Prehabilitation and early rehabilitation after spinal surgery: randomized clinical trial

http://cre.sagepub.com/Clinical Rehabilitation

http://cre.sagepub.com/content/24/2/137The online version of this article can be found at:

DOI: 10.1177/0269215509347432

2010 24: 137Clin RehabilPer Rotbøll Nielsen, Lars Damkjær Jørgensen, Benny Dahl, Tom Pedersen and Hanne TønnesenPrehabilitation and early rehabilitation after spinal surgery: randomized clinical trial

Published by:

http://www.sagepublications.com

can be found at:Clinical RehabilitationAdditional services and information for

http://cre.sagepub.com/cgi/alertsEmail Alerts:

http://cre.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://cre.sagepub.com/content/24/2/137.refs.htmlCitations:

What is This?

- Jan 26, 2010Version of Record >>

at Copenhagen University Library on March 11, 2013cre.sagepub.comDownloaded from

http://cre.sagepub.com/

http://cre.sagepub.com/content/24/2/137

http://www.sagepublications.com

http://cre.sagepub.com/cgi/alerts

http://cre.sagepub.com/subscriptions

http://www.sagepub.com/journalsReprints.nav

http://www.sagepub.com/journalsPermissions.nav

http://cre.sagepub.com/content/24/2/137.refs.html

http://cre.sagepub.com/content/24/2/137.full.pdf

http://online.sagepub.com/site/sphelp/vorhelp.xhtml


Clinical Rehabilitation 2010; 24: 137–148

Prehabilitation and early rehabilitation after spinalsurgery: randomized clinical trialPer Rotbøll Nielsen Anaesthesiology Department, Centre of Head and Orthopaedics, Rigshospitalet and WHO CollaboratingCentre for Health Promotion in Hospitals & Health Services, Bispebjerg Hospital, Lars Damkjær Jørgensen, Benny DahlOrthopaedic Clinic, Centre of Head and Orthopaedics, Rigshospitalet. University of Copenhagen, Tom Pedersen CentreDirector, Rigshospitalet. University of Copenhagen and Hanne Tønnesen WHO Collaborating Centre, Bispebjerg Hospital,Copenhagen, Denmark

Received 23rd December 2008; returned for revisions 14th April 2009; revised manuscript accepted 28th July 2009.

Objective: To evaluate the outcome after spinal surgery when adding

prehabilitation to the early rehabilitation.

Design: A randomized clinical study.

Setting: Orthopaedic surgery department.

Subject: Sixty patients scheduled for surgery followed by inpatient rehabilitation

for degenerative lumbar disease.

Interventions: The patients were computer randomized to prehabilitation and early

rehabilitation (28 patients) or to standard care exclusively (32 patients). The inter-

vention began two months prior to the operation. The prehabilitation included an

intensive exercise programme and optimization of the analgesic treatment. Protein

drinks were given the day before surgery. The early postoperative rehabilitation

included balanced pain therapy with self-administered epidural analgesia, doubled

intensified mobilization and protein supplements.

Main measures: The outcome measurements were postoperative stay,

complications, functionality, pain and satisfaction.

Results: At operation the intervention group had improved function, assessed by

Roland Morris Questionnaire (P¼ 0.001). After surgery the intervention group

reached the recovery milestones faster than the control group (1–6 days versus

3–13, P¼ 0.001), and left hospital earlier (5 (3–9) versus 7 (5–15) days, P¼ 0.007).

There was no difference in postoperative complications, adverse events, low

back pain and radiating pain, timed up and go, sit-to-stand or in life quality.

Patient satisfaction was significantly higher in the intervention group compared

with the control group.

Conclusion: The integrated programme of prehabilitation and early

rehabilitation improved the outcome and shortened the hospital stay – without

more complications, pain or dissatisfaction.

Introduction

An increasing number of patients undergo spinalsurgery. Between 1999 and 2001 the frequency incre-ased by more than 200% in the USA, to 61 opera-tions per 100 000 inhabitants.1 In Denmark the

Address for correspondence: Per Rotbøll Nielsen, TheNeuroscience Centre, Department of Neuroanaesthesiology,Multidisciplinary Pain Centre, Rigshospitalet, University ofCopenhagen, Blegdamsvej 9, DK-2100 Copenhagen Ø,Denmark. e-mail: [email protected]

