Behandlung von Verletzungen des thorakolumbalen Übergangs. Teil 1: Epidemiologie. Unfallchirurg 102:924‐935

57. Knop C, Fabian HF, Bastian L, Blauth M (2001) Late results of thoracolumbar fractures after posterior instrumentation and transpedicular bone grafting. Spine 26:88‐99

58. Knop C, Oeser M, Bastian L, Lange U, Zdichavsky M, Blauth M (2001) Entwicklung und Validierung des VAS‐Wirbelsäulenscores. Unfallchirurg 104:488‐497

59. Knop C, Reinhold M, Roeder C, Staub L, Schmid R, Beisse R, Buhren V, Blauth M (2006) Internet based multicenter study for thoracolumbar injuries: a new concept and preliminary results. Eur Spine J 15:1687‐1694

60. Kostuik JP (1984) Anterior fixation for fractures of the thoracic and lumbar spine with or without neurologic involvement. Clin Orthop Relat Res 189:103‐115

61. Krabbe PF, Stouthard ME, Essink‐Bot ML, Bonsel GJ (1999) The effect of adding a cognitive dimension to the EuroQol multiattribute health‐status classification system. J Clin Epidemiol 52:293‐301

CHAPTER 1

26

62. Kraemer WJ, Schemitsch EH, Lever J, McBroom RJ, McKee MD, Waddell JP (1996) Functional outcome of thoracolumbar burst fractures without neurological deficit. J Orthop Trauma 10:541‐544

63. Kriek JJ, Govender S (2006) AO‐classification of thoracic and lumbar fractures; reproducibility utilizing radiographs and clinical information. Eur Spine J 15:1239‐1246

64. Lahde RE (1983) Luque rod instrumentation. AORN J 38:35‐43 65. Laxer E (1994) A further development in spinal instrumentation. Technical Commission for

Spinal Surgery of the ASIF. Eur Spine J 3:347‐352 66. Lee HM, Kim HS, Kim DJ, Suk KS, Park JO, Kim NH (2000) Reliability of magnetic resonance

imaging in detecting posterior ligament complex injury in thoracolumbar spinal fractures. Spine 25:2079‐2084

67. Lee JY, Vaccaro AR, Lim MR, Öner FC, Hulbert RJ, Hedlund R, Fehlings MG, Arnold P, Harrop J, Bono CM, Anderson PA, Anderson DG, Harris MB, Brown AK, Stock GH, Baron EM (2005) Thoracolumbar injury classification and severity score: a new paradigm for the treatment of thoracolumbar spine trauma. J Orthop Sci 10:671‐675

68. Lee YH, Chiou WK, Chen WJ, Lee MY, Lin YH (1995) Predictive model of intersegmental mobility of lumbar spine in the sagittal plane from skin markers. Clinical Biomechanics 10:413‐420

69. Leferink VJM, Keizer HJE, Oosterhuis JK, van der Sluis CK, ten Duis HJ (2003) Functional outcome in patients with thoracolumbar burst fractures treated with dorsal instrumentation and transpedicular cancellous bone grafting. Eur Spine J 12:261‐267

70. Leferink VJM, Nijboer JMM, Zimmerman KW, Veldhuis EFM, ten Vergert EM, ten Duis HJ (2002) Thoracolumbar spinal fractures: segmental range of motion after dorsal spondylodesis in 82 patients: a prospective study. Eur Spine J 11:2‐7

71. Leferink VJM, Zimmerman KW, Veldhuis EFM, ten Vergert EM, ten Duis HJ (2002) Classificational problems in ligamentary distraction type vertebral fractures: 30% of all B‐type fractures are initially unrecognised. Eur Spine J 11:246‐250

72. Lewis G (2007) Percutaneous vertebroplasty and kyphoplasty for the stand‐alone augmentation of osteoporosis‐induced vertebral compression fractures: present status and future directions. J Biomed Mater Res B Appl Biomater 81:371‐386

73. Liebenson C (1996) Rehabilitation and chiropractic practice. J Manipulative Physiol Ther 19:134‐140

74. Liebenson C, Yeomans S (1997) Outcomes assessment in musculoskeletal medicine. Manual Therapy 2:67‐74

75. Louis R (1977) Fractures instables du rachis. III. L’instabilité. A. Les theories de l’instabilité. Rev Chir Orthop Reparatrice Appar Mot 63:423‐425

76. MacKenzie EJ, Morris JA, Jurkovich GJ, Yasui Y, Cushing BM, Burgess AR, DeLateur BJ, McAndrew MP, Swiontkowski MF (1998) Return to work following injury: the role of economic, social, and job‐related factors. Am J Public Health 88:1630‐1637

77. Magerl F, Aebi M, Gertzbein SD, Harms J, Nazarian S (1994) A comprehensive classification of thoracic and lumbar injuries. Eur Spine J 3:184‐201

78. Magerl FP (1984) Stabilization of the lower thoracic and lumbar spine with external skeletal fixation. Clin Orthop Relat Res 189:125‐141

79. Mayer TG, Barnes D, Kishino ND, Nichols G, Gatchel RJ, Mayer H, Mooney V (1988) Progressive isoinertial lifting evaluation. I. A standardized protocol and normative database. Spine 13:993‐997

80. Mayer TG, Kondraske G, Beals SB, Gatchel RJ (1997) Spinal range of motion. Accuracy and sources of error with inclinometric measurement. Spine 22:1976‐1984


27

81. McAfee PC, Yuan HA, Fredrickson BE, Lubicky JP (1983) The value of computed tomography in thoracolumbar fractures. An analysis of one hundred consecutive cases and a new classification. J Bone Joint Surg Am 65:461‐473

82. McCormack T, Karaikovic E, Gaines RW (1994) The load sharing classification of spine fractures. Spine 19:1741‐1744

83. McEwen J, McKenna SP (1996) Nottingham Health Profile. In: Spilker B (ed) Quality of life and pharmacoeconomics in clinical trials. 2nd ed, pp 281‐286. Lippincott‐Raven, Philadelphia

84. McGill SM, Childs A, Liebenson C (1999) Endurance times for low back stabilization exercises: clinical targets for testing and training from a normal database. Arch Phys Med Rehabil 80:941‐944

85. McHorney CA, Ware JE, Raczek AE (1993) The MOS 36‐Item Short‐Form Health Survey (SF‐36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Med Care 31:247‐263

86. McLain RF (2004) Functional outcomes after surgery for spinal fractures: return to work and activity. Spine 29:470‐477

87. Meerding WJ, Mulder S, van Beeck EF (2006) Incidence and costs of injuries in The Netherlands. Eur J Public Health 16:272‐278

88. Mellin G (1987) Correlations of spinal mobility with degree of chronic low back pain after correction for age and anthropometric factors. Spine 12:464‐468

89. Memmert M (1999) Ein Versuch, die Geschichte der Wirbelsäulenchirurgie zu umreissen. In: Memmert M, Memmert G (eds) Die Wirbelsäule in der Anschauung. Spurensuche in Kunst, Geschichte und Sprache, pp 247‐270. Springer‐Verlag, Berlin Heidelberg

90. Million R, Hall W, Nilsen KH, Baker RD, Jayson MI (1982) Assessment of the progress of the back‐pain patient. 1981 Volvo Award in Clinical Science. Spine 7:204‐212

91. Miyasaka K, Ohmori K, Suzuki K, Inoue H (2000) Radiographic analysis of lumbar motion in relation to lumbosacral stability. Investigation of moderate and maximum motion. Spine 25:732‐737

92. Moreland DB, Egnatchik JG, Bennett GJ (1990) Cotrel‐Dubousset instrumentation for the treatment of thoracolumbar fractures. Neurosurgery 27:69‐73

93. Muller U, Duetz MS, Roeder C, Greenough CG (2004) Condition‐specific outcome measures for low back pain. Part I: validation. Eur Spine J 13:301‐313

94. Mumford J, Weinstein JN, Spratt KF, Goel VK (1993) Thoracolumbar burst fractures. The clinical efficacy and outcome of nonoperative management. Spine 18:955‐970

95. Nattrass CL, Nitschke JE, Disler PB, Chou MJ, Ooi KT (1999) Lumbar spine range of motion as a measure of physical and functional impairment: an investigation of validity. Clin Rehabil 13:211‐218

96. Ng JK, Kippers V, Richardson CA, Parnianpour M (2001) Range of motion and lordosis of the lumbar spine: reliability of measurement and normative values. Spine 26:53‐60

97. Nicoll EA (1949) Fractures of the dorso‐lumbar spine. J Bone Joint Surg Br 31‐B:376‐394 98. Okuyama K, Abe E, Chiba M, Ishikawa N, Sato K (1996) Outcome of anterior decompression

and stabilization for thoracolumbar unstable burst fractures in the absence of neurologic deficits. Spine 21:620‐625

99. Öner FC, van Gils APG, Faber JAJ, Dhert WJ, Verbout AJ (2002) Some complications of common treatment schemes of thoracolumbar spine fractures can be predicted with magnetic resonance imaging: prospective study of 53 patients with 71 fractures. Spine 27:629‐636

100. Öner FC, van Gils APG, Dhert WJ, Verbout AJ (1999) MRI findings of thoracolumbar spine fractures: a categorisation based on MRI examinations of 100 fractures. Skeletal Radiol 28:433‐443

CHAPTER 1

28

101. Öner FC, Verlaan JJ, Verbout AJ, Dhert WJ (2006) Cement augmentation techniques in traumatic thoracolumbar spine fractures. Spine 31:S89‐S95

102. Parker JW, Lane JR, Karaikovic EE, Gaines RW (2000) Successful short‐segment instrumentation and fusion for thoracolumbar spine fractures: a consecutive 41/2‐year series. Spine 25:1157‐1170

103. Poitras S, Loisel P, Prince F, Lemaire J (2000) Disability measurement in persons with back pain: a validity study of spinal range of motion and velocity. Arch Phys Med Rehabil 81:1394‐1400

104. Post RB, Keizer HJE, Leferink VJM, van der Sluis CK (2006) Functional outcome 5 years after non‐operative treatment of type A spinal fractures. Eur Spine J 15:472‐478

105. Post RB, van der Sluis CK, ten Duis HJ (2006) Return to work and quality of life in severely injured patients. Disabil Rehabil 28:1399‐1404

106. Raja‐Rampersaud Y, Fisher C, Wilsey J, Arnold P, Anand N, Bono CM, Dailey AT, Dvorak M, Fehlings MG, Harrop JS, Öner FC, Vaccaro AR (2006) Agreement between orthopedic surgeons and neurosurgeons regarding a new algorithm for the treatment of thoracolumbar injuries: a multicenter reliability study. J Spinal Disord Tech 19:477‐482

107. Reid DC, Hu R, Davis LA, Saboe LA (1988) The nonoperative treatment of burst fractures of the thoracolumbar junction. J Trauma 28:1188‐1194

108. Reinhold M, Knop C, Lange U, Bastian L, Blauth M (2003) Nichtoperative Behandlung von Verletzungen der thorakolumbalen Wirbelsäule. Klinische Spätergebnisse nach 16 Jahren. Unfallchirurg 106:566‐576

109. Resnik L, Dobrykowski E (2005) Outcomes measurement for patients with low back pain. Orthop Nurs 24:14‐24

110. Rissanen A, Alaranta H, Sainio P, Harkonen H (1994) Isokinetic and non‐dynamometric tests in low back pain patients related to pain and disability index. Spine 19:1963‐1967

111. Robertson PA (2007) Ns10 anterior approaches for thoracolumbar fractures. ANZ J Surg 77 Suppl 1:A54

112. Roer N van der, de Bruyne MC, Bakker FC, van Tulder MW, Boers M (2005) Direct medical costs of traumatic thoracolumbar spine fractures. Acta Orthop 76:662‐666

113. Rohlmann A, Neller S, Bergmann G, Graichen F, Claes L, Wilke HJ (2001) Effect of an internal fixator and a bone graft on intersegmental spinal motion and intradiscal pressure in the adjacent regions. Eur Spine J 10:301‐308

114. Roland M, Fairbank J (2000) The Roland‐Morris Disability Questionnaire and the Oswestry Disability Questionnaire. Spine 25:3115‐3124

115. Roland M, Morris R (1983) A study of the natural history of back pain. Part I: development of a reliable and sensitive measure of disability in low‐back pain. Spine 8:141‐144

116. Romero J, Vilar G, Bravo P (1994) Fractures of the dorsolumbar spine with neurological lesions. A comparison of different treatments. Int Orthop 18:157‐163

117. Roy‐Camille R, Saillant G, Berteaux D, Salgado V (1976) Osteosynthesis of thoraco‐lumbar spine fractures with metal plates screwed through the vertebral pedicles. Reconstr Surg Traumat 15:2‐16

118. Sanderson PL, Fraser RD, Hall DJ, Cain CM, Osti OL, Potter GR (1999) Short segment fixation of thoracolumbar burst fractures without fusion. Eur Spine J 8:495‐500

119. Shen WJ, Liu TJ, Shen YS (2001) Nonoperative treatment versus posterior fixation for thoracolumbar junction burst fractures without neurologic deficit. Spine 26:1038‐1045

120. Shen WJ, Shen YS (1999) Nonsurgical treatment of three‐column thoracolumbar junction burst fractures without neurologic deficit. Spine 24:412‐415


29

121. Siebenga J, Segers MJM, Leferink VJM, Elzinga MJ, ten Duis HJ, Rommens PM, Patka P (2007) Cost‐effectiveness of the treatment of traumatic thoracolumbar spine fractures: Nonsurgical or surgical therapy? Indian J Orthop 41:332‐336

122. Siebenga J, Leferink VJM, Segers MJM, Elzinga MJ, Bakker FC, Haarman HJ, Rommens PM, ten Duis HJ, Patka P (2006) Treatment of traumatic thoracolumbar spine fractures: a multicenter prospective randomized study of operative versus nonsurgical treatment. Spine 31:2881‐2890

123. Simmonds MJ, Olson SL, Jones S, Hussein T, Lee CE, Novy D, Radwan H (1998) Psychometric characteristics and clinical usefulness of physical performance tests in patients with low back pain. Spine 23:2412‐2421

124. Singer BR, McLauchlan GJ, Robinson CM, Christie J (1998) Epidemiology of fractures in 15,000 adults: the influence of age and gender. J Bone Joint Surg Br 80:243‐248

125. Stichting Prismant (2007) Landelijke Medische Registratie (LMR). www.prismant.nl 126. Stratford PW, Binkley JM (1999) Applying the results of self‐report measures to individual

patients: an example using the Roland‐Morris Questionnaire. J Orthop Sports Phys Ther 29:232‐239

127. Swinkels RAHM (2004) The ICF classification as a system for structuring outcome measurement. Physiotherapy Singapore 7:7‐13

128. Tezer M, Erturer RE, Ozturk C, Ozturk I, Kuzgun U (2005) Conservative treatment of fractures of the thoracolumbar spine. Int Orthop 29:78‐82

129. The EuroQol Group (1990) EuroQol: a new facility for the measurement of health‐related quality of life. Health Policy 16:199‐208

130. Tropiano P, Huang RC, Louis CA, Poitout DG, Louis RP (2003) Functional and radiographic outcome of thoracolumbar and lumbar burst fractures managed by closed orthopaedic reduction and casting. Spine 28:2459‐2465

131. Ustun TB, Chatterji S, Bickenbach J, Kostanjsek N, Schneider M (2003) The International Classification of Functioning, Disability and Health: a new tool for understanding disability and health. Disabil Rehabil 25:565‐571

132. Vaccaro AR, Lehman RA, Hurlbert RJ, Anderson PA, Harris M, Hedlund R, Harrop J, Dvorak M, Wood K, Fehlings MG, Fisher C, Zeiller SC, Anderson DG, Bono CM, Stock GH, Brown AK, Kuklo T, Öner FC (2005) A new classification of thoracolumbar injuries: the importance of injury morphology, the integrity of the posterior ligamentous complex, and neurologic status. Spine 30:2325‐2333

133. Vaccaro AR, Silber JS (2001) Post‐traumatic spinal deformity. Spine 26:S111‐S118 134. Vaccaro AR, Zeiller SC, Hulbert RJ, Anderson PA, Harris M, Hedlund R, Harrop J, Dvorak

M, Wood K, Fehlings MG, Fisher C, Lehman R, Anderson DG, Bono CM, Kuklo T, Öner FC (2005) The thoracolumbar injury severity score: a proposed treatment algorithm. J Spinal Disord Tech 18:209‐215

135. Verlaan JJ, Dhert WJ, Verbout AJ, Öner FC (2005) Balloon vertebroplasty in combination with pedicle screw instrumentation: a novel technique to treat thoracic and lumbar burst fractures. Spine 30:E73‐E79

136. Verlaan JJ, Diekerhof CH, Buskens E, van der Tweel I, Verbout AJ, Dhert WJ, Öner FC (2004) Surgical treatment of traumatic fractures of the thoracic and lumbar spine: a systematic review of the literature on techniques, complications, and outcome. Spine 29:803‐814

137. Verlaan JJ, van Helden WH, Öner FC, Verbout AJ, Dhert WJ (2002) Balloon vertebroplasty with calcium phosphate cement augmentation for direct restoration of traumatic thoracolumbar vertebral fractures. Spine 27:543‐548

138. Vialle LR, Vialle E (2005) Thoracic spine fractures. Injury 36 Suppl 2:B65‐B72

CHAPTER 1

30

139. Waddell G, Main CJ (1984) Assessment of severity in low‐back disorders. Spine 9:204‐208 140. Waddell G, Somerville D, Henderson I, Newton M (1992) Objective clinical evaluation of

physical impairment in chronic low back pain. Spine 17:617‐628 141. Ware JE, Sherbourne CD (1992) The MOS 36‐item short‐form health survey (SF‐36). I.

Conceptual framework and item selection. Med Care 30:473‐483 142. Watson‐Jones R (1943) Fractures and Joint Injuries. Third edition. E&S Livingstone Ltd.,

Edinburgh 143. Weinstein JN, Collalto P, Lehmann TR (1988) Thoracolumbar “burst” fractures treated

conservatively: a long‐term follow‐up. Spine 13:33‐38 144. White AA, Panjabi MM (1990) Clinical biomechanics of the spine. Lippincott, Philadelphia 145. Whitesides T (1977) Traumatic kyphosis of the thoracolumbar spine. Clin Orthop Relat Res

128:78‐92 146. Wood K, Butterman G, Mehbod A, Garvey T, Jhanjee R, Sechriest V (2003) Operative

compared with nonoperative treatment of a thoracolumbar burst fracture without neurological deficit. A prospective, randomized study. J Bone Joint Surg Am 85‐A:773‐781

147. Wood KB, Khanna G, Vaccaro AR, Arnold PM, Harris MB, Mehbod AA (2005) Assessment of two thoracolumbar fracture classification systems as used by multiple surgeons. J Bone Joint Surg Am 87:1423‐1429

148. World Health Organization (1946) Constitution of the World Health Organization. WHO, Geneva

149. World Health Organization (1980) International Classification of Impairments, Disabilities and Handicaps. WHO, Geneva

150. World Health Organization (2001) International Classification of Functioning, Disability and Health Problems. WHO, Geneva

151. Yi L, Jingping B, Gele J, Baoleri X, Taixiang W (2006) Operative versus non‐operative treatment for thoracolumbar burst fractures without neurological deficit. Cochrane Database Systematic Review: CD005079

152. Young PC, Petersilge CA (1999) MR imaging of the traumatized lumbar spine. Magn Reson Imaging Clin N Am 7:589‐602

31

Chapter 2

Spinal mobility: Sagittal range of motion measured with the SpinalMouse, a new non‐invasive device RB Post, VJM Leferink Archives of Orthopaedic and Trauma Surgery (2004) 124:187‐192

CHAPTER 2

32

Abstract

Introduction: In this paper the SpinalMouse, a new computerized external device for measuring sagittal spinal range of motion (ROM), was tested for inter‐rater reliability and use in clinical practice. Materials and methods: To assess inter‐rater reliability, two investigators each measured 111 subjects. Results: Correlation coefficients were found to be r=0.90 for flexion, r=0.85 for extension and r=0.90 for total inclination. Intra‐class coefficients were 0.95 for flexion, 0.92 for extension and 0.95 for total inclination. A poor agreement (kappa=0.22) was found for the presence of outliers from normal values for intersegmental ROM. Conclusion: We conclude the device is a useful, reliable tool for measuring sagittal spinal ROM in clinical practice, considering the small load it confers on patients and the short amount of time the measurement involves. The SpinalMouse might be more accurate after following the recommendations we make.

SPINALMOUSE

33

Introduction

During the past 30 years, numerous studies have been published about spinal range of motion (ROM), and many methods of measurement have been described. One of the most frequently used methods involves radiological analysis which can be performed in multiple ways (CT, plain‐ and biplanar radiography) [4, 6, 8, 10, 11, 12, 19, 22, 24]. A limitation of this method is the relatively high dose of radiation it requires, which precludes the use of the radiograph as a routine measurement in clinical practice. Therefore, many non‐invasive, external methods have been developed: goniometers, skin markers, inclinometers, spondylometers, measurement of back surface curvature and opto‐electronic systems, some of them computer aided, each with its specific characteristics [2, 9, 14, 16, 17, 21, 22]. External methods for measuring spinal ROM are now commonly used because they are easy to apply, non‐invasive and take little clinical time [21]. We used the SpinalMouse to measure sagittal spinal ROM. It is a non‐invasive electronic computer‐aided device, which is manually guided paravertebrally along the spinal column. It measures spinal ROM as well as intersegmental ROM. A relatively unique feature of the device is that it measures thoracic, lumbar and sacral/hip mobility separately. Data about the SpinalMouse have not yet been published. This study was designed to test the SpinalMouse’s inter‐rater reliability as well as to judge the device on its merits in clinical practice.

Materials and methods

Measurement of spinal ROM was done with the SpinalMouse (Idiag, Volkerswill, Switzerland), an electronic computer‐aided measuring device, which measures sagittal spinal ROM and intersegmental angles in a non‐invasive way, a so‐called surface‐based technique (Fig. 1). The device is connected radiographically via an analog‐digital converter to a standard PC. After supplying basic data of the patient, including height, weight, sex and age, the SpinalMouse is run paravertebrally along the spinal column from C7 to the rima ani (S3). The patient is asked to take three consecutive positions: erect, in maximal flexion and maximal extension of the spine. In each position a measurement is performed. When manually guided paravertebrally along the spine of a subject, the system records the outline of the skin over the spinal column in the sagittal plane. The local angle or inclination relative to a perpendicular line is given at any position by

CHAPTER 2

34

an internal pendulum connected to a potentiometer. An “intelligent recursive algorithm” computes information concerning the relative position of the vertebral bodies of the underlying bony spinal column. Raw data of the SpinalMouse measurements are the superficial back length from C7 to S3 and the local angle of each point of this length relative to the plumb line. In this manner, spinal ROM and 17 segments (Th1/2‐L5/S1) are evaluated.

Fig. 1 The SpinalMouse is run paravertebrally from C7 to S3

All measurements are summarized in a table, which consists of six columns: the first three columns contain the values for posture in the sequence upright, flexion and extension (kyphotic angles are expressed as positive values, lordotic angles as negative values). The last three columns refer to the calculated mobility in the sequence flexion minus upright, upright minus extension, and flexion minus extension (the latter being equal to ROM). At the foot the contribution of the thoracic spine, lumbar spine and sacral spine/hip to total mobility is expressed, as well as length of the back (in mm) and inclination (inclination as flexion minus extension is similar to total inclination) (Table 1). Figure 2 shows the superficial back shape; in the right corner the inclination is drawn, which represents full mobility of the trunk, composed of thoracic, lumbar and sacral/hip mobility.