� The Author(s), 2010.Reprints and permissions: http://www.sagepub.co.uk/journalsPermissions.nav 10.1177/0269215509347432



number increased from 26 to 46 per 100 000between 1997 and 2003. A review of the literatureshows that elective spinal surgery for degenerativedisease is still related to a significant hospital stay,second surgery rate2,3 and risk of developing post-operative complications.4 The development ofbetter patient clinical pathways with improvedoutcome in this field is therefore attractive.In recent years early rehabilitation including

fast-track surgery has been introduced to over-come complications, fatigue and delayed convales-cence after several surgical interventions.5,6 Earlyrehabilitation involves coordinated effort, combin-ing modern concepts of patient education withnewer anaesthetic and analgesic techniques andminimally invasive surgical techniques; the inten-sion is to reduce the stress response, and to mini-mize pain and discomfort.7

In addition, prehabilitation is effective in redu-cing the risk of postoperative complications andhospital stay. Prehabilitation is defined as aug-menting functional capacity prior to surgery.8

Evidence-based programmes have already beendeveloped regarding smoking and alcohol cessa-tion intervention, nutrition among others.9–11

However, to our knowledge an integrated perio-perative programme including both prehabilitationand early rehabilitation has not been evaluated pre-viously. We hypothesized that an integrated pro-gramme for elective spine surgery would result inwell-informed patients prepared to play a veryactive role in the pre- and postoperative period.After the operation patients would experience ashorter time with disability.The aim of this study was to evaluate the effect

of an integrated programme combining prehabili-tation and early rehabilitation on outcome afterelective spinal surgery for degenerative disease,compared with the clinic’s routine procedures.

Methods

In the 18-month trial period from 2003 to 2005, 88consecutive patients were referred and scheduledfor surgery at the list for elective lumbar spinalsurgery for degenerative disease with low backpain and radiating pain. Routinely, thesepatients had their final indication for surgery

established by the surgeon shortly before the oper-ation. Due to the eight weeks prehabilitation pro-gramme the patients needed to be included in duetime before surgery, and sometimes before thefinal indication was established. Therefore, it wasaccepted to include some patients, who in the enddid not need an operation. The surgeons made thedecision of surgery or no surgery without knowl-edge of the allocation.

The routine included inpatient rehabilitation.Inclusion criteria were age over 18 years, beingable to understand and actively participate in theintegrated programme. Fifteen patients did notfulfil these criteria, since they underwent early sur-gery. Therefore, 73 patients were included and ran-domized after informed consent 6–8 weeks beforescheduled operation. The exclusion criteria werecontraindications to surgery in general as well asspecific criteria to this study. They include contra-indications to the use of epidural catheter, allergyto paracetamol, opiates or analgesics for regionalanaesthesia, liver disease, nephropathy or pregnancy.In six of the 73 patients there was no indication ofsurgery at the time of the planned procedure; one ofthese six patients cancelled the operation becausethe pain resolved. Three other patients underwentsurgery at another hospital, and 4 withdrew theirinformed consent (Figure 1).

The allocation was based on computer random-ization in blocks of 10 patients, to either integratedprogramme or standard programme. Informationon intervention or routine procedure was enclosedin sealed opaque envelopes with consecutive num-bers. The randomization was performed by ahealth professional, which did not otherwise takepart in the trial. The randomized study designincluded several pragmatic aspects that relate toeffectiveness, rather than explanatory (efficacy)trials,12,13 such as measuring health outcomes(functionality and quality of life), having a rela-tively long study duration, testing clinically rele-vant treatment modalities, assessing adverse eventsand performing intention to treat analyses.

AssessmentThe Brief Pain Inventory Questionnaire was desi-

gned to measure the subjective intensity of pain andthe impairment caused by pain. It has been

138 PR Nielsen et al.



validated in several languages, including Scandina-vian languages.14 The Brief Pain Inventory Ques-tionnaire was simple for most patients to completeand sensitive to the changes in pain. The score isnormally separated into three pain qualities: painrelated to tissue damage – in this case low backpain; pain related to abdomen; and radiating pain –in this case leg pain. In our study we scoredthe patients for low back pain and radiating pain.The physiotherapist and a nurse especially trained

in pain score, intervention and follow-up per-formed all assessments. They were not member ofthe clinical staff and did not participate in the dailypatient care. They were not blinded, however,because we could not guarantee that the patientswould not disclose their allocation group, and itwould be impossible to blind the patients.

The Roland Morris Questionnaire was designedto be completed by patients to assess physical dis-ability due to low back pain. It contained 24

Assessed for eligibility (n=88)

Total number of patient registered N=73

Excluded (n=15)

15 underwent early surgery

0 refused to participate

Allocated to intervention (n=35)

Received allocated intervention (n=28)

Did not receive allocated intervention(n=7; of whom 1 was operatedelsewhere, 2 were cancelled and 4withdrew informed consent)

Allocated to control (n=38)

Received allocated intervention (n=32)

Did not receive allocated intervention(n=6; of whom 2 were operated elsewhere and 4 cancelled)

Lost to follow-up (n=0)

Discontinued intervention (n=0)

Lost to follow-up (n=0)

Discontinued intervention (n=0)

Analysed (n=28)

Excluded from analysis (n=0)

Analysed (n=32)

Excluded from analysis (n=0)

Figure 1 Flowchart.

Prehabilitation and early rehabilitation after spinal surgery 139



unweighted items, ranging from 0 (no disability) to24 (maximum disability). The Roland MorrisQuestionnaire has been translated and validatedin several languages, including Danish.15 It has agood correlation to other disability instruments.16

The sit-to-stand test is a simple and reliablephysical performance test17 measuring the timerequired for the patient to rise from sitting in achair to standing and return to sitting as quicklyas possible five times. The time was measured witha stopwatch.Timed up and go is a quick and practical test of

basic mobility in elderly patients.17 The patientrose from a chair, stood, walked forward to aline 3m away, turned, walked back to the chair,and sat down. Patients were timed as they per-formed this test.The 15D test is a generic, comprehensive,

15-dimensional, standardized, self-administeredmeasure of health-related quality of life(HRQoL) that can be used both as a profile andsingle-index score measure. The 15D scores areshown to be highly reliable, sensitive and respon-sive to change and generalizable, at least inWestern-type societies.18

The milestones the patient can achieve underhospitalization are defined as: assisted positionalchange in bed, independent positional change inbed, assisted mobilization in bedside, independentmobilization to bedside, assisted mobilization inwalking frame, independent mobilization in walk-ing frame, independent personal hygiene, indepen-dent daily function on ward, walking without aids,complete training programme, independent stairclimb and discharge (Figure 2).

Follow-upPain was measured with a modified Brief Inven-

tory Questionnaire. Functionality was assessed byRoland Morris Questionnaire, sit-to-stand andtimed up and go. Furthermore, the days the clin-ical milestones in the recovery process werereached was recorded (Figure 2).

Postoperative complications were diagnosed bythe surgeons and only registered if they requiredtreatment. The surgeons also decided on discharge.The surgeons were unaware of the allocation,which was not noted in the medical records.

0

1

2

3

4

5

6

7

8

Assistance

∗∗

∗ ∗

∗

∗ ∗

∗∗

∗

∗

In bedBedside

Walking frameDaily funct

WalkTrain

StairsDischarge

Figure 2 Milestones on ward. *Significant.




Quality of life was measured traditionally withself-reported 15D.19 Satisfaction was assessedusing a 5-point scale: very satisfied¼ 5, satisfied,neither satisfied nor dissatisfied, dissatisfied, verydissatisfied¼ 1.

After registration and inclusion, all patientswere assessed for pain and functionality on theday before surgery, on the first, third and fifthpostoperative days, at discharge, and again after1, 3 and 6 months. Quality of life was measured atthe same points.

Postoperative complications, adverse events,patient satisfaction were assessed at the follow-up one month after operation.

Preoperative programmesIn general, treatment at hospitals is free of charge

in Denmark, and so were the programmes in thistrial.

Routine programme for the control groupThe control group followed departmental rou-

tines. Once the surgeon took the decision on sur-gery, the surgeon informed the patient about theoperation and advised regarding cessation of smok-ing and harmful drinking before surgery. At admis-sion the anaesthesiologist informed the patientabout anaesthesia and postoperative pain treat-ment and the nurse in the surgical departmentinformed the patient about the plan for postopera-tive rehabilitation in hospital, including pain treat-ment, diet and physiotherapy following theoperation day, aiming for discharge on the eighthpostoperative day. The physiotherapist mobilizedthe patients if possible on the day of the operationand trained daily 30minutes with the patients in thefollowing days. There was no reporting on furtherself-rehabilitation for control patients.

Prehabilitation programme for the interventiongroup

All intervention patients received a 6- to 8-week,individualized, preoperative training programmefor exercise at home. It consisted of a daily 30-minute programme and the patients monitoredtheir training using a logbook. The programme

focused on improvement of muscle strength forthe back and abdomen and included cardiovascu-lar conditioning. A physiotherapist supervised thelearning process at the day of inclusion and againtwo weeks before the operation. The complianceof the intervention group was measured for everytype of exercise by daily self-reporting in alogbook.

The patients were informed and advised regard-ing cessation of smoking and harmful drinkingbefore surgery.

Fourteen days before the operation the inter-vention group, eventually with relatives, met withthe physiotherapist, and once again received infor-mation about the operation, postoperative mobi-lization and rehabilitation.