SPINALMOUSE

35

Table 1 Data as stored and supplied by the software of the SpinalMouse (in deg). Static test data are angles measured in three positions, ROM data are calculated from the static test data (Sac/Hip=sacral/hip, Thor.Sp=thoracic, Lum.Sp=lumbar, Inclin=inclination, Lth=length of the back in mm; kyphotic values are positive, lordotic values are negative)

Static test data Range of motion data Segment Upright Flexion Extension Flex‐upr Upr‐ext Flex‐ext

Th1/2 9 0 4 ‐8 5 ‐3 Th2/3 6 4 3 ‐2 3 0 Th3/4 4 5 9 0 ‐4 ‐4 Th4/5 3 3 5 0 ‐2 ‐2 Th5/6 4 1 3 ‐3 1 ‐2 Th6/7 5 4 6 ‐1 ‐2 ‐2 Th7/8 4 6 4 2 0 2 Th8/9 5 8 3 3 2 5 Th9/10 3 13 4 11 ‐1 9 Th10/11 ‐1 8 4 9 ‐4 5 Th11/12 ‐3 5 ‐3 7 1 8 Th12/L1 ‐1 6 ‐3 7 1 9 L1/2 ‐1 7 ‐1 8 0 8 L2/3 ‐2 10 ‐4 12 2 13 L3/4 ‐4 10 ‐7 15 3 17 L4/5 ‐4 4 ‐7 8 3 11 L5/S1 ‐7 1 ‐7 8 ‐1 7 Sac/Hip 10 57 ‐8 47 18 65 Thor. Sp 39 58 41 19 ‐3 16 Lum. Sp ‐21 37 ‐29 58 8 66 Inclin 0 103 ‐23 103 23 126 Lth 569 704 557 135 12 147

CHAPTER 2

36

Fig. 2 Sagittal shape of the back and inclination in flexed, upright and extended position (same patient as Table 1)

The developer of the SpinalMouse claims to have measured 180 healthy volunteers without a history of back complaints (using a preliminary version of the device) and so created a database of age‐ and gender‐specific normal values for intersegmental ROM. The SpinalMouse software (version 2.3) supplies a graph in which the calculated intersegmental ROM is compared with these normal values (Fig. 3). Flexion ‐ upright

Upright ‐ extension

Flexion ‐ extension

Th1/2 Th 2/3 Th 3/4 Th 4/5 Th 5/6 Th 6/7 Th 7/8 Th 8/9 Th9/10 Th 10/11 Th 11/12 Th 12/L1L1/2 L2/3 L3/4 L4/5 L5/S1

Fig. 3 Graphic demonstration of the calculated measurements of intersegmental ROM, compared to age‐ and gender‐specific normal values

SPINALMOUSE

37

Study group The study group consisted of 111 subjects aged between 21 and 60 years old (mean age 39.2 yrs, mean height 1.78 m, mean weight 79.3 kg, mean BMI 24.7 kg/m2; 75 men, 36 women). Forty‐two were healthy volunteers, 69 subjects had sustained a spinal fracture (35 treated conservatively and 34 treated surgically). The spinal fracture patients were included because we used their ROM in another study [23]. All spinal fracture patients sustained their fracture at least 1.5 years previously; none of them had any neurological deficit. Measurement To assess inter‐rater reliability, two investigators (RBP and VJML) measured the patients in succession. Observer one completed a full measurement (patient upright, in flexion and extension) after which observer two performed a full measurement. Subjects were asked to try to touch their toes with their fingers with their knees straight, with the neck slightly flexed and their feet approximately 30 cm apart. For extension, subjects were asked to extend their back as far as possible, without external help. The SpinalMouse was placed at C7 (found by palpation after instructing the patient to bend the head slightly to the chest) and manually guided to the rima ani in a paravertebral manner. No “warming up” was performed before the measurement. The following was tested for inter‐rater reliability: inclination in flexion; inclination in extension; total inclination; measured length of the back as flexion minus extension. Also, the presence of outliers from normal values for intersegmental ROM was tested for inter‐rater reliability. Statistical analysis Statistical analysis was done with SPSS version 10 (SPSS Chicago, IL, USA). Length of the back and inclination were tested by means of the intra‐class correlation coefficient (ICC) and Pearson’s correlation coefficient r [20, 25]; mean values measured by the two investigators were compared by means of the paired t‐test. No universally accepted levels have been adopted for expressing the reliability of measurements when ICC values are calculated [13]. However, one proposed scheme for defining the amount of reliability with ICC’s has the following values: 0.99‐0.90 high reliability; 0.89‐0.80 good reliability; 0.79‐0.70 fair reliability and 0.69 and below poor reliability [18]. The presence of outliers from normal values for intersegmental ROM was tested by means of Cohen’s kappa. Cohen’s kappa measures the agreement between the evaluations of two raters when both are rating the same object. A value of 1 indicates perfect agreement, a value of 0 indicates that agreement is no better than chance. A p‐value of 0.05 was considered significant.

CHAPTER 2

38

Results

Correlation coefficients and ICC’s are shown in Table 2, as well as mean values measured by the two observers.

Table 2 Intra‐class correlation coefficient (ICC) and Pearson’s correlation coefficient r for total inclination, inclination in flexion, inclination in extension and length of the back

ICC Pearson’s r Mean obs. one vs obs. two

Total inclination 0.95 0.90 (p<0.001) 118.5 vs 116.4 deg (p<0.05) Flexion 0.95 0.90 (p<0.001) 92.3 vs 92.4 deg (p=0.830) Extension 0.92 0.85 (p<0.001) ‐26.3 vs ‐23.9 deg (p<0.05) Length of the back: flex ‐ ext 0.76 0.61 (p<0.001) 119.4 vs 112.6 mm (p<0.05)

For inclination (total inclination, flexion and extension) ICC’s were 0.92 and 0.95; all Pearson’s correlation coefficients were significant (p<0.001). The relationship between total inclination as measured by observer one and observer two is shown in Figure 4. Observer one measured a higher mean total inclination than observer two (p<0.05), and mean extension as measured by observer one was higher than that measured by observer two (p<0.05). There was no significant difference between mean flexion as measured by both observers (p=0.830).

Total inclination (deg) obs. two

200180160140120100806040

Total inclin

ation (deg) o

bs. one

200

180

160

140

120

100

80

60

40

Fig. 4 Total inclination by observer one plotted against that by observer two

SPINALMOUSE

39

Reliability for measurement of the length of the back as flexion – extension was as follows: r=0.61 (p<0.001), ICC=0.76. Mean values measured by observer one were higher than those measured by observer two (p<0.05). Figure 5 shows the relationship between measured length of the back (flexion – extension) between observer one and observer two. Cohen’s kappa for the presence of outliers from normal values for intersegmental ROM was 0.22.

Length of the back (mm) flex ‐ ext obs. two

2402101801501209060300

Leng

th of the back (m

m) flex ‐ ext obs. one

240

210

180

150

120

90

60

30

0

Fig. 5 Length of the back as flexion minus extension: observer one plotted against observer two

Discussion

This study was designed to test a new device for measuring sagittal spinal ROM. We chose to do this by means of inter‐rater reliability. We did not assess intra‐rater reliability, because that would mean patients would have to bow and extend two more times, and should be done on separate days, which was not possible for time‐ and logistic reasons. Inclination in flexion, inclination in extension and total inclination were found to be highly reliable; however, some differences were found between the mean results. When taking into account that both measurements were done successively (so the first measurement might influence the second), and the fact that patients might not take exactly the same position during two consecutive measurements, some differences in results were to expected. As mentioned by Portek et al., subject

CHAPTER 2

40

repeatability is a major contributing factor in the measurement of spinal mobility [22]. A significant difference for extension and total inclination was found. Considering extension, a large part of our study group encountered problems (e.g. maintaining balance) when standing in extension. As reported in the literature, extension in standing is uncomfortable, and the subject may find it difficult to maintain balance [3, 21]. This might explain the mean difference for extension (2.4 deg) found between the two investigators. Besides this subject repeatability, the difference might also result from the device itself, or a cumulative effect of these two entities. For total inclination, a similar difference was found (2.1 deg). However, it can be questioned whether such a small difference is a clinically important finding. Inter‐rater reliability for the difference of the back’s length from flexion to extension was found to be fair. There was a small but significant difference between the two investigators (6.8 mm). As pointed out by Mayer et al. in a previous study, variability among examiners in locating bony landmarks is a major contributing factor in the external measurement of spinal ROM [15]. When performing one complete measurement, C7 needs to be palpated three times and there has to be stopped at S3 three times, which makes a total of six possible errors per measurement. Accumulation of these possible errors might be an explanation for the difference found in measured length. Considering this, a possibility might be to mark C7 and S3 with a skin marker after palpating both precisely. By doing so, the length of the back might be measured more accurately. However, during flexion these skin markers will shift in relation to the underlying vertebra. Another possible explanation for the difference found is a slightly different pathway taken along the spinal curvature. A poor agreement was found for the presence of outliers from normal values for intersegmental ROM (kappa 0.22). We cannot, however, ascertain whether the developer’s graph with normal intersegmental ROM data was obtained in a reliable way, since no accompanying publication is available. For now, we conclude that the SpinalMouse is not very reliable in measuring intersegmental ROM. Comparison of our results with the literature is difficult because no studies have been published concerning the SpinalMouse so far. In a study by Keeley et al., using a two‐inclinometer technique that is to some extent comparable, inter‐rater correlation was found to be good; r=0.92 (p<0.001) [7]. However, in that study only lumbar and hip mobility were measured, whereas thoracic mobility was not.

SPINALMOUSE

41

Chen et al. found a computerized single‐sensor inclinometer to have a poor inter‐rater reliability (ICC 0.39 for extension and ICC 0.69 for flexion) [1]. In another study the inter‐rater reliability for the CA‐6000 spine motion analyzer (a computerized potentiometer) was as follows: flexion r=0.76; extension r=0.84; flexion+extension r=0.84; compared with these data our results seem favourable. However, only lumbar ROM was measured in that study [5]. The data we report for sagittal spinal ROM can not be seen as normal values. First, we measured patients with a spinal fracture in the past as well as healthy volunteers. In addition, we used inclination for comparison, which, as pointed out before, consists of thoracic, lumbar as well as sacral/hip mobility. For assessment of inter‐rater reliability, this is not important, but it should be taken into account when comparing data with earlier studies, which mostly report on lumbar mobility. Mellin reports on thoracolumbar ROM in his study concerning the Myrin inclinometer, but he did not measure hip mobility, and his population consisted of 25 healthy volunteers, so comparison of our data to his data is not reliable [17]. To validate the ability of the device to measure sagittal spinal ROM and intersegmental ROM, a comparison between the SpinalMouse’s data and radiographs could have been made. It is questionable, however, whether radiographs should be chosen as the “gold standard”. The use of radiographs is probably unjustifiable in terms of patient risk and cost (especially when, as in this study, the whole spinal column needs to be assessed). Because of this, and the fact that 2/3 of our study group consisted of patients with a spinal fracture in the past, we have not been able to validate the SpinalMouse. Clinical use When using the SpinalMouse we encountered a number of pitfalls. When measuring subjects much taller than the investigator, it is difficult to place the SpinalMouse exactly on C7 and to see the LED on the mouse, which indicates that the device is ready for use. To overcome these problems, it might be useful to have a little stool for the investigator to stand on. Another problem we encountered was controlling the skin surface in lumbar lordosis in extension. Sometimes, the SpinalMouse could not pass the lumbar angle properly because the angle was too sharp, and the wheels of the mouse slipped as a result. Concerning the clinical time and the user‐friendliness, we are fairly satisfied. One complete measurement takes about 1 minute, the device is not velocity sensitive, and after some training the SpinalMouse is rather simple to use.

CHAPTER 2

42

Limitations Of course, there are some limitations concerning this study. Investigators were not blinded to the subject’s diagnosis. Another issue that should be kept in mind is that comparison with the literature is not possible, since no publication concerning the SpinalMouse is available.

Conclusions

The SpinalMouse seems to be a good, reliable device for measuring sagittal spinal ROM, as tested by inter‐rater reliability. After following the recommendations we made, the device might be more accurate. Correlation coefficients found were fair to good; there was a significant difference between mean total inclination as well as mean extension as measured by the two investigators, but this difference was so small that it would not be of major influence in clinical practice. For measuring intersegmental ROM, the SpinalMouse does not seem to be a reliable tool yet. Considering the short clinical time needed for measurement and the low health risk posed to the patient, the SpinalMouse could be used as a reliable objective tool for measuring sagittal spinal ROM. Intra‐rater reliability needs to be assessed in a further study.

References

1. Chen SP, Samo DG, Chen EH, Crampton AR, Conrad KM, Egan L, Mitton J (1997) Reliability of three lumbar sagittal motion measurement methods: surface inclinometers. J Occup Environ Med 39:217‐223

2. Chiou WK, Lee YH, Chen WJ, Lin YH (1996) A non invasive protocol for the determination of lumbar spine mobility. Clin Biomech 11:474‐480

3. Dillard J, Trafimow J, Andersson GB, Cronin K (1991) Motion of the lumbar spine. Reliability of two measurement techniques. Spine 16:321‐324


5. Dopf CA, Mandel SS, Geiger DF, Mayer PJ (1994) Analysis of spine motion variability using a computerized goniometer compared to physical examination. A prospective clinical study. Spine 19:586‐595


7. Keeley J, Mayer TG, Cox R, Gatchel RJ, Smith J, Mooney V (1986) Quantification of lumbar function. Part 5: Reliability of range‐of‐motion measures in the sagittal plane and an in vivo torso rotation measurement technique. Spine 11:31‐35

SPINALMOUSE

43


9. Lee YH, Chiou WK, Chen WJ, Lee MY, Lin YH (1995) Predictive model of intersegmental mobility of lumbar spine in the sagittal plane from skin markers. Clin Biomech 10:413‐420

10. Leferink VJM, Zimmerman KW, Veldhuis EFM, ten Vergert EM, ten Duis HJ (2001) Thoracolumbar spinal fractures: radiological results of transpedicular fixation combined with transpedicular cancellous bone graft and posterior fusion in 183 patients. Eur Spine J 10:517‐523

11. Lim TH, Eck JC, An HS, McGrady LM, Harris GF, Haughton VM (1997) A noninvasive, three‐dimensional spinal motion analysis method. Spine 22:1996‐2000

12. Lindsey RW, Dick W, Nunchuck S, Zach G (1993) Residual intersegmental spinal mobility following limited pedicle fixation of thoracolumbar spine fractures with the fixateur interne. Spine 18:474‐478

13. Madson TJ, Youdas JW, Suman VJ (1999) Reproducibility of lumbar spine range of motion measurements using the back range of motion device. J Orthop Sports Phys Ther 29:470‐477

14. Mannion A, Troke M (1999) A comparison of two motion analysis devices used in the measurement of lumbar spinal mobility. Clin Biomech 14:612‐619

15. Mayer RS, Chen IH, Lavender SA, Trafimow JH, Andersson GB (1995) Variance in the measurement of sagittal lumbar spine range of motion among examiners, subjects, and instruments. Spine 20:1489‐1493


17. Mellin G (1986) Measurement of thoracolumbar posture and mobility with a Myrin inclinometer. Spine 11:759‐762

18. Meyers CR, Blesh TE (1962) Measurement in physical education. Ronald Press Co, New York, NY

19. Miyasaka K, Ohmori K, Suzuki K, Inoue H (2000) Radiographic analysis of lumbar motion in relation to lumbosacral stability. Investigation of moderate and maximum motion. Spine 25:732‐737

20. Muller R, Buttner P (1994) A critical discussion of intraclass correlation coefficients. Stat Med 13:2465‐2476


22. Portek I, Pearcy MJ, Reader GP, Mowat AG (1983) Correlation between radiographic and clinical measurement of lumbar spine movement. Br J Rheumatol 22:197‐205

23. Post RB, Leferink VJM (2004) Sagittal range of motion after a spinal fracture: does ROM correlate with functional outcome? Eur Spine J 13:489‐494

24. Schlegel JD, Smith JA, Schleusener RL (1996) Lumbar motion segment pathology adjacent to thoracolumbar, lumbar, and lumbosacral fusions. Spine 21:970‐981

25. Shrout PE, Fleiss JL (1979) Intraclass correlations: Uses in assessing rater reliability. Psychological Bulletin 86:420‐428



45

Chapter 3

Sagittal range of motion after a spinal fracture: does ROM correlate with functional outcome? RB Post, VJM Leferink European Spine Journal (2004) 13:489‐494

CHAPTER 3

46

Abstract

Introduction: Literature regarding the effect of a spinal fracture and its treatment in terms of resulting spinal range of motion (ROM) is scarce. However, there is need for data regarding sagittal spinal ROM, since many patients who sustain a spinal fracture are concerned about the back mobility they will have after treatment. In addition, the relationship between ROM and impairment is not clear. The literature gives conflicting results. Methods: To study spinal ROM after a spinal fracture, we measured thoracolumbar ROM in operatively and non‐operatively treated patients (n=76, average 3.7 years follow‐up) as well as controls (n=41). In order to study the relation between ROM and subjective back complaints, we calculated the correlation between thoraco‐lumbar ROM and scores derived from the VAS spine score and RMDQ. To assess impairment after a spinal fracture, we compared VAS and RMDQ scores between operatively and non‐operatively treated patients and healthy controls. Results: Operatively treated patients were found to have lower thoracolumbar ROM than controls (56.7° vs 70.0°, respectively; p<0.01). There was no difference between operatively treated and non‐operatively treated patients (56.7° vs 62.7°, respectively); nor was a difference found between non‐operatively treated patients and controls. Correlation between ROM and subjective impairment was very weak and only significant for ROM and RMDQ scores in the whole study group (rho= –0.25; p<0.01). Patients were more impaired than controls, there was no difference between operatively and non‐operatively treated patients (VAS score 76.3 vs 72.6; RMDQ score 4.5 vs 4.4, respectively). Conclusion: We conclude that patients treated operatively for a thoracolumbar spinal fracture have a lower thoracolumbar ROM than controls. Spinal ROM, however, does not influence impairment. A spinal fracture results in impairment, no matter what therapy is chosen.

RANGE OF MOTION

47

Introduction

The effect of treatment of a spinal fracture on mobility of the spinal column and resulting range of motion (ROM) is uncertain. Literature about total spinal ROM after a fracture is scarce, as most studies address intersegmental ROM [5, 15, 17, 25]. The few studies available concerning spinal ROM after a spinal fracture reveal conflicting results. Axelsson et al. found, in patients treated with a posterolateral fusion for spondylolysis or facet joint arthritis, that the sagittal lumbar ROM increased after a fusion, probably due to relief of protective muscle spasm [1]. Dodd et al. found that spinal ROM does not return to normal after Harrington rod removal in patients treated operatively for a thoracolumbar fracture [8]. In a study by Junge et al., sagittal spinal ROM was found to be normal 2.5 years after operative treatment for a thoracolumbar spinal fracture [9]. There is need for data on spinal ROM in patients treated for a thoracolumbar spinal fracture. In order to study how a spinal fracture and its treatment affect spinal ROM, we measured sagittal thoracolumbar ROM in operatively and non‐operatively treated patients. For comparison, ROM was also measured in a control group consisting of healthy volunteers. As the ROM after a spinal fracture is still uncertain, little is known about the influence of the resulting spinal ROM on the patients’ overall functional outcome, measured in terms of subjective impairment. Poitras et al. stated that thoracolumbar ROM is poorly to moderately related to functional disabilities [24]. Nattrass et al. found that there was no relationship between ROM and impairment [23]. In contrast, a study by Cox et al. reports a significant correlation between sagittal lumbar ROM and impairment [3]. In order to assess the relationship between subjective impairment and spinal ROM, and to reveal whether operative and non‐operative treatment result in different impairment rates, we asked participants to fill in two questionnaires, the Visual Analogue Scale (VAS) spine score and the Roland‐Morris Disability Questionnaire (RMDQ).

The following questions were studied: • Is there a difference in sagittal spinal ROM between operatively treated

patients, non‐operatively treated patients and controls? • Do the average VAS and RMDQ scores differ between operatively treated

patients, non‐operatively treated patients and controls? • Does sagittal spinal ROM correlate with subjective impairment, measured by

the VAS and RMDQ?

CHAPTER 3

48

Materials and methods

Patients Between January 1996 and December 2000, 254 patients with a fracture of the thoracolumbar spine were treated at the traumatology department of the University Hospital Groningen. One hundred and ten patients (mean age 38.0 years) were treated operatively; 144 (mean age 42.4 years) were treated non‐operatively. 153 (60%) patients (74 treated operatively, 79 non‐operatively) met the inclusion criteria (see Table 1). Patients operated on for an A3.3 fracture with the implant still in situ were not included, because the implant would influence the paravertebral measurement.

Table 1 Inclusion and exclusion criteria

Inclusion criteria Exclusion criteria Spinal fracture between T1 and L5 Pathological fracture Age at follow‐up between 18 and 60 years Neurological deficit Time since injury at follow‐up > 1.5 years Psychiatric illness Capable of understanding the Dutch language A3.3 fracture with implant in situ From the 153 included, 125 randomly selected patients (82%) were sent a letter in which the aim of the study was described. Four to 8 days later, an investigator telephoned patients to ask them to participate. Twelve patients were lost to follow‐up; 14 could not be reached. Twelve refused, giving such reasons as “not interested” or “no time”. Eleven patients missed several appointments. In total, 76 (38 treated operatively, 38 non‐operatively) participated in the study (response rate: 76/125=61%; follow‐up: 76/254=30%). Respondents did not differ from non‐respondents for age, gender and time since injury. For both groups, the participating patients did not differ from the non‐participating patients for age, gender and time since injury. The control group consisted of 41 healthy volunteers (without a history of back surgery or medically treated back complaints) from a normal population of hospital personnel. The three groups did not differ from each other for age, gender and number. Average time from injury to follow‐up was 3.7 years (range 1.7–6.4 years). Time since injury was significantly shorter for non‐operatively treated patients (p<0.01). The non‐operatively treated group consisted of more type A fractures and fewer type B and type C fractures, according to the comprehensive classification (CC) [19] than the operatively treated group (see Table 2).

RANGE OF MOTION

49

Table 2 Study‐group patient descriptions (n=117): age, gender, fracture level, follow‐up and type of fracture (according to the CC [19]; nc not classified)

fracture type Treatment n age (yrs) mean (SD) range

gender (M:F)

fracture level

follow‐up (yrs) mean (SD) range A B C nc

Operative 38 40.5 (12.0) 21‐59 25:13 T9‐L5 4.1 † (1.1) 2.5‐6.4 23 10 3 2 Nonoperative 38 40.6 (11.3) 23‐59 22:16 T4‐L5 3.3 † (1.2) 1.7‐5.6 31 2 1 4 Controls 41 39.1 (10.5) 23‐60 28:13 ‐‐‐‐‐‐‐‐ ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ ‐‐ ‐‐ ‐ ‐ † difference significant p<0.01 Treatment Operative treatment consisted of fracture reduction and fixation by means of dorsal instrumentation with the Universal Spine System (Synthes, Oberdorf, Switzerland), combined with transpedicular cancellous bone grafting and dorsal spondylodesis as described by Daniaux and Dick [4, 6, 7]. Fracture reduction was obtained by indirect manipulation using pedicle screws as levers. Cancellous bone was taken from the dorsal iliac crest and put in the reduced vertebral body transpedicularly [4]. The facet joints at the level of the traumatized disc were opened and the cartilage was removed. Cancellous bone was packed around the joints at the dorsolateral side [2]. No ventral operations, discectomies or laminectomies were performed. Postoperatively, all patients were transferred to a rehabilitation centre. They were allowed to walk after about 10 days in a simple reclination brace, worn for 9 months, after which the implants were removed. Three months later, patients were instructed to resume all former activities. A more detailed description of the operative technique was published previously [14]. Non‐operative treatment was initialized in our hospital and continued in a rehabilitation centre or the outpatient clinic. Therapy consisted of bed rest, sometimes on a Stryker frame, for a maximum of 6 weeks, followed by mobilization with a reclination brace for 9 months, combined with physiotherapy. Most A1 and A2 fractures (according to the CC [19]) were treated with a short period of bed rest, followed by functional treatment without a brace. Measurement We used the SpinalMouse (Idiag, Volkerswill, Switzerland), a computer‐aided device, for measuring sagittal spinal ROM and the intersegmental angle in a non‐invasive manner (Fig.1). The device is connected via an analog‐digital converter to a standard PC. Manually guided along the back of a subject, the system records the outline of the spinal column in the sagittal plane. To measure spinal ROM, the

CHAPTER 3

50

SpinalMouse is run paravertebrally along the spinal column from C7 to the rima ani (S3). The local angle or inclination relative to a perpendicular line is given at any position by an internal pendulum connected to a potentiometer. The ROM of each segment (i.e., intersegmental ROM) is computed, from which the relative parts of ROM for thoracic spine, lumbar spine and sacral spine/hip are computed. A more detailed description has been accepted for publication (R.B. Post and V.J.M. Leferink, 2004, Arch Orthop Trauma Surg).

Fig.1 The SpinalMouse is run paravertebrally from C7 to S3

We measured thoracolumbar ROM (T1/2–L5/S1) by adding thoracic ROM to lumbar ROM. Two investigators in succession (RBP and VJML) measured the patient’s back. Participants were asked to bend and extend as far as possible, with their knees straight, without “warming up”. In this manner, two measurements were obtained from each patient. The ROM we used was the average thoracolumbar ROM obtained from the two measurements. With regard to subjective impairment, we asked participants to fill in two questionnaires measuring back pain and restrictions: the Roland‐Morris disability questionnaire (RMDQ) and the VAS spine score (VAS). The RMDQ is a health status measure designed to be completed by patients to assess physical disability due to low back pain. It is self‐administered and takes less than 5 min to be completed [27]. Total scores can vary from 0 (no disability) to 24 (severe disability). The RMDQ has been used extensively and was found to be a sensitive, reliable and valid instrument [12, 13, 16, 27, 28]. In this study the Dutch version of the RMDQ was used [26].