The evening before surgery, the interventiongroup supplemented their usual food intake with200mL protein-rich drinks (Fortimel; NutriciaLtd, Trowbridge, Wilts, UK).

Intraoperative procedures were similar for bothgroups

The anaesthesiological and surgical procedureswere similar for the two groups. All patients hadprophylactic antibiotics (cefuroxime 1.5 g intrave-nously) and thromboembolic prophylaxis(Innohep tinzaparin natrium 20 000 anti-XaIE/mL, 3500 IE subcutaneous) before the opera-tion began.

The patients received propofol, remifentanil andcisatracurium anaesthesia; they underwent intuba-tion and surgery in the prone position. An epi-dural catheter was placed and 20 microgramsufentanil was administered epidurally beforeextubation. A bolus of 16mg ropivacaine and 8microgram sufentanil was installed using a pump(Baxter), when the patient was awake and couldcontrol his or her lower extremities again.

The surgical procedures did not include morethan two levels; 80% of the patients underwentprimary non-instrumented spondylodesis dueto degenerative disc disease with or without legpain and 50% of these procedures were performedin combination with decompression usingautologous bone graft for posterolateral fusion.A small number of patients underwent instru-mented fusion using the same technique. The deci-sion to use pedicle screws was taken by thesurgeon during the operation. A small numberof patients underwent secondary surgery because




of pseudarthrosis. One case of disc prosthesis atthe L5–S1 level was included in each group.No patients undergoing other procedures, such

as discectomy, anterior lumbar interbody fusion orposterior lumbar interbody fusion, were included.The patients were included when they were

booked for surgery 6–8 weeks before the operation.Because at that time the procedure was not alwaysdecided, different kinds of lumbar spinal surgerywere included. The patients were randomlyassigned after the first assessment. Senior surgeonsperformed all the procedures. There was no differ-ence between the groups regarding surgery, anaes-thetics, fluid therapy or blood transfusion.

Postoperative pain treatment and rehabilitationAll the patients had 1000mg paracetamol four

times a day. After three days the epidural catheterwas removed and the treatment was changed tooxycodone CR tablets 20mg, twice a day until dis-charge. There was no difference between the twogroups in the use of analgesics.The nurses actively mobilized all the patients in

both groups for dressing, meals and toileting, etc.

The control groupThe patients in the control group received rou-

tine pain treatment consisting of epidural infusionof 12mg ropivacaine and 6 micrograms sufentanilper hour and supplementary 10mg ropivacaine atbreak-through pain.The control group was mobilized once daily and

followed the routine rehabilitation for inpatients,aiming to discharge after eight days.

The intervention groupThe intervention group had balanced and

patient-controlled epidural analgesia, which con-sisted of a basic dose (8mg ropivacaine and 4microgram sufentanil per hour) and a bolus infu-sion to supplement (6mg ropivacaine and 3 micro-gram sufentanil up to three times every hour forbreak-through pain).The intervention group had an additional reha-

bilitation programme, which aimed to dischargeon the fifth postoperative day. The physiotherapist

intensively mobilized the patient on the day of theoperation and 30 minutes twice daily in the follow-ing days. The postoperative exercise was similar tothe preoperative exercises.

The intervention group began enteral nutritionin the evening on the day of surgery, (150mLFortimel), and continued to drink 4� 150mLa day in addition to the ordinary hospitaldiet, while the control group received the ordinarydiet only.

Analyses and statisticsWe planned to evaluate a minimal relevant dif-

ference of two to three days of hospital stay. Thepostoperative stay for routine procedures in thedepartment was eight days (including inpatientrehabilitation), based on patient records over thepast three years, whereas the hospital stay forpatients in the intervention group was expected tobe five days. Calculating a 5% risk of a type 1 errorand a 20% risk of a type 2 error, we concluded thatabout 30 patients should be included in each group.

The groups were compared using non-pairedtwo-sided tests: Mann–Whitney test and Fisher’sexact test. The statistician was blinded. The resultswere analysed according to intention to treat,except for the postoperative complications andstay, because of the high risk for underestimatedresults due to the inclusion of patients who werenot operated on in those analyses, thereby present-ing a lower frequency of complications and a lowerstay in hospital. Area under curve was used foranalyses of measurements over time in order toreduce the number of tests. The level of significancewas 0.05. The subgroup analysis was performed asplanned by using multivariate analyses (SPSS forWindows; SPSS Inc., Chicago, IL, USA).

Ethical considerationsThe study was performed in accordance with the

Helsinki II Declaration, and all patients partici-pated after informed consent. The RegionalScientific Ethical Committee 01-041/03 approvedthe protocol, and the study has been registered inthe international protocol registration systemwww.ClinicalTrails.gov, ID NCT 00459966.




Results

A total of 38 patients entered the interventiongroup and 35 entered the control group; 28 and32 patients fulfilled the study in the two groups,respectively, thereby giving a drop-out rate of19%. All 13 patients but one stopped within thefirst week. Two months before surgery there wereno differences between the groups regarding thecharacteristics (Table 1). At the time of operation,however, the intervention group had increasedfunction assessed by Roland Morris (P¼ 0.001),although there were no differences in timed upand go or sit-to-stand between groups (Table 2).Exercise compliance in the intervention group washigh; the patients exercised 85% of possible days.

After surgery, the intervention group reached themilestones in the recovery process significantlyfaster than the control group (1–6 days versus 3–13days, P¼ 0.001), and they left the hospital signifi-cantly earlier (median 5 (3–9) days versus 7 (5–15)days, P¼ 0.007, Figure 2). Complications andadverse events tended to be followed by a lengthier

hospital stay in the control group (30 days versus 18days, P¼ 0.06, Table 3). The 15 patients (10 womenand 5 men, median age 45 years, range 24–66) fromthe non-included group underwent surgery andstayed a median of 6 days (range 4–13). The further4 patients (2 women and 2 men, age 67 years, range38–78) who withdrew their informed consent stayedfor 8 days (range 5–13). These numbers were toosmall for statistical analysis.

The intervention group experienced significantlyless pain, measured as area under curve (P¼ 0.03),and less low back pain intensity (area under curve,P¼ 0.02). There were no differences in average andworst low back pain or all kinds of radiating painsbetween groups. Patients who had major complica-tions or stayed longer at hospital did not experiencedmore pain compared with patients without majorcomplications or who stayed for a shorter time.

Significantly more patients in the interventiongroup were very satisfied with the overall treatmentand outcome compared with the control group (15of 28 patients versus 7 of 32, P¼ 0.02). The qualityof life was similar for both groups.

Table 1 Characteristics of all patients at inclusion time; number (per cent) or median (range)

Interventionn¼ 35

Controln¼ 38

P-value

Women 21 (61%) 22 (59%) 1.0Age (years) 48 (31–80) 52 (23–88) 0.92Weight (kg) 79 (48–118) 74 (51–107) 0.59BMI (kg/m2) 25 (21–33) 26 (17–33) 0.49Daily smokers 3 (11%) 5 (16%) 0.72Harmful drinkersa 0 0 –ASA Classification (I–V)b I (I–II) I (I–II) 0.18Comorbidity (clinical diseases requiring treatment) 4 (14%) 1 (3%) 0.18

Cardiovascular disease 2 0Pulmonary disease 1 1Diabetes (type 1/type 2) 1 0

Functionality: RMQ (0–23)c 16 (8–21) 17 (6–22) 0.17Sit to stand to sit � 5 (seconds) 17 (10–27) 18 (9–46) 0.57Timed up and go (seconds to leave an armchair,

walk 3 m, return, sit down again)10 (7–20) 11 (6–27) 0.83

Radiating pain (visual analogue scale: 0–100)d 38 (0–95) 37 (0–94) 0.39Low back pain (visual analogue scale: 0–100)d 49 (0–87) 54 (0–91) 0.81

aHarmful drinkers were defined by an intake over the limits recommended by the Danish National Boardof Health (less than 21 drinks per week for men and 14 for women; one drink contained 12 g of ethanol).bAmerican Society of Anaesthesiology, classification of risk at surgery; I (no increased risk) to V (moribundpatient).cRonald Morris Questionnaire for evaluation function level for patients with low back pain and radiatingpain; the score ranges from 0 (no functional limitation) to 23 (the worst.40)dBrief Pain Inventory Questionnaire for the pain score level and causal factors.14




The only predictor for major postoperativecomplication was preoperative low back painintensity (odds ratio 10.45; 95% confidenceinterval 2.29–47.70; multivariate analyses) and theonly predictor for prolonged stay was belonging tothe control group (7.40; 1.92–28.50).

Discussion

We found that the integrated programme, whichincluded prehabilitation and early postoperativerehabilitation, was followed by better postopera-tive functionality, faster recovery and shorter hos-pital stay among patients undergoing lumbar spinesurgery. The control group left the hospital afterseven days, which was not significantly differentfrom the expected eight days. The integrated pro-gramme actively involved the patients withoutsimultaneously increasing the pain level or compli-cation rate; instead the patients expressed higherlevel of satisfaction.