RANGE OF MOTION

51

The VAS spine score, developed for use with spinal fracture patients, asks the patient to rate the functional outcome in 19 items on an analogue 10 cm visual scale. The patient’s perception of pain and restriction in activities related to back problems is measured. Higher scores represent better results, recalculated to percentages of the maximum score (0–100%). In previous studies, it has proved to be a reliable and valid instrument [11, 16]. Statistical analysis Statistical analysis was done with SPSS version 10 (SPSS, Chicago). Comparison of VAS and RMDQ scores and ROM between groups was done by means of one‐way ANOVA (posthoc Bonferroni). Correlation was computed by means of Pearson’s correlation coefficient r. RMDQ and VAS scores for the total study group did not show a normal distribution, so correlation between VAS scores and RMDQ scores and ROM for the total group was tested non‐parametrically, by means of Spearman’s rho. Significance was accepted at 0.05.

Results

ROM Operatively treated patients had lower thoracolumbar ROM than did controls (56.7° vs 70.0°, p<0.01). There was no difference found between operatively and non‐operatively treated patients (56.7° vs 62.7°, p=0.429) or between non‐operatively treated patients and controls (62.7° vs 70.0°, p=0.210), see Table 3.

Table 3 Thoracolumbar ROM (°; mean, SD and range) in operatively and non‐operatively treated patients and controls

Treatment mean SD range Operative 56.7 † 16.3 25.0 ‐ 88.5 Non‐operative 62.7 19.7 16.5 ‐ 105.5 Controls 70.0 † 16.7 38.5 ‐ 108.0

† difference significant p<0.01 VAS and RMDQ scores Comparison of VAS and RMDQ scores showed that the mean VAS score in operatively and non‐operatively treated patients was less than the mean VAS score of controls (p<0.001). VAS scores did not differ between operatively and non‐operatively treated patients (p=1.000). Operatively as well as non‐operatively treated patients had a higher mean RMDQ score than did controls (p<0.001). Mean

CHAPTER 3

52

RMDQ scores did not differ between operatively treated patients and non‐operatively treated patients (p=1.000), see Table 4. A Spearman’s rho of ‐0.85 (p<0.001) was found for correlation between the VAS score and RMDQ score for the whole study group.

Table 4 VAS and RMDQ scores in operatively and non‐operatively treated patients and controls

Treatment VAS mean SD range

RMDQ mean SD range

Operative 76.3 † 23.3 21.6‐100.0 4.5 † 5.2 0‐17.0 Non‐operative 72.6 ‡ 22.9 22.8‐100.0 4.4 ‡ 4.3 0‐12.0 Controls 92.8 †‡ 9.2 50.2‐100.0 0.5 †‡ 1.4 0‐7.0

† difference significant p<0.001 ‡ difference significant p<0.001 Correlation between ROM and VAS, RMDQ A significant correlation was found between ROM and RMDQ score for the whole study group (Spearman’s rho= ‐0.25, p<0.01). None of the other correlation coefficients was significant (see Table 5).

Table 5 Correlation between ROM, VAS score and RMDQ score in separate groups and total study group

Correlation with ROM Treatment VAS RMDQ

Operative r= ‐0.001 (p=0.99) r= ‐0.19 (p=0.25) Non‐operative r= 0.12 (p=0.50) r= ‐0.23 (p=0.18) Controls r= 0.16 (p=0.35) r= ‐0.10 (p=0.54) Whole study group (Spearman) rho= 0.16 (p=0.08) rho= ‐0.25 (p=0.007) †

† significant p<0.01

Discussion

This study was designed to evaluate sagittal spinal ROM of patients who sustained a thoracolumbar spinal fracture, as well as to study functional outcome and the relation between sagittal ROM and functional outcome.

ROM Our results show that operatively treated patients have a lower thoracolumbar ROM than do controls. ROM did not differ significantly between non‐operatively treated patients and controls; nor was there a significant difference in ROM between operatively treated and non‐operatively treated patients. In these series

RANGE OF MOTION

53

the only statistically significant difference in ROM was found between operatively treated patients and controls. The differences between the other groups were not significant. An explanation could be found in, e.g., the power. The only conclusion possible based on these findings, however, is that operatively treated patients have lower ROM than do controls. Only a few studies regarding this issue have been published, making it difficult to compare our results with literature. In a study by Junge et al. sagittal spinal ROM was found to be normal after mono‐segmental operative treatment of a spinal fracture [9]. ROM was measured by means of finger‐to‐floor distance (11.6 cm) and the Schober technique (10:13.9 cm), as well as a clinical examination. Finger‐to‐floor distance, however, measures gross mobility of the trunk, which is mainly composed of hip movement [20, 30]. Consequently, finger‐to‐floor distance does not seem to be a valid tool for measuring thoracolumbar ROM. The Schober technique [29], although popular, has some important deficiencies: spinal extension and movement in the upper lumbar/lower thoracic region are not assessed [22]. Although Junge states that ROM was within normal range, he did not mention the normal values for either method. Reported values of 111° for flexion and 37° for extension seem to us values representing total trunk mobility, which does not represent thoracolumbar mobility [9]. In contrast to Junge, a recent study (concerning functional outcome of operatively treated patients) reports decreased thoracolumbar ROM 3 years after injury [10]. Thoracolumbar ROM was measured by finger‐to‐floor distance (11.6 cm) and the Schober technique (13.2 cm), as well as the Ott technique. No normal values for these methods were reported. As pointed out before, the first two methods do not represent true thoracolumbar mobility. The Ott technique consists, according to the author, of measuring the lengthening of a 30 cm distance, caudal to C7 in maximal spinal flexion [10]. However, no literature could be found regarding this technique, so it is not clear that it is reliable and valid for measuring thoracolumbar ROM. Taking into account the limitations of the papers discussed above, it is difficult to compare our results to literature. On the other hand, considering thoracolumbar ROM in normal subjects, a striking difference was found between our results and values reported in the literature (Mellin: 106°, Louis: 133°) [18, 21]. Why operatively treated patients have lower sagittal thoracolumbar ROM than controls is unknown. In our opinion, it seems unlikely that a single fusion is responsible for a decline in thoracolumbar range of motion. Fear of re‐fracture or pain might be a possible explanation, as recently mentioned by Cox et al. [3]. Psychological aspects, for example the impact an operation implies, may lead to

CHAPTER 3

54

less functional use of the back post‐traumatically, which might result in decreased ROM. Another explanation for the lower ROM could be in the invasiveness of the operation, which results in scar tissue formation. VAS and RMDQ scores Evaluation of subjective impairment reveals that patients are impaired after a spinal fracture. Both operatively treated patients and non‐operatively treated patients have a higher mean RMDQ score than do controls, as well as lower mean VAS scores. Both indicate more impairment. Scores between operatively treated and non‐operatively treated patients did not differ. These data indicate that a spinal fracture, regardless of its treatment, results in subjective impairment that is similar for both types of treatment. However, it should be taken into account that average time since injury was shorter for non‐operatively treated patients than for operatively treated patients, possibly biasing results. In literature, a VAS score of 66 is reported for patients treated operatively for a spinal fracture at a follow‐up of 23 months. A control group achieved scores of 92 [11]. For the control group, these data are comparable with our results, whereas operatively treated patients in our study achieve higher VAS scores. A possible explanation is our longer follow‐up. Recently, Leferink et al. studied functional outcome in patients treated operatively for a thoracolumbar burst fracture. In his study, a mean RMDQ score of 4.0 was found, together with a mean VAS score of 79 [16]. Our results were obtained from a different group of patients, but are comparable. Kraemer et al. found a mean RMDQ score of 15.6 after a follow‐up of 3.8 years, in patients treated operatively as well as non‐operatively for a thoracolumbar burst fracture [12]. As in our study, there was no difference in RMDQ scores between operatively and non‐operatively treated patients. Correlation between RMDQ and VAS was found to be good (rho= –0.85). Only one published study reports correlation between these two questionnaires, in which a correlation of –0.72 was found [16]. Correlation between ROM and VAS, RMDQ Another issue is whether ROM influences impairment. We found weak correlation between RMDQ score and ROM for the whole study group (rho= –0.25). Negative rho indicates that an increase in ROM is accompanied by a lower score on the RMDQ, indicating less impairment. However, correlation was very weak and correlation for separate groups was not significant, either. Consequently, it seems unlikely that ROM influences impairment. There is a growing amount of literature

RANGE OF MOTION

55

concerning the relationship between ROM and impairment. Poitras et al. found that kinematic variables, including thoracolumbar ROM, correlate moderately to poorly to disability, and do not appear to be a valid measure of disability [24]. In a study by Nattrass et al. no relationship was found between lumbar ROM (measured with a long‐arm goniometer and dual inclinometer) and impairment measured by the Oswestry Disability Index and the Waddell Disability Index [23]. These findings support our data that ROM is of no (or minor) influence on impairment. However, Nattras measured lumbar ROM, whereas in our series thoracolumbar ROM was measured. In contrast, a study by Cox et al. reports a significant correlation (r=0.52) between ROM and impairment measured by the Quebec Back Pain Disability Questionnaire [3]. The author states that simple parameters of the functional examination, such as ROM, are strongly correlated with the cognitive state. For example, fear will influence (voluntary) ROM [3]. Limitations A limitation of this study is that the average time since injury was shorter for non‐operatively treated patients than for operatively treated patients, which makes these two groups not completely comparable. The shorter follow‐up might have affected the results in some way. Another issue to keep in mind is the response rate, which might have biased results.

Conclusions

Sagittal thoracolumbar ROM 4 years after operative treatment of a spinal fracture seems to be lower than the thoracolumbar ROM of healthy individuals. It is unclear why operative treatment of thoracolumbar fractures might result in lower spinal ROM. Further research should be done in this field. Patients who sustained a spinal fracture are more impaired than healthy controls. ROM does not seem to influence this impairment, however. Both kinds of treatment (operative vs non‐operative) apparently result in similar impairment rates.

References

1. Axelsson P, Johnsson R, Stromqvist B, Arvidsson M, Herrlin K (1994) Posterolateral lumbar fusion. Outcome of 71 consecutive operations after 4 (2‐7) years. Acta Orthop Scand 65:309‐314

CHAPTER 3

56

2. Blauth M, Bastian L, Jeanneret B, Knop C, Moulin P, Müller‐Vahl H, Schmidt U, Schratt H E, Wippermann B (1998) Wirbelsäule. In: Tscherne H, Blauth M (eds) Tscherne Unfallchirurgie, vol 3. Springer, Berlin Heidelberg New York, pp 314‐320 and pp 333‐338


4. Daniaux H (1982) Technik und erste Ergebnisse der transpedikulären Spongiosaplastik bei Kompressionsbrüchen im Lendenwirbelsäulenbereich. Acta Chir Austr 43 [suppl]:79

5. Dekutoski MB, Conlan ES, Salciccioli GG (1993) Spinal mobility and deformity after Harrington rod stabilization and limited arthrodesis of thoracolumbar fractures. J Bone Joint Surg Am 75:168‐176


7. Dick W, Kluger P, Magerl F, Woersdorfer O, Zach G (1985) A new device for internal fixation of thoracolumbar and lumbar spine fractures: the “fixateur interne”. Paraplegia 23:225‐232


9. Junge A, Gotzen L, von‐Garrel T, Ziring E, Giannadakis K (1997) Die monosegmentale Fixateur interne‐Instrumentation und Fusion in der Behandlung von Frakturen der thorakolumbalen Wirbelsäule. Indikation, Technik und Ergebnisse. Unfallchirurg 100:880‐887




13. Leclaire R, Blier F, Fortin L, Proulx R (1997) A cross‐sectional study comparing the Oswestry and Roland‐Morris Functional Disability scales in two populations of patients with low back pain of different levels of severity. Spine 22:68‐71


15. Leferink VJM, Nijboer JMM, Zimmerman KW, Veldhuis EFM, ten Vergert EM, ten Duis HJ (2002) Thoracolumbar spinal fractures: segmental range of motion after dorsal spondylodesis in 82 patients: a prospective study. Eur Spine J 11:2‐7


17. Lindsey RW, Dick W, Nunchuck S, Zach G (1993) Residual intersegmental spinal mobility following limited pedicle fixation of thoracolumbar spine fractures with the fixateur interne. Spine 18:474‐478

18. Louis R (1983) Surgery of the spine: surgical anatomy and operative approaches. Springer‐Verlag, Berlin Heidelberg New York, p 70

RANGE OF MOTION

57


20. Mayer TG, Tencer AF, Kristoferson S, Mooney V (1984) Use of noninvasive techniques for quantification of spinal range‐of‐motion in normal subjects and chronic low‐back dysfunction patients. Spine 9:588‐595


22. Miller MH, Lee P, Smythe HA, Goldsmith CH (1984) Measurements of spinal mobility in the sagittal plane: new skin contraction technique compared with established methods. J Rheumatol 11:507‐511



25. Rohlmann A, Neller S, Bergmann G, Graichen F, Claes L, Wilke HJ (2001) Effect of an internal fixator and a bone graft on intersegmental spinal motion and intradiscal pressure in the adjacent regions. Eur Spine J 10:301‐308



28. Roland M, Morris R (1983) A study of the natural history of low‐back pain. Part II: development of guidelines for trials of treatment in primary care. Spine 8:145‐150

29. Schober P (1937) Lendenwirbelsäule und Kreuzschmerzen. Munch Med Wochenschr 84:336‐338

30. Winter RB, Carr P, Mattson HL (1997) A study of functional spinal motion in women after instrumentation and fusion for deformity or trauma. Spine 22:1760‐1764


59

Chapter 4

Functional outcome 5 years after non‐operative treatment of type A spinal fractures RB Post, HJE Keizer, VJM Leferink, CK van der Sluis European Spine Journal (2006) 15:472‐478

CHAPTER 4

60

Abstract

Introduction: This study was conducted to study the functional outcome after non‐operative treatment of type A thoracolumbar spinal fractures without neurological deficit. Functional outcome was determined following the International Classification of Functioning, Disability and Health (ICF), measuring restrictions in body function and structure, restrictions in activities, and restrictions in participation/quality of life. Methods: All patients were treated non‐operatively for a type A thoracolumbar (T11‐L4) spinal fracture at the University Hospital Groningen, the Netherlands. Thirty‐three of the 81 selected patients agreed to participate in the study (response‐rate 41%). Respondents were older than non‐respondents (mean 50.5 years versus 39.2 years), but did not differ from each other concerning injury‐related variables. Patients with neurological deficits were excluded. Treatment consisted either of mobilization without brace, or of bedrest followed by wearing a brace. Restrictions in body function and structure were measured by physical tests (dynamic lifting test and bicycle ergometry test); restrictions in activities were measured by means of questionnaires, the Roland‐Morris Disability Questionnaire (RMDQ) and the Visual Analogue Scale Spine Score (VAS). Restrictions in participation/quality of life were assessed with the Short Form 36 (SF‐36) and by means of return to work status. Results: Thirty‐seven per cent of the patients were not able to perform the dynamic lifting test within normal range. In the ergometry test, 40.9% of the patients performed below the lowest normal value, 36.4% of the patients achieved a high VO2‐max. Mean RMDQ‐score was 5.2, the mean VAS‐score was 79. No significant differences between patients and healthy subjects were found in SF‐36 scores, neither were differences found between braced and unbraced patients in any of the outcome measures. Concerning the return to work status: 10% of the subjects had stopped working and received social security benefits, 24% had arranged changes in their work and 14% had changed their job. Conclusion: We conclude that patients do reasonably well 5 years after non‐operative treatment of a thoracolumbar fracture, although outcome is diverse in the different categories and physical functioning seems restricted in a considerable number of patients.

PHYSICAL CAPACITY

61

Introduction

With respect to the patients’ status after treatment of a spinal fracture, literature mostly focuses on radiological aspects. However, the result of a spinal fracture and its treatment can be seen much more widely than radiological results alone, for example in terms of remaining back pain or exercise tolerance, referred to as functional outcome. Although most patients are more concerned about disability than about radiological results, literature concerning functional outcome after a spinal fracture is scarce. Recent work shows that patients treated operatively for a spinal fracture have an almost equal functional outcome as healthy people [10]. However, in our clinic spinal fracture patients are more often treated non‐operatively than operatively [14]. This study describes the functional outcome (measured by questionnaires and physical tests) of patients treated non‐operatively for a thoracolumbar spinal fracture.

Materials and Methods

Patients Patients treated non‐operatively for a type A (Comprehensive Classification [11]) thoracolumbar spinal fracture (T10‐L4) between 1993 and 1998 in the University Hospital Groningen, aged between 18 and 60 (at the time of injury) and without neurological deficits, were included. Exclusion criteria were spinal disorders in their medical history, pathological fractures and insufficient command of the Dutch language. Within these criteria, a group of 81 patients was identified, to whom a letter was sent asking to take part in the study. Thirty‐two persons did not respond despite several attempts to contact them, 8 patients did not want to join and 8 patients did not show up at several appointments. Eventually, 33 patients participated in the study (response rate=41%). Details of the study group (n=33) are: mean age at the time of examination 50.5 years (S.D. 11.6, range 27–67); mean follow‐up time 5.3 years (S.D. 1.7, range 3‐8); 20 patients were male, 13 patients were female. Co‐morbidity was: one patient suffered from diabetes mellitus, two patients suffered from cardiovascular disease and two patients suffered from chronic obstructive pulmonary disease (COPD). Etiological factors were traffic accidents (n=12), sports (n=4) and falls (n=17). Fracture levels are shown in Table 1; most fractures (64%) occurred at the thoracolumbar junction (T12/L1), the greater part (82%) was classified as type A1.1 and A1.2 (Table 2) [11].

CHAPTER 4

62

Table 1 Fracture level

level n T11 3 T12 11 L1 10 L2 7 L3 1 L4 1

Table 2 Comprehensive Classification in 33 patients

class subclass n A1.1 12 A1.2 15 A1

A1.3 0 A2.1 2 A2.2 1 A2 A2.3 0 A3.1 2 A3.2 0 A3 A3.3 1

No difference was found in gender, follow‐up, co‐morbidity or fracture severity between respondents and non‐respondents; respondents, however, were older than non‐respondents (50.5 years versus 39.2 years) (p<0.001). The study protocol was approved by the Medical Ethics Committee of the University Hospital Groningen (Nr. 99/12/206). Treatment Treatment was initialized in our hospital, and continued in the outpatient clinic or in a rehabilitation centre. Treatment varied and consisted of mobilization without brace (n=15, “unbraced group”), or two to six weeks of bedrest (or strykerframe) followed by a three‐point reclination brace (n=18, “braced group”). Comparing both groups, patients did not differ in number, age, gender or follow‐up. The decision for brace application was made by a senior staff member: A2 and A3 type fractures were braced, more severe type A1.2 and A1.3 fractures (e.g. those with a large anterior wedge angle) were also braced. By protocol, patients were seen in the outpatient clinic by the surgeon and the rehabilitation specialist after 6 weeks and 3, 6, 9, 12 and 24 months. Patients were mobilized with the guidance of

PHYSICAL CAPACITY

63

a physiotherapist or an occupational therapist, mobilization was also conducted by protocol. After three months weight bearing exercises were introduced. The brace was worn for 9 months, the first 6 months night and day, the last 3 months only in the daytime. Patients were allowed to drive a car or ride a bicycle after 3 and 9 months, respectively. Functional outcome In this study functional outcome was defined according to the International Classification of Functioning, Disability and Health (ICF) by three distinct entities: restrictions in body function and structure, restrictions in activities, and restrictions in participation/quality of life [21]. Restrictions in body function and structure: Dynamic lifting tests as well as an ergometry exercise test were carried out to measure restrictions in body function and structure.

‐ Dynamic lifting test: Patients are asked to lift a box containing a weight from the floor to a 75cm‐high table four times in 20 seconds. The starting weight for men is 5.85 kg, for women 3.6 kg. After this exercise the patient rests for 20 seconds. After each break, the patient decides whether to go on with a heavier weight (men 4.5 kg more, women 2.5 kg more), or to stop. The test is stopped when the cardiac frequency rises above the personal maximum value (maximum cardiac frequency ={220‐age}×0.85), when the personal maximum lifting weight is achieved (maximum weight =0.6×bodymass), when the patient cannot complete the exercise within 20 seconds, or when the patient wants to stop for any other reason [12]. The highest lifted weight is called the maximum lifted load. This load is compared to the National Institute for Occupational Safety and Health (NIOSH) norm, which is the maximum occupational load people are allowed to lift (14.8 kg) [23]. The loading‐degree is then calculated according to the formula: loading‐degree = maximum lifted load/14.8 kg. Twenty‐seven patients (82%) carried out the test. Three patients did not participate in the test because their cardiac frequency in rest exceeded 90 beats per minute (bpm) or their diastolic blood pressure in rest exceeded 100 mm Hg. Two patients did not participate because of cardiovascular medication usage, one patient did not participate for other reasons.

CHAPTER 4

64

‐ Ergometry test: The relative VO2‐max (maximum oxygen uptake in milliliters/minute.kg) was calculated after a sub‐maximal bicycle ergometry test (excalibur 600 sport, LODE). The starting load at 60 revolutions per minute is 50% of the lean body mass (LBM) during 2 minutes. The load is raised to 150%, 200% and 250% of the LBM with a two‐minute interval until the cardiac rate is 120 bpm or more. When the cardiac rate is 120 bpm, the load is not raised further, and at this load 6 minutes of exercise follow. The VO2‐max is then calculated according to the following formula [1]:

for men: VO2‐max= (174.2 × load + 4020) / (103.2 × cardiac rate ‐ 6299),

for women: VO2‐max= (163.8 × load + 3780) / (104.4 × cardiac rate ‐ 7514).

Twenty‐two patients (67%) did the ergometry test. Five patients did not participate because of the reasons mentioned above (cardiovascular), 6 patients didn’t participate for other reasons. A more detailed description of the tests used has been published before [10]. Results of both tests were compared to normal values [1, 12]. Restrictions in activities: Restrictions in activities were measured by two disease‐specific questionnaires; the Roland‐Morris Disability Questionnaire (RMDQ) and the Visual Analogue Scale Spine Score (VAS). The Dutch version of both questionnaires was used. The RMDQ has been used extensively before to measure restrictions in activities due to back pain. The form consists of 24 statements concerning back‐related activities, which can be ticked as positive (restricted) or negative (not restricted). Scores can vary from 0 to 24, a lower score indicating less impairment [16‐18]. The VAS, developed to be used with spinal fracture patients, asks the patient to rate the functional outcome in 19 items on a 10cm visual scale. The patient’s perception of pain and restriction in activities related to back‐problems is measured. Higher scores represent better results, converted to percentages of the maximum score (0‐100). In previous studies, it has proved to be a reliable and valid instrument [8, 10, 14]. Restrictions in participation/quality of life: The Dutch version of the RAND 36‐item health survey Short Form 36 (SF‐36) and the return to work status were used to assess restrictions in participation/quality of life.

PHYSICAL CAPACITY

65

The Short Form 36 scale contains nine sub‐scales measuring: physical functioning, social functioning, role restriction due to physical problems, role restriction due to emotional problems, mental health, energy and vitality, pain, general perception of health and change in health over the past year. Scores can vary from 0 to 100, higher scores indicate better results [6, 13, 22]. Resulting scores were compared to normal data (healthy subjects, age 18‐64 years) [7]. To assess return to work status, patients were asked about employment in the past and at present. Statistical analysis Statistical analysis was carried out with SPSS 11.0 (SPSS inc. Chicago, Illinois). For the total study group, RMDQ, VAS and SF‐36 scores were compared to literature using the Student t‐test. To compare the braced and the unbraced group, results were tested non‐parametrically by means of the Wilcoxon test. Correlation was tested using Pearson’s correlation coefficient r. A p‐value of 0.05 was considered significant.

Results

Restrictions in body function and structure Results of the dynamic lifting test and bicycle ergometry test, compared to normal values, are shown in Table 3 (for the total study group, the braced and the unbraced group). In the total study group, 37% of the patients were not able to perform the dynamic lifting test within normal range. No differences were found between the braced and the unbraced group (p=0.792). In the ergometry test, 40.9% of the patients in the total study group performed below the lowest normal value, 36.4% of the patients achieved a high VO2‐max. There was no significant difference between the braced and the unbraced group (p=0.300). For both tests, scores are corrected for age and gender. Restrictions in activities For the total study group, a mean RMDQ‐score of 5.2 was found. The mean VAS‐score was 79. No differences in mean RMDQ‐score or mean VAS‐score between the braced and the unbraced group were found (p=0.442 and p=0.190, respectively) (Table 4).

CHAPTER 4

66

Table 3 Restrictions in body function and structure as measured by the dynamic lifting test (L.D. = loading degree) and ergometry test (VO2‐max in ml/min.kg), compared to normal data for the total study group, the braced and the unbraced group

n mean SD range under N‐value

VO2‐max low medium high

total 27 1.9 0.8 0.3‐2.7 37.0% ‐ ‐ ‐ braced 15 1.9 0.9 0.3‐2.7 40.0% ‐ ‐ ‐ L.

D.

unbraced 12 2.0 0.7 0.9‐2.7 33.3% ‐ ‐ ‐

total 22 34 12 16‐65 40.9% 13.6% 9.1% 36.4% braced 11 36 14 16‐65 27.3% 18.2% 9.1% 45.5%

V02‐m

ax

unbraced 11 32 11 20‐59 54.5% 9.1% 9.1% 27.3%

Table 4 Restrictions in activities as measured by the RMDQ and VAS for the total study group, the braced and the unbraced group

mean SD range total 5.2 5.9 0‐17 braced 4.4 5.5 0‐17

RMDQ

unbraced 6.1 6.4 0‐17 total 79 19 36‐100 braced 82 19 39‐100 V

AS

unbraced 75 19 36‐97 Restrictions in participation/quality of life Table 5 shows results of the SF‐36 for the total study group, the braced and the unbraced group. Scores were compared to normal data; no significant differences in any of the sub‐scales were found for neither group, or between groups. Correlation between RMDQ‐scores, the ergometry test, the dynamic lifting test, VAS‐scores, SF‐36 physical functioning and SF‐36 general health are shown in Table 6. Before injury, 21 patients had paid work. At follow‐up, 22 patients had paid work (three patients were in search of a job before injury, and were in paid work at follow‐up). Two patients (10%) had stopped working and received social security benefits, 5 patients (24%) had arranged changes in the kind of work or in the intensity or duration of their work. Three patients (14%) had changed their job due to back‐complaints.

PHYSICAL CAPACITY

67

Table 5 Restrictions in participation/quality of life as measured by the SF‐36 (mean; (S.D.) range) for the total study group, the braced and the unbraced group

SF‐36 sub‐scale total braced unbraced Physical functioning 80; (20) 25‐100 84; (18) 50‐100 76; (22) 25‐100 Social functioning 85; (19) 38‐100 83; (20) 38‐100 86; (18) 63‐100 Phys. role restriction 72; (39) 0‐100 68; (39) 0‐100 77; (41) 0‐100 Emotion. role restr. 81; (32) 0‐100 72; (40) 0‐100 91; (15) 67‐100 Mental health 79; (17) 24‐100 75; (20) 24‐100 83; (11) 64‐100 Energy / Vitality 69; (20) 20‐100 68; (21) 35‐100 71; (20) 20‐100 Pain 78; (25) 0‐100 82; (21) 22‐100 73; (28) 0‐100 General health 74; (15) 30‐95 79; (9) 65‐95 68; (19) 30‐90 Change in health 54; (19) 25‐100 58; (19) 25‐100 48; (16) 25‐100 Table 6 Correlation coefficient r between RMDQ, ergometry test, dynamic lifting test (dyn. lift test), VAS, SF‐36 physical functioning (SF‐36 phys.) and SF‐36 general health (SF‐36 gen.)

RMDQ ergometry test

dyn. lift test

VAS SF‐36 phys.

SF‐36 gen.

RMDQ 1.00 ‐0.37 ‐0.62 † ‐0.85 † ‐0.87 † ‐0.63 † ergometry test ‐0.37 1.00 0.38 0.26 0.41 0.33 dyn. lift test ‐0.62 † 0.38 1.00 0.71 † 0.59 † 0.37 VAS ‐0.85 † 0.26 0.71 † 1.00 0.71 † 0.52 † SF‐36 phys. ‐0.87 † 0.41 0.59 † 0.71 † 1.00 0.65 † SF‐36 gen. ‐0.63 † 0.33 0.37 0.52 † 0.65 † 1.00

† significant at p< 0.05

Discussion

This study was developed to gain insight into the functional outcome in patients treated non‐operatively for a thoracolumbar spinal fracture. In order to construct “outcome” in a broad manner, we used the concepts as described by the ICF of the World Health Organization [2, 21]. To obtain subjective and objective data, questionnaires as well as physical tests were used; use of the latter is relatively unique in this field of research.

Restrictions in body function and structure Results of the dynamic lifting test show that 37% of the patients were not able to perform this test within normal values, indicating that these patients have a lower physical capacity than healthy people. Almost equal results were found in the bicycle ergometry test, in which 41% of the patients achieved scores under the lowest normal value. Surprisingly, nearly the same proportion of patients achieved a high VO2‐max (within a normal distribution). Although examination took place

CHAPTER 4

68

approximately 5 years after injury, and no further neurological deficit occurred, this still means that a large part of the study population is impaired in the light of restriction in body function and structure. No difference was found between the braced patients and the unbraced patients. To our best knowledge, no other publication is available concerning VO2‐max in non‐operatively treated spinal fracture patients, which makes comparison to other series a delicate issue. Pulmonary function was studied by Schlaich et al. in patients with an osteoporotic spinal wedge fracture [19]. They found that the vital capacity (VC) and forced expiratory volume in 1 second (FEV1), corrected for age and gender, were lower than in healthy subjects. According to the authors, this might be a result of spinal deformity (hyperkyphosis) which leads to disturbed mechanical function. Why so many patients in our series perform under normal values is unknown. It might be that pain leads to fewer leisure‐activities, resulting in decreased functional capacity. However, this cannot be the only explanation, since remaining pain is not severe, considering the VAS, RMDQ and SF‐36 scores found. Another possible explanation might be found in cognitive factors; as mentioned by Cox et al., fear of refracture may lead to a less functional use of the back, which may result in a lower level of activity [5]. Restrictions in activities Concerning restrictions in activities, a mean RMDQ‐score of 5.2 was found and a mean VAS‐score of 79. These findings indicate that patients are impaired and restricted in activities, but not in a severe manner. It should be kept in mind though, that only patients without neurological deficits were included. As in the physical tests, no difference was found between braced and unbraced patients, so it seems that brace‐usage does not influence impairment in the long term. Weinstein et al. reported a RMDQ‐score of 13.2, measured 20 years after non‐operative treatment for a thoracolumbar burst fracture [24]. Comparison makes our results seem favourable. However, 22% of the patients had some neurological deficit in the afore‐mentioned study. In a recent study, RMDQ‐ and VAS‐scores in non‐operatively treated patients were found to be 4.4 and 72.6 respectively [14]. These findings are comparable to our results. A RMDQ‐score of 3.9 was reported recently in patients 3.7 years after non‐operative treatment of a spinal fracture [25]. Knop et al. found a VAS‐score of 66 for patients treated operatively for a spinal fracture at a follow‐up of 23 months [8]. Our results seem better, though our longer follow‐up time and the different treatment strategies do not make a comparison completely valid.

PHYSICAL CAPACITY

69

Restrictions in participation/quality of life No significant differences between our population and healthy subjects were found concerning SF‐36 scores, neither were significant differences found between braced and unbraced patients. Our results are more favourable than those reported by Kraemer et al. in 1996 [9]. Comparing our study to Kraemer’s paper, we cannot find an explanation for the higher scores found in our series. Correlation coefficients of the different outcome measures were in some cases significant and fairly strong. Surprisingly, the ergometry test did not correlate with any of the other measures. The correlation coefficient of the RMDQ and dynamic lifting test was negative, indicating the lower the RMDQ (less impairment), the more weight was lifted. The same relationship was found between the VAS and SF‐36 physical functioning on the one hand, and dynamic lifting test on the other: the higher VAS and SF‐36 scores (fewer restrictions in activity), the more weight was lifted. Only 10% of patients had stopped working due to back‐problems associated with their spinal fracture. In a social security system like in the Netherlands, where patients receive substantial benefits in case of illness or disablement, a drop‐out of 10% seems a good result. Thirty‐nine per cent of the patients had changed their job or changed the intensity or duration of their work. These data might be influenced by the fact that respondents were quite old (mean age 50 years). In a study by Shen et al., concerning patients treated non‐operatively for a thoracolumbar burst fracture, 76% of the patients returned to their original employment and 8% stopped working [20]. Those results are comparable to ours. In a study by Reid et al. (describing patients treated non‐operatively for a thoracolumbar burst fracture without neurological deficits), 19% was unable to return to work [15]. Two other studies (both concerning non‐operatively treated thoracolumbar burst fractures without neurological deficits) show comparable return to work status: 95% and 87% respectively [3, 4].

There are some limitations in this study. The low response rate may have biased our results despite the fact that no differences were found in gender, follow‐up, co‐morbidity or fracture severity between respondents and non‐respondents. Respondents were 11 years older than non‐respondents. The difference in age does not seem to affect the physical capacity tests since results and normal values were corrected for age. In contrast, it might be that the return to work status would have been even better if younger patients had taken part in the study. Another limitation of the study is the fact that we cannot prove that braced or unbraced

CHAPTER 4

70

patients have comparable outcomes. Our results show a trend that there are no differences in functional outcome between braced and unbraced patients. However, to answer this question properly, this issue should preferably be investigated in a randomized clinical trial.

Conclusions

Functional outcome in patients 5 years after non‐operative treatment for a type A thoracolumbar fracture seems reasonably good, though diverse in the light of the ICF. In physical capacity tests a large part of patients seems restricted. On the other hand, patients are only mildly restricted in activities. No restriction is present concerning participation or the quality of life. Why patients perform less well than healthy people in physical tests remains unknown and should be studied in further research.

References

1. Astrand PO, Rohdahl K (1986) Textbook of Workphysiology: physiological bases of exercise. McGraw‐Hill Book Company, New York, pp 360‐369

2. Bickenbach JE, Chatterji S, Badley EM, Ustun TB (1999) Models of disablement, universalism and the international classification of impairments, disabilities and handicaps. Soc Sci Med 48:1173‐1187




6. Grevitt M, Khazim R, Webb J, Mulholland R, Shepperd J (1997) The short form‐36 health survey questionnaire in spine surgery. J Bone Joint Surg Br 79:48‐52

7. Jenkinson C, Coulter A, Wright L (1993) Short form 36 (SF36) health survey questionnaire: normative data for adults of working age. BMJ 306:1437‐1440




PHYSICAL CAPACITY

71


12. Mayer TG, Barnes D, Kishino ND, Nichols G, Gatchel RJ, Mayer H, Mooney V (1988) Progressive isoinertial lifting evaluation. I. A standardized protocol and normative database. Spine 13:993‐997

13. McHorney CA, Ware JE, Raczek AE (1993) The MOS 36‐Item Short‐Form Health Survey (SF‐36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Med Care 31:247‐263





18. Roland M, Morris R (1983) A study of the natural history of low‐back pain. Part II: development of guidelines for trials of treatment in primary care. Spine 8:145‐150

19. Schlaich C, Minne HW, Bruckner T, Wagner G, Gebest HJ, Grunze M, Ziegler R, Leidig‐Bruckner G (1998) Reduced pulmonary function in patients with spinal osteoporotic fractures. Osteoporos Int 8:261‐267


21. Ustun TB, Chatterji S, Bickenbach J, Kostanjsek N, Schneider M (2003) The International Classification of Functioning, Disability and Health: a new tool for understanding disability and health. Disabil Rehabil 25:565‐571

22. Ware JE, Sherbourne CD (1992) The MOS 36‐item short‐form health survey (SF‐36). I. Conceptual framework and item selection. Med Care 30:473‐483

23. Waters TR, Putz‐Anderson V, Garg A, Fine LJ (1993) Revised NIOSH equation for the design and evaluation of manual lifting tasks. Ergonomics 36:749‐776

24. Weinstein JN, Collalto P, Lehmann TR (1988) Thoracolumbar “burst” fractures treated conservatively: a long‐term follow‐up. Spine 13:33‐38

25. Wood K, Butterman G, Mehbod A, Garvey T, Jhanjee R, Sechriest V (2003) Operative compared with nonoperative treatment of a thoracolumbar burst fracture without neurological deficit. A prospective, randomized study. J Bone Joint Surg Am 85‐A:773‐781



73

Chapter 5

Non‐operatively treated type A spinal fractures: mid‐term versus long‐term functional outcome RB Post, CK van der Sluis, VJM Leferink, PU Dijkstra, HJ ten Duis Accepted International Orthopaedics (2008). In press.

CHAPTER 5

74

Abstract

Introduction: The type A spinal fracture (Comprehensive Classification) without neurological deficit is the most common type of spinal fracture. Generally, these fractures are treated non‐operatively. Little is known about the mid‐term and long‐term functional outcomes of these fractures. According to the literature, severe pain and late onset neurological injury may occur a long time after a spinal fracture. Objective: This study focuses on the mid‐term (4 years) and long‐term (10 years) functional outcomes of patients treated non‐operatively for a type A spinal fracture without primary neurological deficit. Methods: Functional outcome was measured using the Visual Analogue Scale Spine Score (VAS) and the Roland‐Morris Disability Questionnaire (RMDQ). Results: The 50 patients included were on average 41.2 years old at the time of injury. Four years post‐injury a mean VAS score of 74.5 and a mean RMDQ score of 4.9 were found. Ten years after the accident, the mean VAS and RMDQ scores were 72.6 and 4.7, respectively (ns). No significant relationships were found between the difference scores of the VAS and RMDQ on the one hand, and age, gender, fracture sub‐classification, and time between measurements on the other hand. Three (6%) patients had a poor long‐term outcome. None of the patients required surgery for late onset pain or progressive neurological deficit. Conclusions: Functional outcome after a non‐operatively treated type A spinal fracture is good, both 4 years and 10 years post‐injury. For the group as a whole, 4 years after the fracture a steady state exists in functional outcome, which does not change systematically for at least 10 years after the fracture.

LONG‐TERM OUTCOME

75

Introduction

The type A spinal fracture, according to the Comprehensive Classification (CC), is the most common type of spinal fracture, usually presenting without neurological deficit [13]. This type of fracture is characterized by compression of the vertebral body without injury to the posterior ligamentous complex, and in the absence of sagittal translation [13]. Type A fractures are often treated non‐operatively. A large amount of literature is available concerning radiological results and short‐term results regarding these fractures. Besides radiological results, an issue of great importance is the functional outcome of these patients. Little is known about functional outcome after type A spinal fractures [18, 19, 21]. Most of the published data concentrate on relatively short‐term results (one year follow‐up). Literature regarding long‐term outcome (10 years and more) is scarce [8, 19, 26]. One might expect increasing pain over time due to altered facet joint motion and hyperextension of adjacent spinal regions, leading to ongoing degenerative processes [15, 25]. Furthermore, fatigue pain from the soft tissues surrounding the spinal misalignment and the injured disc may be of influence on back pain in the long term [3]. Even late onset neurological deficit may occur years after the trauma, demanding operative intervention [4, 25]. As such, more information on long‐term functional outcome after spinal fractures is needed to understand the problems patients are confronted with years after a spinal fracture. In this article the mid‐term (4 years) and long‐term (10 years) functional outcome is described of a consecutive cohort from patients treated non‐operatively for type A (A1.1‐A3.2) spinal fractures without primary neurological deficit.

Methods

Patients Patients aged between 18 and 60 without primary neurological deficit who were treated non‐operatively for a type A thoracolumbar (T6‐L5) spinal fracture (according to the Comprehensive Classification [13]) at the University Medical Centre Groningen, the Netherlands, were eligible for the study. All patients were treated between 1993 and 2000. Exclusion criteria were previous spinal disorders in the medical history, psychiatric illnesses, pathological fractures, and insufficient command of the Dutch language. Only patients who had taken part in our previous studies and whose mid‐term outcome was known were included [17, 18].

CHAPTER 5

76

Patients were sent a letter requesting their participation along with two questionnaires for completion. Medical files of all included patients were reviewed to obtain data on late onset pain and late onset neurological deficit. Treatment Treatment was initialized in our hospital and continued in the outpatient clinic or in a rehabilitation centre. A senior staff member was responsible for deciding on the preferred method of therapy. Treatment consisted of two to six weeks bed rest (or strykerframe). After this period, type A1.3, A2, and A3 fractures were braced and type A1.1 and A1.2 fractures were treated without brace. However, depending on the severity of pain, some type A1.1 and A1.2 fractures were treated without bedrest, by direct mobilization without brace. Patients were mobilized with the guidance of a physiotherapist or occupational therapist. Three months post‐injury weight bearing exercises were introduced. The brace was worn for nine months, the first six months night and day, the last 3 months only during the daytime. Functional outcome Functional outcome was measured using two disease‐specific questionnaires: the Visual Analogue Scale Spine Score (VAS) and the Roland‐Morris Disability Questionnaire (RMDQ) [10, 20]. The VAS, developed for use in patients with a spinal fracture, consists of 19 items. The patients rate their functional outcome on 10 cm visual scales. As such, the patient’s perception of restriction in activities due to back‐problems is measured. Higher scores represent better results, which are converted to percentages of the maximum score (0‐100). In previous studies, the VAS Spine Score has proven to be a reliable and valid instrument [10, 17, 18, 19]. The RMDQ measures restrictions in activities due to back pain. Twenty‐four statements concerning back‐related activities are marked as positive (restricted) or negative (not restricted). Scores can vary from zero to 24, a lower score indicating less impairment [20]. The RMDQ was found to be a sensitive, reliable, and valid instrument for measuring physical impairment due to back pain [22]. Statistical analysis Statistical analysis was carried out using SPSS 11.0 (SPSS Inc., Chicago, Illinois). VAS and RMDQ scores 4 years and 10 years after the trauma were compared by means of the paired‐sample t‐test. To analyze the effect of independent variables

LONG‐TERM OUTCOME

77

(i.e. age, gender, fracture type, and duration of time between measurements) on VAS and RMDQ difference scores (i.e. mid term ‐ long term), a linear regression analysis was performed. A p‐value of 0.05 was considered to be of statistical significance.

Results

Patients Sixty‐two patients had taken part in preceding studies and their mid‐term functional outcome was known. From these 62 patients, seven were lost to follow‐up and five patients refused to take part for a variety of reasons. Fifty patients (50/62=81%) returned the questionnaires and comprised the study group. The study group (n=50) consisted of 31 (62%) men and 19 (38%) women. Mean age at the time of injury was 41.2 years (S.D. 12.0, range 19‐60 years). Mean follow‐up time for the mid‐term functional outcome was 4.3 years (S.D. 1.7, range 2‐7 years) and for the long‐term 9.8 years (S.D. 2.0, range 7‐14 years). Fracture levels ranged from T6 to L5, 67% occurred at the thoracolumbar junction (T12‐L1). Eight patients (16%) had two or more spinal fractures. Of these, only the most severe fracture type was registered and the other fracture was not taken into account. Fracture types according to the Comprehensive Classification are shown in Table 1 [13].

Table 1 Fracture types according to the Comprehensive Classification (n=50) [13]

Class n Subclass n A1.1 9 A1.2 17 A1 28 A1.3 2 A2.1 3 A2.2 1 A2 4 A2.3 ‐ A3.1 17 A3.2 1 A3 18 A3.3 ‐

n=absolute number of patients Etiological factors were accidental falls (n=23), traffic accidents (n=17), sports accidents (n=6), and occupational injuries (n=4). None of the patients required surgery for late onset pain or late occurring neurological deficit. No differences were found between respondents and non‐respondents concerning age, gender, fracture type, or follow‐up time.

CHAPTER 5

78

Functional outcome No significant differences were found between VAS and RMDQ scores at mid‐ and long‐term follow‐up (p=0.291 and p=0.733, respectively). The mean difference scores of the VAS and RMDQ were 1.9 and 0.2, respectively (see Table 2).

Table 2 Outcome‐ and difference scores

mid‐term outcome long‐term outcome difference scores VAS RMDQ VAS RMDQ VAS RMDQ mean 74.5 4.9 72.6 4.7 1.9 0.2 median 82.0 4.5 74.0 2.5 1.6 0.0 S.D. 21.2 4.6 22.0 5.4 13.1 4.1 range 23 ‐ 100 0 ‐ 17 14 ‐ 98 0 ‐ 21 –23 ‐ 25 –13 ‐ 8

Three patients (6%) had a significantly higher RMDQ score at the long‐term assessment. Two of them (patients one and two) displayed a RMDQ difference score of ‐13 (along with a VAS difference score of +9 and ‐4) (see Fig. 1). In boxplots, cases which are outside the box by more than 1.5 times the interquartile range (interquartile range = 3rd quartile ‐ 1st quartile) are outliers [24]. As such, these two patients are considered outliers.

VAS difference score RMDQ difference score

30

20

10

0

‐10

‐20

‐30

Fig. 1 Box‐plot graph showing the VAS‐ and RMDQ difference scores. The graph illustrates the median (inner black line), the upper and lower quartiles (the box), the range of data falling within 1.5 x interquartile range (the whiskers) and outliers (□ patients one and two, ● patient three)

LONG‐TERM OUTCOME

79

The third patient (patient three) showed a RMDQ difference score of ‐8 (VAS difference score +17) and could also be considered as an outlier (see Fig. 1). Characteristics of these three patients will be described in the Discussion section. None of these three patients differed significantly from the study group with regard to age, gender, follow‐up time, or type of fracture. Regression analysis with VAS and RMDQ difference scores as dependent variables showed no correlation with age at the time of injury, gender, fracture classification, and time between the measurements.

Discussion

Approximately 66% of spinal fractures can be classified as being type A fractures. Of those, 86% present without neurological deficit [13]. Often, these fractures are treated non‐operatively. Consequently, non‐operatively treated type A spinal fractures form the large majority of spinal fractures. This retrospective, cross‐sectional study was conducted to assess long‐term functional outcome after non‐operatively treated type A spinal fractures.

Our data show good results at both 4 years as well as 10 years after the fracture. Functional outcome as measured by the VAS and RMDQ appears good at both measurements. A VAS score of 73‐75 can be interpreted as an admirable figure. Furthermore, an average RMDQ score of 5 reflects almost no disability. Compared to a healthy reference group (VAS score 92‐93, RMDQ score of 0.5), patients are only slightly impaired [10, 17]. When considering these numbers, however, it should be kept in mind that patients with primary neurological deficits were not included in this study. It is well known that neurological injury negatively affects functional outcome in patients with a spinal fracture [14]. Our results concerning the RMDQ are similar to that of Siebenga et al. who investigated functional outcome 7 years after non‐operatively treated type A spinal fractures; an average RMDQ score of 4.6 was found [21]. A study concerning functional outcome 16 years after non‐operatively treated type A spinal fractures reported a mean VAS score of 58 points [19]. Our results concerning VAS scores are more favourable. An explanation might be that in the afore‐mentioned paper the study group comprised some neurologically injured patients. Recently a VAS score of 67 was found 5 years after non‐operative treatment for type A fractures [5]. Our mid‐term VAS score is to some extent comparable. Similar to our results, Tezer et al. as well as did Butler et al. [6, 23] also found satisfactory results 6 years after non‐operative treatment of

CHAPTER 5

80

compression and “burst” fractures. Weinstein et al., as one of the first to study functional outcome after spinal fractures, found a mean RMDQ score of 13, measured 20 years after non‐operatively treated “burst” fractures [26]. Our results seem better; a clarification might be found in the fact that Weinstein’s paper included patients with neurological deficits and that the fracture classification and distribution was dissimilar to our data. No difference was found between functional outcomes at the mid‐ and long‐term assessments. Furthermore, the time between the two measurements did not show a correlation with VAS and RMDQ difference scores. This indicates that a “steady state” on a group level exists from (at least) 4 to 10 years post‐injury. Previous studies mention a status quo in functional outcome ranging from 2 to 4 years after a spinal fracture [1, 7]. Data on the course of outcome after 2 to 4 years, however, are not available. To our knowledge, there is no previously published paper available studying the course of functional outcome after a spinal fracture in the same cohort. As mentioned before, pain and neurological deficit can arise long after a spinal fracture (mostly after non‐operative treatment, but rarely after operative treatment) [25]. In these cases, operative treatment might be necessary [4, 25]. Late onset pain requiring operative interference has also been reported for type A fractures [3]. In our series however, none of the patients required surgery for late neurological symptoms or pain. Three patients had a higher RMDQ score at the long‐term assessment (difference scores ‐13 and ‐8), indicating more impairment. Since the minimal clinically important change for the RMDQ is 3.5 points, the alteration in RMDQ scores in these patients indicates a clinically important change [16]. When looking closer at these patients, none of them had a significantly different gender, age, follow‐up time or fracture type compared to the rest of the study group. Patient one, a 44‐year‐old woman, sustained a type A1.2 fracture of T12 due to a car‐accident in 1999. She was treated by direct mobilization without brace. At the mid‐term measurement (3 years post‐injury) she had a RMDQ score of 7 and a VAS score of 23. At the long‐term follow‐up (8.1 years post‐injury), a RMDQ score of 20 was found and a VAS score of 14. After the regular clinical control visits, she never contacted us concerning severe pain. Her general practitioner had requested an X‐ray in 2002 since the patient suffered from back pain (see Fig. 2). This X‐ray showed osteoarthritis in the facet joints, which could explain her complaints. On the other hand, her VAS score had deteriorated 9 points, which is less than 20 points, the minimal clinically important change when using a VAS [16].