The programme comprised several elements,which have been shown per se to improve the out-come after elective procedures other than spine sur-gery.20–22 Hospital stay duration is influenced byseveral factors, such as the development of compli-cations and second surgery, traditions and expecta-tions from the staff and the patient.23 The expectationcould influence the length of stay, exclusively.A study concerning day case discectomy notedthat the patients admitted as inpatients failed tobe discharged as day cases, whereas those admittedto the day case unit did.24 Furthermore, differencesbetween therapists’ and patients’ expectations andgoals may influence the length of hospital stay.

In order to avoid the latter we used a fixed mile-stones assessment to obtain identical functional-ities before discharge. Despite this, we found thatnearly all the patients stayed in the hospital forone day or more after they had met the milestonesand were ready for discharge. This delay in dis-charge was more pronounced for the interventiongroup. The prolongation was due to delayed orderof nursing in home, culture and routines in dis-charge procedures.

The intervention group reduced the postopera-tive stay by 29%. Part of the explanation could bethe intensive exercise programme. ThoughT

able

2Functionalit

yand

pain

inth

ein

terv

ention

and

contr

olgro

up

over

tim

e(m

edia

nand

range)

Inclu

sio

nA

dm

issio

nD

ischarg

e1

Month

3M

onth

s6

Month

s

Int.

gro

up

n¼

35

Cont.

gro

up

n¼

38

Int.

gro

up

n¼

28

Cont.

gro

up

n¼

32

Int.

gro

up

n¼

28

Cont.

gro

up

n¼

28

Int.

gro

up

n¼

28

Cont.

gro

up

n¼

28

Int.

gro

up

n¼

28

Cont.

gro

up

n¼

28

Int.

gro

up

n¼

28

Cont.

gro

up

n¼

28

Rola

nd

Morr

isQ

uestionnaire

16

(score

)

16

(8–21)

17

(6–22)

14

(1–21)

17

(7–23)

––

12

(3–21)

17

(1–23)

8(0

–20)

11

(0–22)

8(0

–20)

11

(0–23)

Sit

tosta

nd

17

(seconds)

17

(10–27)

18

(9–46)

14

(8–22)

18

(8–31)

––

12

(8–28)

14

(5–32)

11

(8–23)

12

(8–33)

12

(7–20)

13

(7–27)

Tim

ed

up

and

go

(seconds)

10

(7–20)

11

(6–27)

9(6

–31)

11

(7–27)

14

(8–26)

13

(8–35)

8(5

–31)

9(5

–18)

8(5

–17)

9(5

–14)

8(5

–18)

8(5

–13)

HR

QoL,

15D

18

(score

)0.8

3(0

.67–1.0

0)

0.7

9(0

.63–0.9

5)

0.8

5(0

.74–0.9

9)

0.8

2(0

.65–0.9

2)

0.8

3(0

.62–1.0

0)

0.7

9(0

.48–0.9

4)

0.8

8(0

.74–0.9

9)

0.8

6(0

.66–0.9

6)

0.9

0(0

.73–1.0

0)

0.8

9(0

.62–1.0

0)

0.9

1(0

.73–1.0

0)

0.9

0(0

.69–1.0

0)

Low

back

pain

(VA

S)

45

(0–87)

54

(0–91)

41

(0–78)

53

(0–96)

31

(2–77)

21

(0–85)

17

(0–50)

31

(0–87)

10

(0–93)

16

(0–81)

10

(0–69)

21

(0–62)

Radia

ting

pain

(VA

S)

38

(0–95)

39

(0–94)

41

(0–78)

53

(0–96)

––

17

(0–50)

31

(0–87)

10

(0–93)

16

(0–81)

10

(0–69)

21

(0–62)

Valu

es

are

media

n(r

ange).

Cont.

,contr

olgro

up;

Int.

,in

terv

ention

gro

up;

HR

Qol,

health-r

ela

ted

qualit

yof

life;

VA

S,

vis

ualanalo

gue

scale

.




intensive physical training is an integrated part ofthe rehabilitation programmes for non-surgicalpatients with low back pain,25,26 we have notfound any randomized studies looking at perio-perative training in relation to spinal surgery inthe literature. This could be explained by the rou-tines and clinical principles used by which patientslie down for a longer period after surgery inorder to reduce the risk of spinal instability.Furthermore, the classical discrepancy betweenmedical and surgical interventions may previouslyhave been a barrier to the evaluation of a trainingprogramme for patients scheduled for spinesurgery.

Research from other surgical procedures, how-ever, has presented conflicting results for physicaltraining. In some studies of hip replacement ther-apy the outcome improved,10,27 some only showedmoderate effect,28 while other authors did not seeany effect of preoperative exercise.29 Preoperativetraining did not influence the outcome after knee

replacement.30 Hitherto, no authors have pub-lished aggravated functions, more complicationsor prolonged stay as a result of training pro-grammes in the surgical period.