LONG‐TERM OUTCOME

81

Fig. 2 Radiograph of a T12 fracture (type A1.2) in a 44‐year‐old woman, three years post‐injury, showing facet joint arthritis (indicated by arrows) The second patient was a 44‐year‐old man, who had sustained a T12 fracture in 1997, type A1.2, due to a fall. He was treated by 2 weeks of bed rest followed by a brace. At the mid‐term measurement (5.4 years post‐injury) he had a RMDQ score of 8 and a VAS score of 23. At the long‐term follow‐up (10.3 years post‐injury), a RMDQ score of 21 and a VAS score of 27 were found. After the regular control visits, the patient was seen by a neurologist in 2004 because of low back‐pain and a strange feeling in both legs. No pathological neurological conditions were found. An X‐ray showed spondylosis at L2‐L3 (see Fig. 3). This could explain the patient’s discomfort. The third patient was a 38‐year‐old male, who sustained a type A1.1 fracture of L3 in 2000 following a sports accident. In addition, he sustained a femoral fracture and an acetabulum fracture. The femoral fracture was treated by intramedullary nailing, the acetabulum fracture was treated non‐operatively. At the mid‐term measurement (2.1 years post‐injury) he had a RMDQ score of 0 and a VAS score of 89. At the long‐term follow‐up (7.3 years post‐injury), a RMDQ score of 8 and a VAS score of 72 were found. Why this patient deteriorated is unknown.

CHAPTER 5

82

Fig. 3 Radiograph of the type A1.2 T12 fracture in a 44‐year‐old male, seven years post‐injury, showing spondylosis at L2‐L3 (indicated by arrows)

Given the favourable outcomes found in our series, non‐operative treatment is an accurate approach for type A spinal fractures without primary neurological deficit. Three patients however, had poor outcomes, although their fractures were classified as being the rather “simple” type A1.1 and A1.2. Why these patients had poor outcomes is unknown. As we made no MRI scans at that time, potential posterior ligamentous complex (PLC) injuries can not be excluded. The PLC is important in maintaining stability in the spinal column, and rupture may result in instability and severe back‐pain [15, 23, 25]. As reported by Leferink et al., 30% of type A fractures (classified on CT images) appeared to be type B fractures (PLC lesions present) during operation [12]. We could not demonstrate a relationship between VAS‐ and RMDQ difference scores compared to age, gender, fracture type, or time between measurements. Since no correlation was found between outcome and sub‐classification, it is open to discussion whether such an extensive classification as the CC is required in daily practice. Perhaps there is need for a new, less extensive classification system, which gives more direction to treatment and uses MRI for detecting PLC injuries. The thoracolumbar injury classification and severity score (TLICS), as recently developed by the Spine Trauma Study Group, recognizes the afore‐mentioned criteria; possibly it will replace the CC in the future [11].

LONG‐TERM OUTCOME

83

This study is subject to certain restrictions and limitations that are worth mentioning. When assessing outcome 10 years after a fracture, scores on the questionnaires might be influenced by back pain unrelated to the spinal fracture. For example, pain may arise solely due to the normal process of ageing and osteoarthritis, not per se at the level of the injured vertebra. Furthermore, it is known that a variety of other factors, which we did not consider (e.g. chronic illness, lower education level) might influence back pain [9]. Finally, the small sample size might have introduced a type 2 statistical error [2].

Conclusions

Functional outcome after a non‐operatively treated type A spinal fracture is good, both at 4 years as well as 10 years post‐injury. Patients are only slightly disabled. For the group as a whole, 4 years after the fracture a steady state exists in functional outcome, which does not change systematically for at least 10 years after the fracture. A small number of patients have a poor outcome, though none of our patients required surgery for late onset pain or late onset neurological deficit. Further research in this group of patients is advocated to reveal contributing factors.

References


2. Bailey CS, Fisher CG, Dvorak MF (2004) Type II error in the spine surgical literature. Spine 29:1146‐1149




6. Butler JS, Walsh A, O’Byrne J (2005) Functional outcome of burst fractures of the first lumbar vertebra managed surgically and conservatively. Int Orthop 29:51‐54



CHAPTER 5

84

9. Harris IA, Young JM, Rae H, Jalaludin BB, Solomon MJ (2007) Factors associated with back pain after physical injury: a survey of consecutive major trauma patients. Spine 32:1561‐1565



12. Leferink VJM, Zimmerman KW, Veldhuis EFM, ten Vergert EM, ten Duis HJ (2002) Classificational problems in ligamentary distraction type vertebral fractures: 30% of all B‐type fractures are initially unrecognised. Eur Spine J 11:246‐250



15. Öner FC, van Gils APG, Faber JAJ, Dhert WJA, Verbout AJ (2002) Some complications of common treatment schemes of thoracolumbar spine fractures can be predicted with magnetic resonance imaging: prospective study of 53 patients with 71 fractures. Spine 27:629‐636

16. Ostelo RW, de Vet HC (2005) Clinically important outcomes in low back pain. Best Pract Res Clin Rheumatol 19:593‐607



19. Reinhold M, Knop C, Lange U, Bastian L, Blauth M (2003) Non‐operative treatment of thoracolumbar spinal fractures. Long‐term clinical results over 16 years. Unfallchirurg 106:566‐576


21. Siebenga J, Segers MJ, Elzinga MJ, Bakker FC, Haarman HJ, Patka P (2006) Spine fractures caused by horse riding. Eur Spine J 15:465‐471

22. Stratford PW, Binkley JM, Riddle DL, Guyatt GH (1998) Sensitivity to change of the Roland‐Morris Back Pain Questionnaire: part 1. Phys Ther 78:1186‐1196


24. Tukey JW (1977) Exploratory Data Analysis. Addison‐Wesley, Reading, MA, pp39‐43 25. Vaccaro AR, Silber JS (2001) Post‐traumatic spinal deformity. Spine 26:S111‐S118 26. Weinstein JN, Collalto P, Lehmann TR (1988) Thoracolumbar “burst” fractures treated

conservatively: a long‐term follow‐up. Spine 13:33‐38


85

Chapter 6

Long‐term functional outcome after type A3 spinal fractures: operative versus non‐operative treatment RB Post, CK van der Sluis, VJM Leferink, HJ ten Duis Submitted

CHAPTER 6

86

Abstract

Introduction: The optimal treatment of the type A3 “burst” fracture remains a challenging issue. Literature regarding short‐term functional outcome after operative and non‐operative treatment of these fractures shows conflicting results. Regarding the long‐term outcome, hardly any data is available. Some authors do however fear complications in the long term, like late onset pain and late onset neurological deficit. Objective: To reveal the long‐term (5 years) functional outcome after operative (dorsal stabilization) and non‐operative treatment for a type A3 spinal fracture (Comprehensive Classification) without neurological deficit. Methods: Functional outcome was measured by means of two disease‐specific questionnaires: the Visual Analogue Scale Spine Score (VAS) and the Roland‐Morris Disability Questionnaire (RMDQ). Results: The 63 patients included (38 treated operatively, 25 treated non‐operatively) were on average 37 years old at the time of injury. The mean VAS scores in the operatively and non‐operatively treated groups were 82.6 and 80.8, respectively (NS). The mean RMDQ scores in the operatively and non‐operatively treated groups were 3.3 and 3.1, respectively (NS). None of the patients required surgery for late onset pain or late onset neurological deficit. Conclusion: Functional outcome appears to be good five years after operative as well as non‐operative treatment of type A3 “burst” fractures. Both treatment modalities show equal outcomes. Since outcome after both treatments is comparable, other factors than the type of fracture should be taken into account when deciding which therapy should be chosen.

TYPE A3 FRACTURE

87

Introduction

The optimal treatment for the type A3 spinal fracture (Comprehensive Classification [25]) remains a subject of debate [9, 14, 19, 38, 39, 45]. This type of fracture, also referred to as “burst” fracture, is characterized by comminution of the vertebral body with centrifugal extrusion of fragments, whereas the posterior ligamentous complex is intact [25]. Advocates of operative treatment point out the benefits of surgical approaches, namely the improvement in spinal alignment, decreased deformity, early mobilization and improvement (or no further deterioration) in neurological function [2, 12, 15, 28]. On the other hand, non‐operative treatment lacks the risks of surgery, like deep wound infection, iatrogenic neurological damage and implant failure [10, 27, 31, 37]. In addition, costs of non‐operative treatment are lower [17, 34, 40, 45]. Concerning radiological results and short‐term clinical results a large amount of literature is available [1, 18, 23, 46]. However, the results of treatment can be seen in a broader perspective than radiological results alone. In what way do patients participate in normal daily activities and do they experience back pain? The measurement of patients’ health status is referred to as functional outcome. Although studying radiological results is useful, there appears to be no relationship between functional outcome and radiological appearance (e.g. anterior wedge angle, vertebral height loss) [22, 27, 39, 44]. Some literature regarding short‐term outcome is on hand [9, 12, 14, 19, 28, 38]. However, literature regarding long‐term outcome is less available. Several authors fear complications in the long term, like progressive kyphosis resulting in back pain or even late onset neurological injury [5, 7, 42]. This study analyzes the long‐term (5 years) functional outcome after operative and non‐operative treatment for a type A3 “burst” fracture in patients without neurological deficits.

Methods

Patients Patients aged between 18 and 60 (at the time of injury) who sustained a type A3 thoracolumbar (T7‐L5) spinal fracture according to the Comprehensive Classification [25], without neurological deficit treated at the University Medical Centre Groningen were eligible for this study. To obtain the diagnosis, an X‐ray and CT‐scan were made, no standard MRI’s were made at that time. All patients were initially treated between 1996 and 2000. Exclusion criteria were previous

CHAPTER 6

88

spinal disorders in the medical history, psychiatric illnesses, pathological fractures or insufficient command of the Dutch language. Medical files of all included patients were reviewed to obtain data on late onset pain or late onset neurological deficits. Treatment A senior staff member, taking into account radiological and clinical findings, made the decision whether an operative or non‐operative procedure was preferred.

Operative treatment consisted of fracture reduction and fixation by means of dorsal instrumentation using the Universal Spine System (Synthes Cooperation, Bochum, Germany), combined with transpedicular cancellous bone grafting and dorsal spondylodesis following Dick and Daniaux [11, 13]. Fracture reduction, i.e. angular reduction and distraction, was acquired by indirect manipulation via pedicle screws as lever. Cancellous bone (taken from the dorsal iliac crest) was put transpedicularly in the reduced vertebral body and packed around the opened facet joints at the dorsolateral side afterwards as well [6, 11]. This spondylodesis was done at the level of the destructed endplate, for example only the upper segment in a type A3.1 fracture and both segments in type A3.3 fractures. No ventral operations, discectomies or laminectomies were performed. Postoperatively, all patients were transferred to a rehabilitation centre. They were allowed to walk after about 10 days in a reclination brace, which was worn for 9 months. In the final 3 months, patients only wore the brace during daytime. After 9 months the implants were removed. Non‐operative treatment was initialized in our hospital and continued in a rehabilitation centre. Treatment consisted of 6 weeks of bedrest (or strykerframe), followed by a reclination brace. Patients were mobilized with the guidance of a physiotherapist or an occupational therapist. After 3 months weight bearing exercises were introduced. The brace was worn for 9 months, the first 6 months night and day, the last 3 months only during the daytime. Patients were allowed to drive a car or ride a bicycle after 3 and 9 months, respectively. Functional outcome measurement Functional outcome was measured by two disease‐specific questionnaires: the Visual Analogue Scale Spine Score (VAS) and the Roland‐Morris Disability Questionnaire (RMDQ) [21, 35]. The VAS, developed to be used in spinal fracture patients, consists of 19 items measuring restriction in activities due to back‐related problems. Patients are asked

TYPE A3 FRACTURE

89

to value the functional outcome in these 19 items on a 10 cm visual scale. Higher scores indicate better results, converted to percentages of the maximum score (0‐100). In previous studies, it has proven to be a reliable and valid instrument [21, 24, 29, 39, 40]. The RMDQ is a health status measure designed to be completed by patients to assess physical disability due to back pain. Twenty‐four statements regarding back‐related activities can be ticked as positive (restricted) or negative (not restricted). Scores can vary from 0 to 24, a lower score indicating less impairment [35]. The Dutch version of the RMDQ was used [36]. Statistical analysis Statistical analysis was carried out using SPSS 11.0 (SPSS inc. Chicago, Illinois). Categorical data were analyzed by applying chi‐square tests. Since RMDQ and VAS scores in the operative group were skewed, the Mann‐Whitney test was used to compare means between the operative and non‐operative group. In order to analyze the influence of follow‐up time and age on the outcome, a linear regression analysis was performed with VAS and RMDQ scores as dependent variables and age and follow‐up time as independent variables. A p‐value of 0.05 was considered significant.

Results

Patients Seventy‐six patients met the inclusion criteria. From this group of 76 patients (46 treated operatively, 30 treated non‐operatively), 2 had died (8 and 9 years after treatment, due to unrelated causes) and 7 were lost to follow‐up. Sixty‐seven patients were sent two postal questionnaires and an informed consent agreement. Sixty‐three patients returned the questionnaires (follow‐up rate 63/67=94%) and comprised the study group. No differences were found in age, gender, follow‐up time or fracture classification and distribution between respondents and non‐respondents.

Details of the study group (n=63) were as follows: ‐ Patients treated operatively: Twenty‐six out of the 38 patients were males (68%) (see Table 1). Fracture levels ranged from T9 to L5, most fractures (74%) occurred at the thoracolumbar junction (T12/L1). Five patients had multiple spinal fractures, the most severe was registered, the others were not taken into account.

CHAPTER 6

90

‐ Patients treated non‐operatively: Fifteen out of the 25 patients were males (60%) (see Table 1). Fracture levels ranged from T7 to L5, most fractures (60%) occurred at the thoracolumbar junction (T12/L1). Four patients had multiple spinal fractures, the most severe was registered, the others were not taken into account.

Table 1 Details of the study group (n=63)

operative non‐operative n 38 25 gender (♂ : ♀) 26 (68%) : 12 (32%) 15 (60%) : 10 (40%) age (years) mean (S.D.) range 37.2 (11.8) 18‐56 37.4 (12.2) 19‐58 follow‐up (years) mean (S.D.) range 5.7 (2.9) 2.5‐10.6 4.8 (2.9) 2.1‐10.4

accidental falls (n) 13 10 traffic accidents (n) 12 11 sports injuries (n) 10 2

etiological factors

occupational (n) 3 2 A3.1 15 (40%) 22 (88%) A3.2 18 (47%) 3 (12%)

comprehensive classification

A3.3 5 (13%) 0 When comparing the operative and non‐operative group, no differences were found in gender, age, follow‐up time or fracture distribution. The operative group consisted of significantly more type A3.2 and A3.3 fractures and less type A3.1 fractures (p<0.01). None of the patients required surgery for late onset pain or late onset neurological deficit. Functional outcome No differences were found between operative and non‐operative patients concerning VAS and RMDQ scores (see Table 2). The distribution of VAS and RMDQ scores is shown in Figure 1 and Figure 2.

Table 2 VAS and RMDQ scores for the treatment groups

treatment VAS mean median SD range

RMDQ mean median SD range

operative 82.6 94.1 21.9 17‐100 3.3 0 5.1 0‐17 non‐operative 80.8 84.0 19.4 31‐100 3.1 1.0 3.7 0‐12

TYPE A3 FRACTURE

91

non‐operativeoperative

VAS score

110

100

90

80

70

60

50

40

30

20

10

Fig. 1 Box‐plot graph showing the VAS scores in both treatment groups. The graph illustrates the median (inner black line), the upper and lower quartiles (the box), the range of data excluding outliers (the whiskers) and outliers (○ displaying outliers)

non‐operativeoperative

RMDQ score

20

15

10

5

0

Fig. 2 Box‐plot graph showing the RMDQ scores in both treatment groups. The graph illustrates the median (inner black line), the upper and lower quartiles (the box), the range of data excluding outliers (the whiskers) and outliers (○ displaying outliers)

CHAPTER 6

92

When comparing patients with type A3.1 fractures to those with type A3.2 fractures, no significant differences were found between VAS and RMDQ scores (80 versus 86 and 3.5 versus 2.7, respectively). Within the patients with a type A3.1 fracture, no significant differences were found in VAS and RMDQ scores between those treated operatively and those treated non‐operatively (80 versus 80 and 4.0 versus 3.1, respectively). Regression analysis showed no correlation between age or follow‐up time and VAS and RMDQ scores. A strong correlation was found between VAS and RMDQ scores (Spearman’s rho ‐0.84, p<0.01).

Discussion

This study was conducted to compare the functional outcome after operative (dorsal stabilization) and non‐operative treatment in patients with type A3 spinal fractures. Literature comparing functional outcome after operative and non‐operative treatment for type A3 spinal fractures is available. However, these studies mostly focus on short‐term results [9, 12, 14, 19, 28, 38]. Papers comparing long‐term outcome (approximately 5 years or over) are reasonably scarce [3, 22, 39, 45]. Although our data were obtained in a retrospective, cross‐sectional setting, a closer look at the results reveals some interesting information. VAS scores No difference was found between the operatively and non‐operatively treated group with respect to the mean VAS scores. As such, both groups seem to suffer equal disability. In previous studies, VAS scores in healthy individuals were found to be 92‐93 [21, 29]. Comparing our data to these numbers, VAS scores in our collective were lower, indicating that both groups suffer from some disability compared to healthy subjects. However, this disability seems to be quite low. Previously, Siebenga et al. found mean VAS scores of 81 and 61, measured 4 years after operative and non‐operative treatment for a type A3 fracture, respectively [39]. Our non‐operatively treated patients perform better. This difference might be explained by two patients developing late neurological deficits in Siebenga’s non‐operative group. Furthermore, in contrast to our series, Siebenga’s cohort of non‐operatively treated patients comprised a few type A3.3 fractures. Outcomes comparable to ours were found by Resch et al. [33]. Four years after operative and non‐operative treatment of type A fractures (mainly type A3), a Hannover spine score of 85 was reported [33]. This outcome measure is fairly comparable to the

TYPE A3 FRACTURE

93

VAS spine score [8, 20]. In another study, a Hannover spine score of 82 was found, nine years after dorsal instrumentation for type A3 fractures [2]. Our results are comparable. In a paper concerning late outcome after non‐operatively treated type A fractures (follow‐up 16 years) a mean VAS score of 58 points was found [32]. Our patients seem to do better, an explanation might be found in the fact that the afore‐mentioned study included subjects with neurological injuries. As known from literature neurological deficit in spinal fracture patients affects the outcome in a negative manner [26]. Similar to our series no late onset neurological deterioration occurred [32]. RMDQ scores Impairment measured by the RMDQ seems reasonably low in both the operatively and non‐operatively treated group (mean scores 3.3 and 3.1, respectively). A RMDQ score of 12.5 was reported in an age‐matched sample of subjects with non‐specific low back pain (duration of onset 1‐6 weeks) [41]. Our patients do remarkably well compared to this figure. Wood et al. found in a prospective setting regarding “burst” fractures a RMDQ score of 8 in operatively treated patients along with a RMDQ score of 4 in those treated non‐operatively [45]. Our data are comparable with respect to the non‐operative group. Concerning the operatively treated group, our patients display a lower RMDQ score (indicating less impairment). An explanation might be found in the presence of dorsal instrumentation which can give rise to mechanical complaints and pain. In our treatment protocol implants are removed. Wood et al. did not mention so, hence possibly the presence of implants resulted in back pain. Others found an average RMDQ score of 3 after operative treatment for a type A3 fracture and a RMDQ score of 9 after non‐operative treatment [39]. Our results are comparable concerning the operative group, but in the non‐operative group these numbers contrast to our data. A possible explanation might be the late onset neurological deficit which occurred in two patients in the non‐operative group [39]. Kraemer et al. found a RMDQ score of 8, measured 4 years after operative as well as non‐operative treatment for “burst” fractures [22]. Our patients show considerably lower impairment, an explanation is not at hand. Operative versus non‐operative Studies trying to find the most favourable treatment in type A3 fractures show contradicting results. Concerning short‐term outcome, Denis et al. found operative treatment to give superior outcome over non‐operative treatment, measured 3

CHAPTER 6

94

years post‐injury [12]. In his series, 17% of patients treated non‐operatively developed neurological problems versus no deterioration in the operative group. In addition, return to work (RTW) was better in those treated operatively. A prospective study by Öner et al., using MRI‐scans, found better results (measured by using Denis’ pain scale) for operative treatment at 2 year follow‐up [12, 28]. In contrast, a recent study comparing outcome 3 years after treatment of L1 “burst” fractures reported less pain (on Denis’ scale) and higher RTW in non‐operatively treated neurologically intact patients [9, 12]. Other authors found equal outcomes 1‐2 years after operative and non‐operative treatment for “burst” fractures [14, 19, 38]. Regarding long‐term outcome, two multi‐centre prospective randomized trials have been published, which show conflicting results [39, 45]. One study reported better results for those patients treated operatively, according to more favourable VAS and RMDQ scores plus higher RTW rates [39]. In contrast, equal outcomes (as measured by the SF‐36 and Oswestry disability questionnaires) and similar RTW rates were reported by Wood et al., four years after treatment of thoracolumbar “burst” fractures [16, 43, 45]. However, RMDQ scores showed better outcomes in those patients treated non‐operatively [45]. Comparable to our findings, other authors found equal outcomes for operative and non‐operative treatment 4‐6 years after “burst” fractures [3, 22]. The duration of follow‐up time did not correlate with functional outcome. This indicates that the functional outcome involving the period of our follow‐up time (2 to 10 years) does not alter considerably. This is in accordance with a study by Andress et al., who could not demonstrate a correlation between outcome and duration of follow‐up time (3 to 9 years in his series) after operative treatment for type A3 fractures [2]. As in our series, no difference was found in outcome between the sub‐classifications (i.e. A3.1, A3.2, A3.3) [2]. Also Weinstein et al. found no correlation between the length of follow‐up time and outcome in non‐operatively treated “burst” fractures [44]. We could not demonstrate differences in outcome 5 years after operative and non‐operative treatment of type A3 fractures. Long‐term complications, such as late onset neurological deficit or late onset pain as reported in literature did not occur [5, 7, 42]. None of our patients required surgery for late onset pain or late onset neurological deficit. When comparing our long‐term VAS and RMDQ scores to literature, patients seem to do reasonably well, and outcome does not seem to deteriorate on the long term for neither group. Considering our results, one can

TYPE A3 FRACTURE

95

conclude that functional outcome in the long term is equal for both treatment modalities and is independent from age and duration of follow‐up time. As such, benefits and drawbacks of both treatment modalities should carefully be taken into account when deciding which treatment is preferred in patients with type A3 fractures without neurological deficit. Both approaches are relatively safe and major complications are rare, so other factors like co‐morbidities, (in)direct costs and short‐term clinical complications, such as urinary tract infections, pressure sores or pulmonary embolism, should play a role in decision making. In this light, it is noteworthy that costs for non‐operative treatment are considerably lower than those for operative treatment [8, 17, 34, 40, 45]. Limitations Certain limitations are present in this study. Data were obtained in a cross‐sectional setting, which has several weaknesses compared to prospective study designs. Furthermore, the small size of the study group might have introduced a type 2 statistical error [4]. When considering the results found, it should be kept in mind that the operative group consisted of more patients suffering from type A3.2 and A3.3 fractures and less type A3.1 fractures. On the other hand, those patients who had sustained a type A3.1 fracture did not show different VAS or RMDQ scores compared to those who had sustained a type A3.2 fracture. To make a definite judgement concerning long‐term outcome after type A3 fractures, larger, prospective studies are needed. To assess outcome in a broad manner, besides questionnaires physical capacity tests could be considered as well [24, 30].

Conclusions

Functional outcome appears to be good five years after operative (dorsal stabilization) as well as non‐operative treatment of type A3 spinal fractures. Both treatment modalities show equal outcomes. We did not observe late onset neurological problems or late onset pain requiring surgery. When making a decision on the treatment for a patient with a type A3 spinal fracture without neurological deficit, factors other than the type of fracture should be taken into

account.

CHAPTER 6

96

References

1. Alanay A, Acaroglu E, Yazici M, Aksoy C, Surat A (2001) The effect of transpedicular intracorporeal grafting in the treatment of thoracolumbar burst fractures on canal remodeling. Eur Spine J 10:512‐516


3. Andreychik DA, Alander DH, Senica KM, Stauffer ES (1996) Burst fractures of the second through fifth lumbar vertebrae. Clinical and radiographic results. J Bone Joint Surg Am 78:1156‐1166



6. Blauth M, Bastian L, Jeanneret B, Knop C, Moulin P, Müller‐Vahl H, Schmidt U, Schratt HE, Wippermann B (1998) Wirbelsäule. In: Tscherne H, Blauth M (eds) Tscherne Unfallchirurgie, vol 3. Springer, Berlin Heidelberg New York, pp 241‐372





11. Daniaux H (1982) Technik und erste Ergebnisse der transpedikulären Spongiosaplastik bei Kompressionsbrüchen im Lendenwirbelsäulenbereich. Acta Chir Austriaca 43 (suppl):79




15. Esses SI, Botsford DJ, Wright T, Bednar D, Bailey S (1991) Operative treatment of spinal fractures with the AO internal fixator. Spine 16:S146‐S150

16. Fairbank JC, Couper J, Davies JB, OʹBrien JP (1980) The Oswestry low back pain disability questionnaire. Physiotherapy 66:271‐273

17. Hitchon PW, Torner JC, Haddad SF, Follett KA (1998) Management options in thoracolumbar burst fractures. Surg Neurol 49:619‐626

18. Klerk LW de, Fontijne WP, Stijnen T, Braakman R, Tanghe HL, van Linge B (1998) Spontaneous remodeling of the spinal canal after conservative management of thoracolumbar burst fractures. Spine 23:1057‐1060

TYPE A3 FRACTURE

97


20. Knop C, Blauth M, Bastian L, Lange U, Kesting J, Tscherne H (1997) Frakturen der thorakolumbalen Wirbelsäule. Spätergebnisse nach dorsaler Instrumentierung und ihre Konsequenzen. Unfallchirurg 100:630‐639







27. Mumford J, Weinstein JN, Spratt KF, Goel VK (1993) Thoracolumbar burst fractures. The clinical efficacy and outcome of nonoperative management. Spine 18:955‐970






33. Resch H, Rabl M, Klampfer H, Ritter E, Povacz P (2000) Operative vs. konservative Behandlung von Frakturen des thorakolumbalen Übergangs. Unfallchirurg 103:281‐288






CHAPTER 6

98


40. Siebenga J, Segers MJM, Leferink VJM, Elzinga MJ, ten Duis HJ, Rommens PM, Patka P (2007) Cost‐effectiveness of the treatment of traumatic thoracolumbar spine fractures: Nonsurgical or surgical therapy? Indian J Orthop 41:332‐336

41. Stratford PW, Binkley JM, Riddle DL, Guyatt GH (1998) Sensitivity to change of the Roland‐Morris Back Pain Questionnaire: part 1. Phys Ther 78:1186‐1196

42. Vaccaro AR, Silber JS (2001) Post‐traumatic spinal deformity. Spine 26:S111‐S118 43. Ware JE, Sherbourne CD (1992) The MOS 36‐item short‐form health survey (SF‐36). I.

Conceptual framework and item selection. Med Care 30:473‐483 44. Weinstein JN, Collalto P, Lehmann TR (1988) Thoracolumbar “burst” fractures treated

conservatively: a long‐term follow‐up. Spine 13:33‐38 45. Wood K, Butterman G, Mehbod A, Garvey T, Jhanjee R, Sechriest V (2003) Operative

compared with nonoperative treatment of a thoracolumbar burst fracture without neurological deficit. A prospective, randomized study. J Bone Joint Surg Am 85‐A:773‐781

46. Yazici M, Atilla B, Tepe S, Calisir A (1996) Spinal canal remodeling in burst fractures of the thoracolumbar spine: a computerized tomographic comparison between operative and nonoperative treatment. J Spinal Disord 9:409‐413

99

Chapter 7

General discussion

CHAPTER 7

100

Introduction

In Chapter 1 we gave an overview of questions and aims of this thesis. In Chapter 2 to Chapter 6 these questions, regarding the range of motion (ROM) after a spinal fracture and the correlation between functional outcome and ROM were discussed. Furthermore, possible techniques to measure the ROM were studied, the mid‐ and long‐term outcomes after non‐operatively treated type A fractures were presented and the optimal treatment in the type A3 “burst” fracture was considered. In this section, we will discuss the most important findings from the afore‐mentioned chapters and compare our results to literature. At the end of this chapter, general conclusions are presented and directions for future research will be given.

Discussion

Patients All patients described in this thesis were between 18 and 60 years old at the time of injury. Patients with neurological deficits due to the spinal fracture were excluded. The occurrence of neurological deficits after a spinal fracture ranged from 18‐22 to 45 percent in three large surveys [26, 40, 48]. As such, patients without neurological injury make up the majority of spinal fracture patients. Within the inclusion criteria (18‐60 years old at the time of injury), the mean age of patients in our series was 37 to 45 years, the male to female proportion was approximately 2 to 1. Patients sustaining a traumatic spinal fracture [26, 37, 38, 74], as well as trauma patients in general [35] display a peak in young males. Since those patients have many productive life‐years ahead, a good‐quality outcome and satisfying return to work rates are indispensable. Functional outcome Although no universally accepted definition of the concept “functional outcome” exists, Liebenson’s description of functional outcome as being “the measurement of a patient’s status, either symptomatically or functionally” seems to be feasible [46]. The International Classification of Functioning, Disability and Health (ICF) describes a person’s health status by measuring restrictions in the categories 1) body function/structure, 2) activity and 3) participation; all of these are influenced by personal and environmental factors [86]. Accordingly, the ICF makes it possible to rate every variable related to patients’ health status. Several instruments are

GENERAL DISCUSSION

101

available for measuring functional outcome, like physical capacity (measuring body function), questionnaires (measuring different domains) and return to work (RTW), the latter evaluating participation [46]. Whether measurement of spinal range of motion (which assesses body function/structure) is a valid instrument for evaluating functional outcome is not clear in literature [17, 56, 59, 65]. Range of Motion For calculating spinal ROM, numerous techniques have been described. One of the first was the Schober test [69]. Subsequently, radiological examination and a variety of external methods have been developed [14, 21, 33, 52, 55]. The Schober test has some important deficiencies: spinal extension and movement in the upper lumbar/lower thoracic region are not assessed [57]. Radiological measurement requires a high dose of radiation, which makes it less suitable for research purposes. Consequently, external, non‐invasive tools are frequently used [60]. In general, these instruments show good reliability. However, certain pitfalls are reported, like the difficulty of palpating bony landmarks and the problem of skin movement over the vertebrae [13, 15, 51, 52, 55, 64]. The SpinalMouse is a computerized external device for quantifying spinal ROM. It was found to be a reliable tool for measuring sagittal spinal ROM (Chapter 2). Inter‐rater intra‐class correlation coefficients (ICC’s) were 0.76 to 0.95. In contrast, a poor agreement (Cohen’s kappa=0.22) was found for the occurrence of outliers from normal values for intersegmental ROM. Comparison of our results to the literature is difficult, since only few studies have been published concerning the SpinalMouse so far [29, 50]. Mannion et al. found comparable inter‐rater ICC’s, ranging from 0.81 to 0.86 [50]. Similar to our findings, they concluded that the device was reliable in assessing sagittal spinal ROM, though it appeared less valid in measuring intersegmental spinal ROM. A recent study, validating the SpinalMouse in assessing lumbar ROM, found ICC’s slightly lower than ours, varying from 0.60 to 0.85 [29]. The authors found the instrument to be valid in measuring lumbar ROM, intersegmental ROM was measured inadequately. In the category of non‐invasive external devices, one can choose from a variety of instruments, like goniometers, skin markers, inclinometers, spondylometers or opto‐electronic systems. Given the results we found, the SpinalMouse could be used for measuring sagittal spinal ROM in both research and clinical environments. For assessing intersegmental ROM it seems less suitable. Advantages of the SpinalMouse are that it is easy to use, and results can be easily

CHAPTER 7

102

saved in a computer and attached to a patient’s file. We conclude that the SpinalMouse may find clinical application in follow‐up of patients with back complaints, e.g. after a spinal fracture. Its measurements could be used for objective evaluation of changes in spinal mobility during treatment.

It is uncertain whether spinal ROM correlates with subjective impairment. In Chapter 3 we measured the relationship between spinal ROM and VAS and RMDQ scores in spinal fracture patients (treated operatively and non‐operatively) as well as in healthy controls. No correlation between ROM and disability could be demonstrated. In literature, contradictory results have been reported concerning the association between ROM and disability. Yurac et al. found no relationship between residual segmental ROM and outcome (as measured by the Greenough Low Back Outcome Scale) after removal of instrumentation for spinal fractures [87]. Another study found that thoracolumbar ROM correlated poorly to moderately with disability, and did not appear to be a valid measure of disability [65]. Neither a relationship was found between lumbar ROM and impairment when using the Oswestry Disability Index and the Waddell Disability Index [59]. Also Gronblad et al. could not demonstrate a correlation between ROM and outcome [28]. These findings support our data that spinal ROM is of no (or minimal) influence on impairment. On the contrary, others did find a correlation between spinal ROM and disability [17, 53, 56]. Cox et al. reported that a patient’s cognitive state is strongly related with ROM. Therefore, fear for example will be of influence on the ROM, reducing the (voluntary) ROM [17]. In patients with a spinal fracture, this fear is conceivable. In summary, spinal ROM does not seem to correlate with subjective impairment, and accordingly appears to be unsuitable for assessing functional outcome. An interesting finding from our research was that spinal fracture patients treated operatively had lower ROM than healthy subjects. The reasons why ROM decreases after operative treatment of a spinal fracture are unknown. Possibly scar tissue influences ROM, or fear of maximum bending may play a role, the latter being reported by Cox et al. [17]. It seems unlikely that the fusion of one single level will result in a decrease of ROM in the whole spinal column. Literature concerning ROM after operative treatment of spinal fractures reveals conflicting results, reporting decreased as well as normal spinal ROM [21, 34, 39]. Whether the decreased ROM is clinically important however is questionable, since the remaining ROM does not seem to correlate with disability.

GENERAL DISCUSSION

103

Type A fractures treated non‐operatively The most common spinal fracture is the type A fracture (Comprehensive Classification [48]), accounting for 66% of all spinal fractures. Of these, 86% present without neurological deficits [48]. Often, these fractures are treated non‐operatively. In Chapter 4 we measured functional outcome 5 years after non‐operative treatment for a type A spinal fracture. In the dimension of body function and structure, approximately 1/3 of the patients performed below normal levels in the exercise tests. Restrictions in activities were quite low, given the VAS and RMDQ scores found (79 and 5.2, respectively). Finally, restrictions in participation/quality of life showed no significant differences between patients and healthy subjects concerning SF‐36 scores [32]. Ten per cent of the subjects had stopped working and received social security benefits, 24% had arranged changes in their work and 14% had changed their job. Quite similar RTW rates were found by other authors [12, 16, 66]. When judging RTW rates however, one has to keep in mind that these figures are highly influenced by different national legal and compensation systems [30, 47]. The exercise tests comprised an ergometry test (measuring cardiopulmonary condition) and a lifting test. Results of the latter indicate that a large proportion of patients have difficulty with lifting a box from the floor to a table. The “problem area” (for example trunk musculature, leg musculature, incapacitating pain) responsible for this reduced lifting capacity cannot be indicated by this test, interesting however is that a substantial part of patients are confronted with a decreased lifting ability after non‐operative treatment of a type A fracture. Summarizing, patients are slightly restricted in their activities, this restriction however does not influence their participation/quality of life, since SF‐36 scores were similar to healthy persons. Our study had the relatively unique feature that it measured outcome in all the domains of the ICF, whereas most studies cover only a distinct fraction of the ICF [77]. There is a clear paucity in data concerning physical capacity after a spinal fracture. To our best knowledge, no other publication is available regarding residual physical capacity in non‐operatively treated spinal fracture patients, which makes comparison to other series impossible. Leferink et al. measured physical capacity after operative treatment for type A fractures [45]. Their results were comparable to ours. Physical capacity testing is rather uncommon in spinal fracture outcome research, whilst we showed that the physical tests might give a good reflection of the outcome. The dynamic lifting test correlated quite strongly with the RMDQ, VAS and SF‐36 physical index. Further research in this field is needed to shed more light on the physical capacity after a spinal fracture.

CHAPTER 7

104

With respect to the duration of follow‐up time in functional outcome research after spinal fractures, most of the published data concentrate on relatively short‐term results (1‐2 years) [19, 22, 36, 63]. Literature regarding long‐term outcome (10 years) is scarce [25, 67, 83]. In the long term, pain may arise due to changed facet joint motion, hyperextension of adjacent spinal regions, fatigue pain from the soft tissues and ongoing degenerative processes [5, 63, 79, 81]. Even late onset neurological deficit can arise years after the spinal fracture [7]. In Chapter 5 we measured functional outcome in the same cohort of patients 4 and 10 years after non‐operative treatment for type A fractures without neurological deficits. Patients were only slightly disabled at both measurement times, outcomes were equal at both instants. Three patients (6%) had a poor long‐term outcome, though none of the patients required surgery for late onset pain or late onset neurological deficit. Late onset pain requiring operative interference has been reported for type A fractures [5]. Numbers on how many patients need operative treatment for late arising pain or neurological symptoms are not available in literature. We found a status quo in functional outcome between 4 and 10 years post‐injury. Previous studies mention a steady state in functional outcome ranging from 2 to 4 years after a spinal fracture [1, 16, 72]. Data on outcomes after 2 to 4 years, however, are scarcely available. Andress et al. measured functional outcome at 4, 5 and 7 years after operative treatment of type A3 fractures [1]. As in our series, no differences were found in functional outcomes at the distinct measurement moments. In summary, the course of functional outcome after a non‐operatively treated type A spinal fracture seems to be in a steady state as from about two years post‐injury. This is in accordance with van der Sluis et al., though be it those numbers concern severely injured patients [76]. Consequently, it seems to be justified to discharge spinal fracture patients from follow‐up control two years post‐injury, as is current practice at our institution. Operative versus non‐operative treatment The debate which treatment modality is the most advisable in spinal fractures has been going on for decades [62, 72]. Comparison of functional outcomes after operative and non‐operative treatment for spinal fractures (type A, B and C fractures according to the Comprehensive Classification [48]) did not demonstrate a difference in our series in Chapter 3. Both patient groups had only minor impairments 3 to 4 years post‐injury. In literature, good outcome is reported after operative as well as after non‐operative treatment for spinal fractures. Most

GENERAL DISCUSSION

105

authors find patients to be slightly impaired both after operative and non‐operative treatment, with RTW rates ranging from 50% to 100% [19, 39, 43, 54, 67, 78]. Studies directly comparing operative and non‐operative treatment in cohorts comprising all types of spinal fractures reveal a trend towards equal outcomes for both treatment modalities (see Table 1).

Table 1 Outline of studies comparing operative and non‐operative treatment in spinal fractures

Author Classification n Outcome measure

Results (operative vs. non‐operative)

Briem [9] CC: type A, B, C fractures

133 VAS spine score, SF‐36, RTW

equal

Domenicucci [22] Denis: “burst” and wedge fractures

31 Denis outcome scale

equal

El Awad [23] CC: type A, B, C fractures

100 Denis outcome scale, neurology

equal

Öner [63] CC: type A, B, C fractures


operative better

Post [chapter 3] CC: type A, B, C fractures

76 VAS spine score, RMDQ

equal

CC: Comprehensive Classification. SF‐36: Short‐Form 36 health survey. RTW: return to work. RMDQ: Roland‐Morris disability questionnaire

The type A3 “burst” fracture Especially the type A3 (Comprehensive Classification [48]) “burst” fracture and its optimal treatment has gained much attention in literature. It has been the issue of many papers, reporting superior, inferior as well as equal outcomes for operative treatment compared to non‐operative treatment [2, 11, 20, 36, 41, 70‐72, 84]. To evaluate both treatment modalities, we measured the long‐term (5 years) functional outcome of this type of fracture in Chapter 6. No differences in outcomes between these groups could be demonstrated. Furthermore, none of the patients required surgery for late onset pain or late onset neurological deficit; age and duration of follow‐up time did not correlate with outcome. In summary, functional outcome after a type A3 spinal fracture is equal for both treatment modalities in the long term and is independent from age and duration of follow‐up time. As such, benefits and drawbacks of both treatment modalities should be carefully taken into account when deciding which treatment is preferred in a neurologically intact patient with a type A3 fracture. Both approaches are relatively safe and major complications are rare (in a large survey by Knop et al., 15% of patients treated by dorsal implantation suffered from, mostly minor, complications [37]). Therefore, other factors like (in)direct costs and short‐term

CHAPTER 7

106

clinical complications (such as urinary tract infections, pressure sores or pulmonary embolism) should play a role in decision making. In this light, it is noteworthy that costs for non‐operative treatment are considerably lower than those for operative treatment [9, 68, 84]. Studies comparing operative and non‐operative treatment in type A3 “burst” fractures find conflicting results (see Table 2). Our results were obtained in a retrospective setting, like most studies concerning outcome after spinal fractures (approximately 90% of all reports in literature are retrospective [82]). Only two prospective, randomized, multi‐centre studies are available, which show contrasting results [72, 84]. The relatively low incidence of traumatic spinal fractures and the specialized care needed make it difficult to form large study groups in spinal fracture research, which may introduce a type II statistical error (i.e. not detecting an existing difference due to small study groups). According to the literature, 87% of the published papers concerning spinal surgery display a type II error [3].

Table 2 Outline of studies comparing operative and non‐operative treatment in type A3 “burst” fractures

Author Classification n Outcome measure Results (operative vs. non‐operative)

Andreychik [2] Denis: “burst” fractures


equal

Butler [11] Denis: “burst” fractures

31 Denis outcome scale, RTW

non‐operative better

Denis [20] Denis: “burst” fractures


operative better

Knight [36] Denis: “burst” fractures


equal

Kraemer [41] Denis: “burst” fractures

24 RMDQ, SF‐36, RTW

equal

Seybold [70] Denis: “burst” fractures

42 Dallas pain questionnaire

equal

Shen [71] Denis: “burst” fractures

80 Denis outcome scale, GLBOS, RTW

equal

Siebenga [72] CC: type A3 fractures

32 VAS spine score, RMDQ, RTW

operative better

Wood [84] Denis: “burst” fractures

47 RMDQ, SF‐36, ODI

equal, trend towards non‐operative better

Post [chapter 6] CC: type A3 fractures

63 VAS spine score, RMDQ

equal

CC: Comprehensive Classification. SF‐36: Short‐Form 36 health survey. RTW: return to work. RMDQ: Roland‐Morris disability questionnaire. GLBOS: Greenough low back outcome score. ODI: Oswestry disability index

GENERAL DISCUSSION

107

Ventral versus dorsal procedure All operatively treated patients mentioned in this thesis were treated by dorsal stabilization with the Universal Spine System (USS). In our clinic, ventral procedures are carried out too infrequent to include enough patients for reliable outcome research concerning this approach. The dorsal approach is the most common operative procedure to treat spinal fractures [82]. With regard to the ventral approach, less literature is available. Ventral surgery could produce a more complete and reliable decompression of the spinal canal and provides a better sagittal alignment than the dorsal technique does [4, 18, 82]. Radiological appearance, however, does not correlate with functional outcome. On the other hand, the ventral approach requires a more sophisticated technique and may result in serious adverse effects [24]. Okuyama et al. described good results after ventral surgery, as well as did Ghanayem et al. [27, 61]. When comparing outcomes after dorsal and ventral approaches, no differences were found by Briem et al. and Verlaan et al. [10, 82]. Predictors of outcome Functional outcome appeared to be independent from age, gender, fracture sub‐type (A3.1 and A3.2) or treatment modality (operative versus non‐operative) as described in Chapters 3, 5 and 6. This is in accordance with other authors, who found age and gender or fracture type to be unrelated to outcome [1, 9, 31, 49, 73]. Outcome between the type A3 sub‐types appears to be equal, though a negative correlation has been reported between increasing severity of lesions (CC) and outcome (measured by the VAS spine score and SF‐36) [9]. Also others found a correlation between the Comprehensive Classification and Frankel scales, more severe fracture‐types (from type A to type C) resulted in poorer neurological func‐tioning [37, 42, 48]. Neurological deficit itself, on the other hand, is known to have a major negative impact on outcome after a spinal fracture [54]. In this thesis we did not consider the socio‐economic status of patients, in future studies these characteristics should be investigated more extensively. With respect to socio‐economic and psychosocial variables, Harris et al. found that lower education level and the presence of chronic illnesses had a strong negative correlation with outcome (measured by the SF‐36) after major trauma [31]. Slover et al. found a negative correlation between the number of co‐morbidities and the difference scores in the Oswestry Disability Index after lumbar surgery [75]. Even so smoking, headache and depression had a negative impact on the baseline and difference scores in the Oswestry Disability Index [75]. Briem et al. found good

CHAPTER 7

108

mental health to have a positive effect on outcome after operative treatment for a spinal fracture [10]. Recently, psychosocial variables (i.e. fear‐avoidance beliefs and depression) were found to explain 20% of variance in outcome scores after spinal surgery [49]. Possibly psychosocial and socio‐economic factors are of major influence on outcome, and may play a larger role in outcome than assumed in the present time. Questionnaires The use of various outcome measures by different authors makes comparison sometimes difficult. For measuring back‐related functional outcome approximately 40 different questionnaires are available [58]. According to Bombardier, a questionnaire measuring outcome in spinal disorders should at least cover the following five domains: back‐specific function, generic health status, pain, work disability and patient satisfaction [8]. To assess outcome in a uniform manner and face all the afore‐mentioned domains, a proposed ideal tool would consist of a combination of the RMDQ or Oswestry Disability Index, the SF‐36, together with the Work Limitation Questionnaire and the Patient Satisfaction Scale [8]. Whether such a large set of questionnaires will be used in clinical practice is doubtful. In our opinion, the Roland‐Morris Disability Questionnaire, VAS spine score and Oswestry Disability Index are helpful tools for measuring outcome, since they have good psychometric properties and are widely used, which makes comparison possible. The Denis outcome scale, although popular in literature, might be too constrained and its reliability and validity are unknown. What is good outcome in spinal surgery? According to a recent paper that tried to quantify good outcome from patients’ and surgeons’ perspective, the main parameters determining good outcome are: absence of pain, high patients’ satisfaction, low disability and good social reintegration [30]. At present, none of the frequently used questionnaires covers all these proposed parameters, and maybe a new questionnaire for outcome research after a spinal fracture should be developed. It should make an effort to involve all the domains of the ICF, furthermore comparison would be easier if this questionnaire was to be used widely. Classification schemes A classification scheme is a model of an observable fact that is supposed to enlighten the severity of the injury and the possible consequences. An ideal classification of different injury patterns should provide a reasonable estimation of the outcome, and give direction to treatment. As stated in the introduction, the

GENERAL DISCUSSION

109

presently most used classifications (Denis classification and Comprehensive Classification) show a moderate reliability and repeatability [6, 42, 85]. Furthermore, both schemes do not entirely recognize the role of the posterior ligamentous complex (PLC). Therefore, it is not surprising that both systems cannot predict outcome (with exception of gross neurological deficits) very precisely. In contrast, a study by Öner et al., using MRI‐scans, was able to predict poor outcome due to PLC lesions [63]. Possibly the ThoracoLumbar Injury Classification and Severity Score (TLICS), as recently developed by Vaccaro et al., might be able to predict the outcome and additionally give more direction to treatment [44, 80]. Outcome research following implementation of this scheme should take place in the future.

Conclusions

Important findings of this thesis are:

• The SpinalMouse offers a quick and reliable method of measuring sagittal spinal ROM.

• The measurement of spinal ROM is not a suitable tool for evaluating disability. • A clear paucity exists regarding literature concerning physical capacity after a

spinal fracture, whereas we found a considerable number of patients to be restricted in their physical ability after non‐operative treatment of type A spinal fractures.

• The results of dynamic lifting tests seem to correlate well with subjective impairment.

• Operative treatment of type A spinal fractures (including the type A3 fracture) does not give superior functional outcome compared to non‐operative treatment.

• The majority of patients are slightly impaired after a spinal fracture without neurological deficit. This impairment does not influence their quality of life.

• Functional outcome is steady from about four years until ten years after non‐operative treatment of type A spinal fractures without neurological deficits.

• Functional outcome after a spinal fracture seems to be influenced by factors other than age, gender, treatment and fracture type classification.

CHAPTER 7

110

Directions for future research

• Concerning fracture classification, the first results after usage of the ThoracoLumbar Injury Classification and Severity Score (TLICS) will have to be waited for. The system itself already showed good reliability, now data on functional outcome after application of this scheme will have to clarify whether it is appropriate for directing treatment.