The shorter hospital stay and better functionalityin this study was not related to reduced develop-ment of complications in the intervention group.The explanation could be the overall low complica-tion rate in this study, and that the study was notdesigned to evaluate minor differences in complica-tions after lumbar surgery. This would require amuch larger study population. The risk of type 2failure should be considered; however, our findingsdid not indicate that a significant difference in post-operative complications between the two groupscould be achieved unless more than 4500 patientswere included. Instead, the early and intensiverehabilitation might have increased the demandsfor bodily resources, so we should be satisfiedwith a status quo regarding development of com-plications in the intervention group.

Table 3 Surgical intervention and outcome; given as number (per cent) or median (range)

Intervention (n¼ 28) Control (n¼32)

Primary spondylodesis instrumental 2 / 28 (7%) 2 / 32 (6%)Non-instrumental 23 / 28 (82%) 26 / 32 (81%)

Secondary spondylodesis (non-instrumental) 2 / 28 (7%) 3 / 32 (9%)Implementation of discus prosthesis 1 / 28 (4%) 1 / 32 (3%)Duration of operation (minutes) 110 (85–260) 135 (65–210)Fluid infusion at operation (mL) 2450 (1300–4500) 2700 (1700–4800)Patients with major complicationsa 8 8Cardiopulmonary insufficiency 0 0Thromboembolic complications 0 0Haematomas 1 2Severe pain (470 on VAS) 6 7Allergic reaction 1 0Patients with minor complications 4 4Pneumonia (antibiotics) 0 0Wound infection (antibiotics) 0 1Urinary tract infection (4105 bacteria/mL) 2 0Urinary retention (tubulation) 0 1Constipation (laxities) 2 2Patients with adverse events 3 1Delayed order of nursing at home 2 0Surgery on Fridayb 1 1Hospital stay after surgery (days) 5 (4–9)* 7 (5–15)Stay until reaching all milestones 4 (1–6)* 6 (3–13)Extra stay due to complications 18* 30

Values are number (%) or median (range).*P50.01.aRequiring intensive treatment or secondary surgery.bNot followed by physiotherapy in the weekend.




Although the intervention group experiencedless pain when measured over time, the numberof patients who developed episodes of severebreakthrough pain was similar in the two groups.Other research groups also found no significanteffect on pain relief and other complications byself-administrated analgesics taken epidurally,but the consumption of local anaesthetics andopioids was smaller and the patients has lessmotor block.31,32 It could be expected that ourintervention group had a higher pain score andhigher need for analgesics due to the intensivetraining programme, but this was not the case.A bias could be that we may have overlooked adifference in the duration of episodes with severepain because we only asked the patients about thepain level. Other limitations of our study are therelatively small number of patients and risk of biasdue to the non-blinded design. Both could add toan overestimation of positive results. Finally, therisk of type 1 failure (giving a significant result bychance) should be considered, which repeating thestudy could overcome.33 The drop-out rate in thepresent study of 19% was acceptable and may addto the strength of the study, especially becausesome of them did not completely drop out.Of the possible predictors only a high score of

preoperative low back pain was significantly asso-ciated with the development of postoperative com-plications. This is in agreement with the literatureshowing that preoperative pain is a predictor forthe incidence and intensity of postoperativepain,34,35 and severe postoperative pain may fur-ther lead to immobilization and increased risk ofpostoperative complications. The multivariateanalyses supported the conclusion that belongingto the control regime was associated with a signif-icantly prolonged postoperative stay.This study has demonstrated the benefits of

adding a prehabilitation programme to the earlyrehabilitation standard which is now under imple-mentation in our institution. In direct costs, thereduced postoperative hospital stay seems tooffset the extra cost of the preoperative outpatientprogramme.36 However, this may not be reflectedin the balances because of a lack of sensitivity inour reimbursement systems, which are based onaverage costs.Even though the differences between the two

groups seem minor, the clinical relevance of the

results (i.e. shorter period of disability) seemsmajor for the individual patient, which is alsoreflected by the higher level of satisfaction.

We could not show any effect after three or sixmonths. It would be interesting to combine theprehabilitation programme with a late rehabilita-tion programme beginning three months post-operatively as described by Christensen et al.37,38

A ‘back-cafe group’ had significant better dailyfunction, lesser pain and lower cost after eightweeks of video-training at home combined withcafe meetings compared with physiotherapyprogramme or video-training, exclusively.37,38

Another relevant issue would be to measurewhether a postoperative long-term exercise regi-men would be beneficial by inhibiting the relativefrequent relapses of pain within the first years afterthe surgical intervention.