• Given the lack of data regarding physical capacity after a spinal fracture, future research should focus on this issue.

• To provide more level‐I evidence in spinal fracture outcome research, large, prospective randomized (multi‐centre?) studies are clearly needed. Possibly, internet‐based data gathering as recently presented by the Spine Trauma Study Group (EPOST) and by Knop et al. might offer an opportunity to form large study groups.

• The search for predictors of outcome should (also) focus on socio‐economic and psychosocial aspects, since these might play an important role and are relatively underexposed in present studies.

References


2. Andreychik DA, Alander DH, Senica KM, Stauffer ES (1996) Burst fractures of the second through fifth lumbar vertebrae. Clinical and radiographic results. J Bone Joint Surg Am 78:1156‐1166


4. Been HD (1991) Anterior decompression and stabilization of thoracolumbar burst fractures using the Slot‐Zielke‐device. Acta Orthop Belg 57 Suppl 1:144‐161


6. Blauth M, Bastian L, Knop C, Lange U, Tusch G (1999) Interobserverreliabilität bei der Klassifikation von thorakolumbalen Wirbelsäulenverletzungen. Orthopäde 28:662‐681


8. Bombardier C (2000) Outcome assessments in the evaluation of treatment of spinal disorders: summary and general recommendations. Spine 25:3100‐3103


GENERAL DISCUSSION

111

10. Briem D, Lehmann W, Ruecker AH, Windolf J, Rueger JM, Linhart W (2004) Factors influencing the quality of life after burst fractures of the thoracolumbar transition. Arch Orthop Trauma Surg 124:461‐468



13. Chiarello CM, Savidge R (1993) Interrater reliability of the Cybex EDI‐320 and fluid goniometer in normals and patients with low back pain. Arch Phys Med Rehabil 74:32‐37

14. Chiou WK, Lee YH, Chen WJ, Lin YH (1996) A non invasive protocol for the determination of lumbar spine mobility. Clin Biomech 11:474‐480

15. Chockalingam N, Dangerfield PH, Giakas G, Cochrane T (2002) Study of marker placements in the back for opto‐electronic motion analysis. Stud Health Technol Inform 88:105‐109



18. Dai LY, Jiang SD, Wang XY, Jiang LS (2007) A review of the management of thoracolumbar burst fractures. Surg Neurol 67:221‐231

19. Defino HL, Canto FR (2007) Low thoracic and lumbar burst fractures: radiographic and functional outcomes. Eur Spine J 16:1934‐1943




23. El Awad AA, Othman W, Al Moutaery KR (2002) Treatment of thoracolumbar fractures. Saudi Med J 23:689‐694

24. Esses SI, Botsford DJ, Kostuik JP (1990) Evaluation of surgical treatment for burst fractures. Spine 15:667‐673


26. Gertzbein SD (1992) Scoliosis Research Society. Multicenter spine fracture study. Spine 17:528‐540

27. Ghanayem AJ, Zdeblick TA (1997) Anterior instrumentation in the management of thoracolumbar burst fractures. Clin Orthop Relat Res 335:89‐100

28. Gronblad M, Hurri H, Kouri JP (1997) Relationships between spinal mobility, physical performance tests, pain intensity and disability assessments in chronic low back pain patients. Scand J Rehabil Med 29:17‐24

29. Guermazi M, Ghroubi S, Kassis M, Jaziri O, Keskes H, Kessomtini W, Ben Hammouda I, Elleuch MH (2006) Validity and reliability of Spinal Mouse to assess lumbar flexion. Ann Readapt Med Phys 49:172‐177

CHAPTER 7

112

30. Haefeli M, Elfering A, Aebi M, Freeman BJ, Fritzell P, Guimaraes Consciencia J, Lamartina C, Mayer M, Lund T, Boos N (2008) What comprises a good outcome in spinal surgery? A preliminary survey among spine surgeons of the SSE and European spine patients. Eur Spine J 17:104‐116

31. Harris IA, Young JM, Rae H, Jalaludin BB, Solomon MJ (2007) Factors associated with back pain after physical injury: a survey of consecutive major trauma patients. Spine 32:1561‐1565

32. Jenkinson C, Coulter A, Wright L (1993) Short form 36 (SF36) health survey questionnaire: normative data for adults of working age. BMJ 306:1437‐1440


34. Junge A, Gotzen L, von Garrel T, Ziring E, Giannadakis K (1997) Die monosegmentale Fixateur interne: Instrumentation und Fusion in der Behandlung von Frakturen der thorakolumbalen Wirbelsäule. Indikation, Technik und Ergebnisse. Unfallchirurg 100:880‐887

35. Kingma J (1994) The young male peak in different categories of trauma victims. Percept Mot Skills 79:920‐922


37. Knop C, Bastian L, Lange U, Oeser M, Zdichavsky M, Blauth M (2002) Complications in surgical treatment of thoracolumbar injuries. Eur Spine J 11:214‐226

38. Knop C, Blauth M, Buhren V, Hax PM, Kinzl L, Mutschler W, Pommer A, Ulrich C, Wagner S, Weckbach A, Wentzensen A, Worsdorfer O (1999) Operative Behandlung von Verletzungen des thorakolumbalen Übergangs. Teil 1: Epidemiologie. Unfallchirurg 102:924‐935


40. Knop C, Reinhold M, Roeder C, Staub L, Schmid R, Beisse R, Buhren V, Blauth M (2006) Internet based multicenter study for thoracolumbar injuries: a new concept and preliminary results. Eur Spine J 15:1687‐1694


42. Kriek JJ, Govender S (2006) AO‐classification of thoracic and lumbar fractures: reproducibility utilizing radiographs and clinical information. Eur Spine J 15:1239‐1246

43. Krompinger WJ, Fredrickson BE, Mino DE, Yuan HA (1986) Conservative treatment of fractures of the thoracic and lumbar spine. Orthop Clin North Am 17:161‐170



46. Liebenson C, Yeomans S (1997) Outcomes assessment in musculoskeletal medicine. Manual Therapy 2:67‐74

GENERAL DISCUSSION

113

47. MacKenzie EJ, Morris JA, Jurkovich GJ, Yasui Y, Cushing BM, Burgess AR, DeLateur BJ, McAndrew MP, Swiontkowski MF (1998) Return to work following injury: the role of economic, social, and job‐related factors. Am J Public Health 88:1630‐1637


49. Mannion AF, Elfering A, Staerkle R, Junge A, Grob D, Dvorak J, Jacobshagen N, Semmer NK, Boos N (2007) Predictors of multidimensional outcome after spinal surgery. Eur Spine J 16:777‐786

50. Mannion AF, Knecht K, Balaban G, Dvorak J, Grob D (2004) A new skin‐surface device for measuring the curvature and global and segmental ranges of motion of the spine: reliability of measurements and comparison with data reviewed from the literature. Eur Spine J 13:122‐136

51. Mayer RS, Chen IH, Lavender SA, Trafimow JH, Andersson GB (1995) Variance in the measurement of sagittal lumbar spine range of motion among examiners, subjects, and instruments. Spine 20:1489‐1493


53. Mayer TG, Tencer AF, Kristoferson S, Mooney V (1984) Use of noninvasive techniques for quantification of spinal range‐of‐motion in normal subjects and chronic low‐back dysfunction patients. Spine 9:588‐595



56. Mellin G (1987) Correlations of spinal mobility with degree of chronic low back pain after correction for age and anthropometric factors. Spine 12:464‐468

57. Miller MH, Lee P, Smythe HA, Goldsmith CH (1984) Measurements of spinal mobility in the sagittal plane: new skin contraction technique compared with established methods. J Rheumatol 11:507‐511

58. Muller U, Duetz MS, Roeder C, Greenough CG (2004) Condition‐specific outcome measures for low back pain. Part I: validation. Eur Spine J 13:301‐313



61. Okuyama K, Abe E, Chiba M, Ishikawa N, Sato K (1996) Outcome of anterior decompression and stabilization for thoracolumbar unstable burst fractures in the absence of neurologic deficits. Spine 21:620‐625

62. Osebold WR, Weinstein SL, Sprague BL (1981) Thoracolumbar spine fractures. Results of treatment. Spine 6:13‐34


64. Petersen CM, Johnson RD, Schuit D, Hayes KW (1994) Intraobserver and interobserver reliability of asymptomatic subjects’ thoracolumbar range of motion using the OSI CA 6000 Spine Motion Analyzer. J Orthop Sports Phys Ther 20:207‐212

CHAPTER 7

114





69. Schober P (1937) Lendenwirbelsäule und Kreuzschmerzen. Munch Med Wochenschr 84:336‐338

70. Seybold EA, Sweeney CA, Fredrickson BE, Warhold LG, Bernini PM (1999) Functional outcome of low lumbar burst fractures. A multicenter review of operative and nonoperative treatment of L3‐L5. Spine 24:2154‐2161



73. Singer BR (1995) The functional prognosis of thoracolumbar vertebrae fractures without neurological deficit: a long‐term follow‐up study of British Army personnel. Injury 26:519‐521

74. Singer BR, McLauchlan GJ, Robinson CM, Christie J (1998) Epidemiology of fractures in 15,000 adults: the influence of age and gender. J Bone Joint Surg Br 80:243‐248

75. Slover J, Abdu WA, Hanscom B, Weinstein JN (2006) The impact of comorbidities on the change in short‐form 36 and oswestry scores following lumbar spine surgery. Spine 31:1974‐1980

76. Sluis CK van der, ten Duis HJ, Geertzen JH (1995) Multiple injuries: an overview of the outcome. J Trauma 38:681‐686

77. Swinkels RAHM (2004) The ICF classification as a system for structuring outcome measurement. Physiotherapy Singapore 7:7‐13


79. Tropiano P, Huang RC, Louis CA, Poitout DG, Louis RP (2003) Functional and radiographic outcome of thoracolumbar and lumbar burst fractures managed by closed orthopaedic reduction and casting. Spine 28:2459‐2465

80. Vaccaro AR, Lehman RA, Hurlbert RJ, Anderson PA, Harris M, Hedlund R, Harrop J, Dvorak M, Wood K, Fehlings MG, Fisher C, Zeiller SC, Anderson DG, Bono CM, Stock GH, Brown AK, Kuklo T, Öner FC (2005) A new classification of thoracolumbar injuries: the importance of injury morphology, the integrity of the posterior ligamentous complex, and neurologic status. Spine 30:2325‐2333

81. Vaccaro AR, Silber JS (2001) Post‐traumatic spinal deformity. Spine 26:S111‐S118 82. Verlaan JJ, Diekerhof CH, Buskens E, van der Tweel I, Verbout AJ, Dhert WJ, Öner FC (2004)

Surgical treatment of traumatic fractures of the thoracic and lumbar spine: a systematic review of the literature on techniques, complications, and outcome. Spine 29:803‐814

GENERAL DISCUSSION

115

83. Weinstein JN, Collalto P, Lehmann TR (1988) Thoracolumbar “burst” fractures treated conservatively: a long‐term follow‐up. Spine 13:33‐38

84. Wood K, Butterman G, Mehbod A, Garvey T, Jhanjee R, Sechriest V (2003) Operative compared with nonoperative treatment of a thoracolumbar burst fracture without neurological deficit. A prospective, randomized study. J Bone Joint Surg Am 85‐A:773‐781

85. Wood KB, Khanna G, Vaccaro AR, Arnold PM, Harris MB, Mehbod AA (2005) Assessment of two thoracolumbar fracture classification systems as used by multiple surgeons. J Bone Joint Surg Am 87:1423‐1429

86. World Health Organization (2001) International Classification of Functioning, Disability and Health Problems. WHO, Geneva

87. Yurac R, Marre B, Urzua A, Munjin M, Lecaros MA (2006) Residual mobility of instrumented and non‐fused segments in thoracolumbar spine fractures. Eur Spine J 15:864‐875


117

Chapter 8

Summary

CHAPTER 8

118

An introduction to different aspects regarding spinal fractures is presented in Chapter 1. The incidence of traumatic thoracolumbar spinal fractures without neurological deficit in the Netherlands is approximately 1.2 per 10,000 per year (in the age group of 20 to 60 years). Concerning total medical costs in the Netherlands, spinal fractures rank 7th (3.8%) in total trauma costs, with a mean cost of EUR 6,600 per patient. The history of classification schemes is discussed, with special reference to the Comprehensive Classification, which is used in this thesis. The Denis classification and the Comprehensive Classification, being the most frequently used schemes, have some important deficiencies: reliability has been shown to be moderate and both systems do not recognize the important role of the posterior ligamentous complex (PLC) in maintaining spinal stability. Consequently, new developments with respect to the use of the MRI in detecting injuries to the PLC are forthcoming, like the ThoracoLumbar Injury Classification and Severity Score (TLICS), a new scheme which does take into account the PLC. Several aspects of operative and non‐operative treatment are considered. Benefits of the operative approach are the improvement of spinal alignment, decreased deformity, early mobilization and rehabilitation and sometimes improvement in neurological function. On the other hand, non‐operative treatment lacks the risks of surgery, such as deep wound infection, iatrogenic neurological injury and implant failure. The history of operative treatment is explored, as well as different techniques of operative treatment (the dorsal and ventral approaches and the recently developed kyphoplasty). An introduction to the concept of functional outcome (i.e. the measurement of a patient’s status, either symptomatically or functionally) is portrayed. In addition, the different health‐models as published by the World Health Organization are described, being the International Classification of Impairments, Disabilities and Handicaps (ICIDH) and its successor, the International Classification of Functioning, Disability and Health (ICF). The ICF describes a person’s health status by measuring restrictions in the categories 1) body function/structure, 2) activity and 3) participation; all of these are influenced by personal and environmental factors. Various outcome measures, like physical capacity, return to work and health related quality of life are discussed. At the end, a review of the literature is presented. From this review, interesting findings worth mentioning include residual radiological vertebral manifestation (like anterior wedge angle, corporal height) not correlating with functional outcome. Furthermore, neurological deficits have major negative effects on the outcome. Finally, the aim and outline of this thesis are presented.

SUMMARY

119

The effect of a spinal fracture and its treatment on resulting spinal range of motion (ROM) is uncertain. Furthermore, the relationship between ROM and impairment is not clear. Before these issues can be studied, however, another topic is how to measure the ROM. Due to the restrictions of radiographs in clinical practice (radiation exposure), many non‐invasive, external methods for measuring spinal ROM have been developed. In Chapter 2, we tested the SpinalMouse, a non‐invasive computer‐aided device for measuring sagittal spinal ROM. When run along the back, the device measures spinal ROM as well as intersegmental ROM. A rather distinctive feature of the apparatus is its ability to measure thoracic ROM as well, whereas literature concerning spinal ROM mostly concentrates on lumbar ROM. To assess inter‐rater reliability, two raters measured 111 subjects. Correlation coefficients were r=0.90 for flexion, r=0.85 for extension and r=0.90 for total inclination. Intra‐class correlation coefficients were 0.95 for flexion, 0.92 for extension and 0.95 for total inclination. A poor agreement (Cohen’s kappa=0.22) was found for the occurrence of outliers from normal values for intersegmental ROM. Following these figures, the SpinalMouse appears to be a reliable tool for measuring sagittal spinal ROM. For measuring intersegmental ROM, though, it is not reliable. Considering the small load it puts on patients and the short amount of time the measurement involves, the SpinalMouse can be a useful tool to survey sagittal spinal mobility, for example during treatment for back‐related disorders. At the end of Chapter 2, we make some recommendations for successful usage in clinical practice.

In Chapter 3 we studied the ROM after a spinal fracture, as well as the relationship between ROM and disability. We measured thoracolumbar ROM in operatively and non‐operatively treated patients (n=76, average 3.7 years follow‐up) as well as in healthy controls (n=41). To evaluate impairment after a spinal fracture and to compare treatment modalities, we measured disability using two disease‐specific questionnaires: the VAS spine score (VAS) and the Roland‐Morris Disability Questionnaire (RMDQ). Scores were determined in operatively and non‐operatively treated patients as well as in healthy controls. In order to study the relation between ROM and subjective back complaints, we calculated the correlation between thoracolumbar ROM and scores derived from the VAS and RMDQ. Operatively treated patients were found to have lower thoracolumbar ROM than controls (56.7° versus 70.0°, respectively). There was no difference between operatively treated and non‐operatively treated patients (56.7° versus 62.7°,

CHAPTER 8

120

respectively); nor was a difference found between non‐operatively treated patients and controls. Correlation between ROM and subjective impairment was very weak and was only significant for ROM and RMDQ scores in the whole study group (rho=‐0.25; p<0.01). Patients were more impaired than controls, but there was no difference between operatively and non‐operatively treated patients (VAS score 76.3 versus 72.6; RMDQ score 4.5 versus 4.4, respectively). Literature concerning the ROM after a spinal fracture shows contradictory results. Some authors find a decreased ROM whilst others find a normal ROM. In addition, the relation between ROM and impairment is not clear in literature. We conclude that sagittal thoracolumbar ROM four years after operative treatment of a spinal fracture is less than thoracolumbar ROM of healthy individuals. Why operative treatment of thoracolumbar fractures results in lower spinal ROM is unclear. Patients who sustain a spinal fracture are more impaired than healthy controls, the ROM does not seem to correlate with this impairment, however. Both kinds of treatment (operative and non‐operative) result in similar impairment rates.

The type A spinal fracture (Comprehensive Classification) without neurological deficit is the most common type of spinal fracture. Often, these fractures are treated non‐operatively. Chapter 4 describes the functional outcome after non‐operative treatment of type A thoracolumbar spinal fractures without neurological deficit. Functional outcome was determined following the WHO’s International Classification of Functioning, Disability and Health (ICF), measuring restrictions in body function and structure, restrictions in activities, and restrictions in participation/quality of life. Thirty‐three patients (mean age 50.5 years, mean follow‐up time 5.3 years) were included. Restrictions in body function and structure were measured by physical tests (dynamic lifting test and bicycle ergometry test), restrictions in activities were measured by means of questionnaires (VAS and RMDQ). Restrictions in participation/quality of life were assessed with the Short Form 36 questionnaire (SF‐36) and by means of return to work (RTW) status. Thirty‐seven per cent of the patients were not able to perform the dynamic lifting test within normal range. This dynamic lifting test assessed the patients’ ability to raise a weight from the floor to a 75cm‐high table, repetitively. The examination was stopped when patients experienced exhaustion or discomfort (psychophysical testing). As such, one third of the patients more rapidly experienced feelings of distress than healthy people performing the same task. In the ergometry test (measuring cardiopulmonary conditions), 41% of the patients performed below the

SUMMARY

121

lowest normal value, 36% of the patients achieved a high VO2‐max. Mean VAS score was 79, the mean RMDQ score was 5.2. No significant differences between patients and healthy subjects were found in SF‐36 scores. The dynamic lifting test correlated quite strongly with the RMDQ, VAS and SF‐36 physical index (r=‐0.62, r=0.71 and r=0.59, respectively). Concerning the return to work status: 10% of the subjects had stopped working and received social security benefits, 24% had arranged changes in their work and 14% had changed jobs. Other authors find quite similar RTW rates, though it should be kept in mind that different national social security laws influence the RTW rates. With regard to the physical capacity tests, there is lack of data relating to this issue. Physical capacity testing is rather exceptional in spinal fracture outcome research, while we showed that the physical tests might give a good reflection of the outcome. In conclusion, patients seem to do reasonably well 5 years after non‐operative treatment of a type A spinal fracture, although outcome is diverse in the different categories and the physical functioning seems restricted in a considerable number of patients. The restrictions in activities that patients experience do not influence their participation/quality of life, however.

Regarding the duration of follow‐up time in functional outcome research after spinal fractures, most of the available studies focus on quite short‐term results (1‐2 years). Literature regarding long‐term outcome (10 years) is scarce. According to the literature, pain and late onset neurological deficit may arise years after the spinal fracture. Chapter 5 focuses on the mid‐term (4 years) and long‐term (10 years) functional outcome in the same cohort of patients treated non‐operatively for a type A spinal fracture without primary neurological deficit. Functional outcome was measured using the VAS and the RMDQ. The 50 patients included were on average 41 years old at the time of injury. Four years post‐injury a mean VAS score of 74.5 and a mean RMDQ score of 4.9 were found; ten years after the accident, the mean VAS and RMDQ scores were 72.6 and 4.7, respectively (NS). The mean difference score of the VAS was 1.9 (S.D. 13.1), the mean difference score of the RMDQ was 0.2 (S.D. 4.1). No significant relationships were found between the difference scores of the VAS and RMDQ on the one hand, and age, gender, fracture sub‐classification and time between measurement moments on the other hand. Three (6%) patients had a poor long‐term outcome. None of the patients required surgery for late onset pain or progressive neurological deficit; furthermore, we found a status quo in functional outcome from 4 to 10 years post‐injury.

CHAPTER 8

122

In short, functional outcome after a non‐operatively treated type A spinal fracture without neurological deficit is good, both 4 years as well as 10 years post‐injury. Patients are only slightly disabled. For the group as a whole, 4 years after the fracture a steady state exists in functional outcome, which does not change systematically for at least 10 years after the fracture. A small number of patients have a poor outcome, though none of our patients required surgery for late onset pain or late onset neurological deficit.

The optimal treatment concerning the type A3 (Comprehensive Classification) “burst” fracture remains a challenging issue. Literature regarding short‐term functional outcome after operative and non‐operative treatment of these fractures shows conflicting results. Regarding long‐term outcome, data is hardly available. Some authors fear complications in the long term, like late onset pain and late onset neurological deficit. In Chapter 6 we studied the long‐term (5 years) functional outcome after operative and non‐operative treatment for this type of fracture. Functional outcome was measured by means of the VAS and the RMDQ. The 63 patients included (38 treated operatively, 25 treated non‐operatively) were on average 37 years old at the time of injury. The mean VAS scores in the operatively and non‐operatively treated groups were 82.6 and 80.8, respectively (NS). The mean RMDQ scores in the operatively and non‐operatively treated groups were 3.3 and 3.1, respectively (NS). Given these scores, both treatment modalities result in equal outcomes. None of the patients required surgery for late onset pain or late onset neurological deficit. In addition, we found that age and duration of follow‐up time did not correlate with outcome. Moreover, outcome was similar for type A3.1 and type A3.2 fractures. This raises the question whether such a detailed scheme as the Comprehensive Classification is required in daily practice. Literature regarding outcome after operative and non‐operative treatment for a type A3 fracture shows contradictory results, though there appears to be a trend towards equal outcomes after operative and non‐operative treatment. Most studies ,however, are retrospective, only two prospective studies are available; one shows superior outcome after operative treatment, the other finds equal outcomes. To summarize, we conclude that functional outcome after a type A3 spinal fracture is good, operative treatment results in similar outcome as non‐operative treatment in the long term. As such, benefits and drawbacks of both treatment modalities should be carefully taken into account when deciding which treatment is preferred for an individual patient. Furthermore, since outcome after both treatments is similar, other factors than the type of fracture should be taken into account when deciding which therapy should be chosen.

SUMMARY

123

In Chapter 7 a general discussion is presented. The results from the studies are considered, as well as some other aspects, like predictors of outcome, the ventral operative approach, the use of questionnaires and classification schemes. The most important conclusions are presented, which are:

• The SpinalMouse offers a quick and reliable method of measuring sagittal spinal ROM.

• The measurement of spinal ROM is not a suitable tool for evaluating disability. • A clear paucity exists regarding literature concerning physical capacity after a

spinal fracture, whereas we found a considerable number of patients to be restricted in their physical ability after non‐operative treatment of type A spinal fractures.

• The results of dynamic lifting tests seem to correlate well with subjective impairment.

• Operative treatment of type A spinal fractures (including the type A3 fracture) does not give superior functional outcome compared to non‐operative treatment.

• The majority of patients are slightly impaired after a spinal fracture without neurological deficit. This impairment does not influence their quality of life.

• Functional outcome is steady from about four years until ten years after non‐operative treatment of type A spinal fractures without neurological deficits.

• Functional outcome after a spinal fracture seems to be influenced by factors other than age, gender, treatment and fracture type classification.

Finally, some recommendations for future research are presented, focussing on a new classification scheme, physical capacity testing after a spinal fracture, the need for large study groups and the search for socio‐economic and psychosocial predictors of outcome.


125

Chapter 9

Nederlandse samenvatting

CHAPTER 9

126

In Hoofdstuk 1 wordt een algemene inleiding gegeven met betrekking tot wervelfracturen. De incidentie van traumatische thoracolumbale wervelfracturen zonder neurologische uitval is in Nederland ongeveer 1,2 per 10.000 personen per jaar (in de leeftijdsgroep van 20 tot 60 jaar). Wat betreft de kosten staat de behandeling van wervelfracturen in Nederland op de zevende plaats (3,8%) in de totale traumatologie‐gerelateerde kosten, met een gemiddelde van 6.600 euro per patiënt. We bespreken de ontwikkeling van classificatiemodellen, met speciale aandacht voor de Comprehensive Classification, welke gebruikt is in dit proefschrift. De thans meest gebruikte modellen, de classificatie volgens Denis en de Comprehensive Classification, hebben enkele nadelen: de betrouwbaarheid is matig en beide modellen erkennen de belangrijke rol van het posterieure ligamentaire complex (PLC) niet. Deze structuur speelt een belangrijke rol in de stabiliteit van de wervelkolom. Dientengevolge zijn nieuwe ontwikkelingen in opkomst, zoals het gebruik van de MRI om laesies van de PLC te ontdekken. De ThoracoLumbar Injury Classification and Severity Score (TLICS) is een nieuw classificatiemodel dat de status van de PLC wel in ogenschouw neemt. Verschillende aspecten van operatieve en conservatieve behandeling worden belicht. Voordelen van operatieve behandeling zijn de verbetering van de stand van de wervels, minder deformatie, vroegtijdige mobilisatie en revalidatie en soms de verbetering van neurologische functie. Een conservatieve behandeling heeft niet de gevaren van operatief ingrijpen, zoals wond infecties, iatrogeen neurologisch letsel of uitbreken van het osteosynthese materiaal. We beschrijven de totstandkoming van de operatieve behandeling door de tijd, evenals de verschillende technieken van operatieve behandeling (de dorsale en ventrale benadering en de recent ontwikkelde vertebroplastiek). Het begrip “functioneel resultaat” wordt gedefinieerd: het meten van de functionele situatie van een patiënt. Het gezondheidsmodel van de Wereld Gezondheids Organisatie (WHO), de International Classification of Functioning, Disability and Health (ICF), beschrijft de gezondheid van een persoon door beperkingen te meten in de volgende categorieën 1) lichaamsfunctie/structuur 2) activiteit en 3) participatie. Persoonlijke‐ en omgevings factoren beïnvloeden deze entiteiten. Verder worden in dit hoofdstuk verschillende instrumenten die functionele resultaten meten besproken, zoals lichamelijke conditie, terugkeer in het arbeidsproces en gezondheid‐gerelateerde kwaliteit van leven. Aan het eind presenteren we een literatuurbespreking. Interessante bevindingen uit deze bespreking zijn onder andere dat de uiteindelijke radiologische weergave

SAMENVATTING

127

van de wervel (zoals de voorste wighoek en de hoogte van het wervellichaam) niet correleert met het functionele resultaat. Tevens blijkt dat neurologische uitval een belangrijk negatief effect heeft op het functioneel resultaat. Ten slotte worden het doel en de inhoud van dit proefschrift beschreven.

Het gevolg van een wervelfractuur (en de behandeling hiervan) op de beweeglijkheid (Range of Motion: ROM) van de wervelkolom is niet duidelijk. Evenzo is de relatie tussen de ROM en het functionele resultaat niet helder. Voordat men deze vraagstukken kan oplossen moet echter eerst een manier gevonden worden om de ROM te meten. Als gevolg van de beperkingen van röntgenfoto’s voor onderzoeksdoeleinden (stralingsbelasting), zijn er vele, niet‐invasieve methoden voor het bepalen van de spinale ROM ontwikkeld. In Hoofdstuk 2 onderzochten we de SpinalMouse, een niet‐invasief, computer‐gestuurd apparaat om de sagittale spinale ROM te meten. De SpinalMouse wordt langs de rug gerold en meet de sagittale spinale ROM en de intersegmentele ROM (de ROM tussen de wervels). Een vrij unieke eigenschap van het apparaat is zijn vermogen om ook de thoracale ROM te meten. In de literatuur wordt doorgaans slechts de lumbale ROM gemeten. Om de inter‐beoordelaar betrouwbaarheid te evalueren werden 111 personen gemeten door 2 beoordelaars. Correlatie coëfficiënten waren r=0,90 voor de flexie, r=0,85 voor extensie en r=0,90 voor de totale inclinatie. De intra‐class correlatie coëfficiënten waren 0,95 voor flexie, 0,92 voor extensie en 0,95 voor totale inclinatie. De overeenkomst tussen uitbijters voor normale waarden van intersegmentele ROM was laag (Cohen’s kappa=0,22). Gezien deze cijfers lijkt de SpinalMouse een betrouwbaar instrument voor het meten van de ROM van de wervelkolom. Voor het bepalen van de intersegmentele ROM is het apparaat niet betrouwbaar. Gegeven het feit dat de belasting voor de patiënten laag is en een meting snel uitgevoerd kan worden, kan de SpinalMouse een nuttig middel zijn om de beweeglijkheid van de wervelkolom te meten, bijvoorbeeld tijdens de behandeling van rugklachten. Aan het eind van Hoofdstuk 2 doen we enkele aanbevelingen voor het gebruik in de praktijk.

In Hoofdstuk 3 onderzochten we de ROM na een wervelfractuur, alsmede de relatie tussen ROM en beperkingen. We bepaalden de thoracolumbale ROM in operatief en conservatief behandelde patiënten (n=76, gemiddelde follow‐up 3,7 jaar) en in een gezonde controlegroep (n=41). Om de beperkingen na een wervelfractuur vast te stellen en om de behandelingen te vergelijken werden twee ziekte‐specifieke vragenlijsten gebruikt: de VAS spine score (VAS) en de Roland‐

CHAPTER 9

128

Morris Disability Questionnaire (RMDQ). De scores werden bepaald in operatief en conservatief behandelde patiënten en in een gezonde controlegroep. Om het verband tussen ROM en subjectieve rugklachten na te gaan, berekenden we de correlatie tussen de thoracolumbale ROM en de VAS‐ en RMDQ scores. Operatief behandelde patiënten hadden een lagere ROM dan gezonde proefpersonen (respectievelijk 56,7° en 70,0°). Er was geen verschil tussen operatief en conservatief behandelde patiënten (respectievelijk 56,7° en 62,7°). Ook werd geen verschil gevonden tussen conservatief behandelde patiënten en de controlegroep. De correlatie tussen de ROM en subjectieve rugklachten was erg zwak en alleen significant voor ROM en RMDQ scores in de gehele studiegroep (rho=‐0,25, p<0,01). Patiënten ondervonden meer beperkingen dan gezonde personen, er was geen verschil tussen operatief en conservatief behandelde patiënten (VAS score 76,3 versus 72,6; RMDQ score respectievelijk 4,5 en 4,4). Literatuur met betrekking tot de ROM na een wervelfractuur laat tegenstrijdige resultaten zien. Sommige auteurs vinden een verminderde ROM terwijl anderen een normale ROM beschrijven. Ook is de samenhang tussen ROM en beperkingen niet duidelijk in de literatuur. Wij concluderen dat de sagittale thoracolumbale ROM 4 jaar na de operatieve behandeling van een wervelfractuur lager is dan die van gezonde personen. Waarom operatieve behandeling van wervelfracturen resulteert in een lagere ROM is niet duidelijk. Patiënten die een wervelfractuur oplopen ondervinden meer beperkingen dan gezonde mensen, de ROM lijkt echter geen invloed te hebben op deze beperkingen. Beide behandelmethoden (operatief en conservatief) resulteren in een vergelijkbaar beperkingenniveau.

De type A wervelfractuur (Comprehensive Classification) zonder neurologische uitval is de meest voorkomende wervelfractuur. Vaak worden deze fracturen conservatief behandeld. Hoofdstuk 4 beschrijft de functionele resultaten na conservatieve behandeling van type A thoracolumbale wervelfracturen zonder neurologische uitval. Het functionele resultaat werd bepaald volgens de International Classification of Functioning, Disability and Health (ICF). Dit model meet beperkingen in de categorieën lichaamsfunctie/structuur, activiteiten en participatie. Drieëndertig patiënten (gemiddelde leeftijd 50,5 jaar, gemiddelde follow‐up 5,3 jaar) werden geïncludeerd. Beperkingen in lichaamsfunctie/structuur werden gemeten door fysieke testen (dynamische tiltest en fiets ergometrie test), beperkingen in activiteiten werden gemeten door middel van vragenlijsten (VAS en RMDQ). Beperkingen in participatie werden bepaald middels de Short Form 36 vragenlijst (SF‐36) en de terugkeer in het arbeidsproces (return to work: RTW).

SAMENVATTING

129

Zevenendertig procent van de patiënten was niet in staat de dynamische tiltest binnen normaalwaarden te volbrengen. Deze tiltest bestond uit het herhaaldelijk optillen van een gewicht van de vloer naar een 75 cm hoge tafel. De test werd gestopt indien patiënten zelf aangaven vermoeid te raken of pijn te krijgen (psychofysisch testen). Aldus blijkt dat bij ongeveer een derde deel van de patiënten eerder klachten van vermoeidheid of pijn optreden tijdens tillen dan bij gezonde personen het geval is. In de ergometrie test (cardiopulmonale conditie) presteerde 41% onder de laagste normaalwaarde, 36% van de patiënten bereikte een hoge maximale zuurstofconsumptie (VO2‐max). De gemiddelde VAS score was 79, de gemiddelde RMDQ score 5,2. In de SF‐36 scores werd geen verschil gevonden tussen patiënten en gezonde personen. De dynamische tiltest correleerde vrij sterk met de RMDQ, VAS en SF‐36 fysieke subschaal (respectievelijk r=‐0,62, r=0,71 en r=0,59). Tien procent van de patiënten was gestopt met werken en ontving een uitkering, 24% had aanpassingen in zijn werk en 14% was van baan veranderd. Andere auteurs vinden vergelijkbare cijfers. Hierbij moet opgemerkt worden dat verschil in nationale regelgeving over sociale zekerheid deze cijfers beïnvloedt. Er is een gebrek aan gegevens over de fysieke conditie na een wervelfractuur. Het onderzoeken van de lichamelijke conditie is vrij zeldzaam in wervelfractuur‐onderzoek, terwijl we hebben aangetoond dat de lichamelijke conditie een goede indruk geeft van het functionele resultaat. Concluderend lijken patiënten het redelijk goed te doen 5 jaar na conservatieve behandeling van een type A wervelfractuur, alhoewel de resultaten uiteenlopend zijn in de verschillende categorieën van de ICF. De lichamelijke conditie lijkt beperkt in een aanzienlijk aantal patiënten. De beperkingen die patiënten ervaren in de categorie activiteiten zijn echter niet van invloed op de participatie.

De lengte van de follow‐up in onderzoek naar de functionele gevolgen voor patiënten met een wervelfractuur is meestal kort. De meeste studies beschrijven een termijn van 1 tot 2 jaar. Literatuur over lange‐termijn resultaten (10 jaar of langer) is schaars. Volgens de literatuur kunnen jaren na de fractuur nog pijn en neurologische uitval ontstaan. Hoofdstuk 5 richt zich op de middellange (4 jaar) en lange‐termijn (10 jaar) functionele resultaten van conservatief behandelde patiënten met een type A wervelfractuur zonder initiële neurologische uitval. De functionele resultaten werden gemeten met behulp van de VAS en RMDQ vragenlijsten. De 50 deelnemende patiënten waren gemiddeld 41 jaar oud ten tijde van het trauma. Vier jaar na het ongeval was de gemiddelde VAS score 74,5 en de gemiddelde RMDQ score 4,9. Tien jaar na het ongeval waren de gemiddelde VAS‐

CHAPTER 9

130

en RMDQ scores respectievelijk 72,6 en 4,7 (NS). De gemiddelde verschilscore van de VAS bedroeg 1,9 (S.D. 13,1), voor de RMDQ bedroeg deze 0,2 (S.D. 4,1). Er werden geen significante correlaties gevonden tussen de verschilscores van de VAS en RMDQ enerzijds, en leeftijd, geslacht, fractuurclassificatie en de tijd tussen de metingen anderzijds. Drie patiënten (6%) hadden een slecht lange‐termijn resultaat. Geen enkele patiënt had een indicatie voor chirurgisch ingrijpen wegens laat optredende pijn of neurologische uitval. Voorts vonden we een status‐quo in het functionele resultaat vanaf 4 tot 10 jaar na het ongeval. Samenvattend kan gesteld worden dat het functionele resultaat na conservatieve behandeling van een type A wervelfractuur goed is, zowel 4 als 10 jaar na het trauma. Patiënten ervaren slechts lichte beperkingen. Voor de groep als geheel bestaat er 4 jaar na het ongeval een constant niveau van het functionele resultaat, dat niet systematisch verandert tot tenminste 10 jaar na het ongeval. Een klein aantal patiënten heeft een slechte uitkomst, hoewel geen van onze patiënten een operatie nodig had voor laat optredende pijn of neurologische uitval.

De optimale behandeling voor de type A3 (Comprehensive Classification) “burst” fractuur blijft een uitdaging. Literatuur met betrekking tot de korte‐termijn resultaten na operatieve en conservatieve behandeling van deze fractuur laat tegenstrijdige resultaten zien. Wat betreft de lange‐termijn resultaten is er nauwelijks literatuur voorhanden. Sommige auteurs vrezen complicaties op de lange termijn, zoals laat optredende pijn of neurologische uitval. In Hoofdstuk 6 bekeken we de functionele resultaten na operatieve en conservatieve behandeling van dit type fractuur op middellange termijn (5 jaar). Het functionele resultaat werd gemeten middels de VAS en RMDQ vragenlijsten. Drieënzestig patiënten namen deel aan de studie (38 operatief behandelden, 25 conservatief behandelden). Zij waren gemiddeld 37 jaar oud ten tijde van het ongeval. De gemiddelde VAS scores in de operatief en conservatief behandelde groep waren respectievelijk 82,6 en 80,8 (NS). De gemiddelde RMDQ scores in de operatief en conservatief behandelde groep waren respectievelijk 3,3 en 3,1 (NS). Beide behandelingen resulteren daarmee in een vergelijkbaar functioneel resultaat. Voor geen enkele patiënt was operatief ingrijpen wegens laat optredende pijn of neurologische uitval noodzakelijk. Verder bleek dat leeftijd en de duur van de follow‐up niet samenhingen met het functioneel resultaat. Het functionele resultaat was gelijk voor de type A3.1 en A3.2 fracturen. Dit werpt de vraag op of een zo gedetailleerd schema als de Comprehensive Classification nodig is in de dagelijkse praktijk. De literatuur over functionele resultaten na operatieve en conservatieve

SAMENVATTING

131

behandeling van type A3 fracturen laat strijdige resultaten zien, hoewel een trend lijkt te bestaan die wijst naar vergelijkbare resultaten na operatief en conservatief behandelen. Echter, de meeste onderzoeken zijn retrospectief van aard, er zijn slechts 2 prospectieve studies verschenen: één laat betere resultaten zien na operatieve behandeling, de ander vindt vergelijkbare resultaten. Resumerend concluderen we dat het functioneel resultaat na een type A3 wervelfractuur goed is. Op de middellange termijn resulteert operatieve behandeling in vergelijkbare resultaten als conservatieve behandeling. Aldus moeten de voor‐ en nadelen van beide behandelingen goed overwogen worden bij het besluit welke behandeling uitgevoerd gaat worden. Ook andere factoren dan het type fractuur dienen meegewogen te worden in de besluitvorming.

Hoofdstuk 7 behandelt de algemene discussie. De resultaten van de studies worden besproken, samen met enkele andere aspecten, zoals voorspellers van het functioneel resultaat, de ventrale en dorsale operatieve benadering, het gebruik van vragenlijsten en classificatieschema’s. De belangrijkste conclusies die getrokken kunnen worden zijn:

• De SpinalMouse kan gebruikt worden om snel en betrouwbaar de sagittale ROM van de wervelkolom te meten.

• Het meten van de beweeglijkheid van de wervelkolom is geen geschikt instrument om beperkingen te meten.

• Er is een duidelijk gebrek aan literatuur over de fysieke conditie na een wervelfractuur, terwijl wij bij een aanzienlijk aantal patiënten beperkingen vonden in de fysieke conditie na conservatieve behandeling van een type A wervelfractuur.

• Resultaten van de dynamische tiltest correleren goed met subjectieve beperkingen.

• Operatieve behandeling van een type A wervelfractuur (inclusief de type A3 fractuur) geeft een vergelijkbaar functioneel resultaat als conservatieve behandeling.

• De meerderheid van patiënten is licht beperkt na een wervelfractuur zonder neurologische uitval. Deze beperking beïnvloedt de kwaliteit van leven niet.

• Het functionele resultaat bevindt zich op een constant niveau vanaf 4 jaar tot 10 jaar na een conservatief behandelde type A wervelfractuur zonder neurologische uitval.

• Het functionele resultaat na een wervelfractuur wordt beïnvloed door andere factoren dan leeftijd, geslacht, behandeling en fractuurclassificatie.

CHAPTER 9

132

Ten slotte worden aan het eind van dit hoofdstuk enkele aanbevelingen gedaan voor verder onderzoek, met de nadruk op een nieuwe classificatie, fysieke conditie‐ metingen, de noodzaak van grote studiegroepen en het mogelijke belang van sociaal‐economische en psychosociale factoren als voorspellers van het functioneel resultaat.



LIST OF ABBREVIATIONS

135

List of abbreviations AO Arbeitsgemeinschaft für Osteosynthesefragen BMI Body Mass Index bpm Beats Per Minute CC Comprehensive Classification COPD Chronic Obstructive Pulmonary Disease CT Computer Tomography deg degrees FEV1 Forced Expiratory Volume in 1 second fig figure GLBOS Greenough Low Back Outcome Scale ICC Intra‐class Correlation Coefficient ICF International Classification of Functioning, Disability and Health ICIDH International Classification of Impairments, Disabilities and Handicaps kg kilogram LBM Lean Body Mass LD Loading Degree LED Light Emitting Diode LSC Load Sharing Classification m meter MRI Magnetic Resonance Imaging MVAS Million Visual Analogue Scale NHP Nottingham Health Profile NIOSH National Institute for Occupational Safety and Health NS not significant ODI Oswestry Disability Index PC Personal Computer PILE Progressive Isoinertial Lifting Evaluation PLC Posterior Ligamentous Complex RMDQ Roland‐Morris Disability Questionnaire ROM Range Of Motion RTW Return to Work S.D. Standard Deviation SF‐36 Short Form 36 item General Health Survey SIP Sickness Impact Profile

LIST OF ABBREVIATIONS

136

SPSS Standard Package for the Social Sciences TLICS ThoracoLumbar Injury Classification and Severity Score TLISS ThoracoLumbar Injury Severity Score TLSO thoracolumbosacral orthosis UHG University Hospital Groningen UMCG University Medical Centre Groningen USS Universal Spine System VAS Visual Analogue Scale Spine Score VC Vital Capacity VO2–max maximum oxygen uptake WDI Waddell Disability Index WHO World Health Organization

DANKWOORD

137

Dankwoord Ten eerste wil ik mijn promotor bedanken, prof. dr. H.J. ten Duis. Dankzij uw voortreffelijke hulp en kritische vragen ligt dit proefschrift er nu. Ik wil u hartelijk bedanken voor uw begeleiding en het in mij gestelde vertrouwen. Uw op‐ en aanmerkingen waren zeer waardevol in de verfijning van het manuscript. Mijn dank hiervoor.

Mw. dr. C.K. van der Sluis, co‐promotor, revalidatie‐arts. Beste Corry, bijzonder hoe wegen elkaar kunnen kruisen. In 1999 (?) heb jij mij de eerste beginselen van wetenschappelijk onderzoek bijgebracht. In 2003 ben ik via Vincent weer bij jou terecht gekomen, hebben we mijn scriptie (inderdaad, uit 1999) gepubliceerd, het wervel onderzoek verder uitgebouwd en uiteindelijk heeft een en ander geresulteerd in dit proefschrift. Hartelijk dank voor je tijd, interesse, advies en de zorgvuldige correcties. Zonder jouw adequate hulp en snelle respons op de manuscripten was het niet gelukt. Jij was de beste co‐promotor die ik me kon wensen.

Dr. V.J.M. Leferink, co‐promotor, chirurg/traumatoloog. Beste Vincent, in 2002 zijn we begonnen met de SpinalMouse, omdat ik nog 4 studie‐punten moest halen (dat is equivalent aan 4 studie weken....). Vier weken werd 4 jaar (en nog een beetje). Jij hebt me vanaf het begin enorm geholpen met de basisvaardigheden van het respectievelijk onderzoeken, schrijven en publiceren. Toen ik begon met schrijven was jouw proefschrift bijna klaar, nooit gedacht dat dit zou volgen. Enorm bedankt voor je advies, interesse, enthousiaste hulp en je immer aanwezige bereidheid tot “ruggespraak”.

Mw. drs. H.J.E. Keizer, revalidatie‐arts en dr. P.U. Dijkstra, statisticus en methodoloog. Beste mede‐auteurs, mijn welgemeende dank voor jullie bijdrage. Evelien, ik denk dat we een mooi artikel geschreven hebben, een prachtig stuk werk van jouw kant. Beste Pieter, er zijn grove leugens, kleine leugens, en statistiek. Ik ben je zeer erkentelijk voor je tijd en hulp als de cijfertjes mij boven het hoofd gingen. Een mens heeft nooit genoeg boeken, en het geheim van de boxplot is voorgoed opgelost.

De hooggeleerde heren prof. dr. S.K. Bulstra, prof. dr. J.H.B. Geertzen en prof. dr. P. Patka, leden van de beoordelingscommissie. Ik wil u allen danken dat u zitting heeft willen nemen in de beoordelingscommissie.

DANKWOORD

138

Drs. R.J. Sol en drs. E.M. Post, paranimfen. Jeroen en Es, ik ben vereerd dat jullie deze zware taak op jullie wilden nemen. Aan de ene zijde een audioloog voor de input, aan de andere zijde een communicatie‐deskundige voor de output. Wat kan er dan nog mis gaan?

Drs. A. Vogel, drs. J. Oldenziel, drs. Y.K. Sze en drs. H.J.J. Glastra‐van Loon. Astrid, Job, Yuk‐Kueng en Hessel, geachte collegae: hartelijk dank voor jullie gewaardeerde hulp met tekst, stickers, plaatjes en lay‐out.

Chirurgen en orthopaedisch chirurgen uit het Scheper Ziekenhuis te Emmen. Mijn dank voor de geboden mogelijkheden tot “UMCG ochtendjes”, “type middagjes” en een “promo weekje” op zijn tijd. Mijn dank voor jullie interesse naar de vorderingen van boek en loopbaan. Michiel, ook deze kan aan de muur gespijkerd worden! Collega‐assistenten, ik ben jullie zeer erkentelijk voor het overnemen van diensten als er weer getypt moest worden. Secretaresses, mijn dank voor jullie hulp en het feit dat ik jullie printer steeds weer mocht leegprinten. Gea, hartelijk dank voor je hulp en interesse.

Uiteraard wil ik alle patiënten bedanken die hebben deelgenomen aan de uitgebreide tests en steeds maar weer de vragenlijsten trouw terugstuurden. Mijn dank hiervoor.

Tenslotte alle mensen die mij op één of andere wijze geholpen hebben met dit proefschrift. Jaap Rinzema en Léonie Roberti‐de Kam wil ik graag bedanken voor de hulp bij het corrigeren van de Engelse teksten. Dekkers, dank voor je hulp bij de afronding. Herman Blauwgeers wil ik danken voor de hulp bij het verkrijgen van landelijke cijfers. Familie en vrienden, Bert, Ineke, Ria, Erik en Fleur, Esther en Mark, mijn dank voor jullie interesse en motiverende woorden. Is het toch maar weer gelukt! Vrienden en vriendinnen, jullie allen bedankt! Ik zal nooit meer zeuren over mijn boekje, beloofd.

Lieve Nicole, eindelijk klaar met promo‐typen! Tijd voor leuke(re) dingen! From London to Paris, nu dan eindelijk tijd voor jouw home‐town: NYC here we come....

CURRICULUM VITAE

139

Curriculum Vitae The author of this thesis was born in 1977 in Westerbork, The Netherlands. After graduating from the Atheneum in 1995 he started his medical study at the University of Groningen. During his study he started working on scientific research with Corry van der Sluis, consultant in Rehabilitation Medicine, and Vincent Leferink, surgeon. After graduating from university at the end of 2002 he is since 2003 resident surgery in the Scheperziekenhuis, Emmen (head dr. M. van den Berg) working on further career in the field of surgery. In this period he started working on this thesis. In his spare time the author loves to sail on his catamaran, as well as playing golf and playing drums in a bigband. Address for correspondence: Richard B. Post [email protected]


Documents

FUNCTIONAL OUTCOME AFTER A SPINAL FRACTURE · 3.7% of total health care costs. Spinal fractures (including spinal cord injury) rank 7th (3.8%) in total trauma costs, with a mean cost