We could not identify the most effective ele-ments or exclude ineffective elements from thiscombined programme due to the pragmaticdesign. Furthermore, it is not possible to excludea positive impact from the expected discharge day.This may be a minor effect caused by the earlierdischarge in the control group than expected.

As this is the first study to evaluate an inte-grated perioperative programme adding prehabili-tation to early rehabilitation, confirmation fromother studies is needed. Future studies shouldfocus on optimization of the treatment, for exam-ple introduction of further optimized surgical pro-cedures39 and organizational improvements thatmight reduce the complication rate. It would alsobe relevant to investigate the effect of the prehabi-litation concept separately as well as the integratedprogramme on different surgical procedures.The new research should also consider economicconsequences and cost-effectiveness analysis. Apossible effect of the integrated programme inother surgical interventions also remains to beinvestigated.

In conclusion, besides surgery a package ofattentional efforts consisting of an integrated pro-gramme of prehabilitation and early rehabilitationimproved the early outcome and shortened thehospital stay for patients undergoing electivesurgery for degenerative spine disease – withoutproducing more pain or dissatisfaction. As a con-sequence of several pragmatic aspects it is not pos-sible from this study to identify the effective




elements in the programme. In an optimal clinicalsetting our new integrated programme wouldallow us to operate on elective spine patients onMondays, aiming the discharge at Fridays, thusreducing the workload over the weekends.However, there is still room for further improve-ment of the outcome.

Clinical messages

� A prehabilitation programme beginning twomonths prior to operation significantlyimproved patients’ functionality at admis-sion for operation.

� After surgery the intervention group reachedthe recovery milestones significantly fasterthan the control group and left hospitalearlier.

� Patient satisfaction was significantly higherin the intervention group than in the controlgroup.

ContributorsPN designed and managed the study, recruited

the patients and obtained, recorded, and inter-preted the data. LJ recruited the patients andobtained and recorded the data. BD recruited thepatients and interpreted the data. TP analysedand interpreted the data. HT designed thestudy and analysed and interpreted the data. PNand HT wrote the paper. LJ, BD and TP editedthe paper.

Conflict of interestNo conflicts of interest.

References

1 Deyo RA, Gray DT, Kreuter W, Mirza S,Martin BI. United States trends in lumbar fusionsurgery for degenerative conditions. Spine 2005; 30:1441–45.

2 Weinstein JN, Lurie JD, Tosteson TD et al.Surgical versus nonsurgical treatment for lumbardegenerative spondylolisthesis. N Engl J Med2007; 356: 2257–70.

3 Resnick DK, Choudhri TF, Dailey AT et al.Guidelines for the performance of fusion proce-dures for degenerative disease of the lumbarspine. Part 1–17. J Neurosurg Spine 2005;2: 637–38.

4 Martin BI, Mirza SK, Comstock BA, Gray DT,Kreuter W, Deyo RA. Reoperation rates follow-ing lumbar spine surgery and the influence ofspinal fusion procedures. Spine 2007; 32: 382–87.

5 Langelotz C, Spies C, Muller JM, Schwenk W.‘Fast-track’-rehabilitation in surgery, a multimo-dal concept. Acta Chir Belg 2005; 105: 555–59.

6 Kehlet H. Fast-track colonic surgery: status andperspectives. Recent Results Cancer Res 2005;165: 8–13.

7 Kehlet H, Wilmore DW. Fast-track surgery.Br J Surg 2005; 92: 3–4.

8 Carli F, Zavorsky GS. Optimizing functionalexercise capacity in the elderly surgicalpopulation. Curr Opin Clin Nutr Metab Care2005; 8: 23–32.

9 Tonnesen H, Nielsen PR, Lauritzen JB, MollerAM. Smoking and alcohol intervention beforesurgery: Evidence for best practice. Br J Anaesth2009; 102: 297–306.

10 Siggeirsdottir K, Olafsson O, Jonsson H,Iwarsson S, Gudnason V, Jonsson BY. Shorthospital stay augmented with education andhome-based rehabilitation improves function andquality of life after hip replacement: randomizedstudy of 50 patients with 6 months of follow-up.Acta Orthop 2005; 76: 555–62.

11 Kehlet H, Dahl JB. Anaesthesia, surgery, andchallenges in postoperative recovery. Lancet 2003;362: 1921–28.

12 Sackett DL, Gent M. Controversy in countingand attributing events in clinical trials. N Engl JMed 1979; 301: 1410–12.

13 Gartlehner G, Hansen RA, Nissman D,Lohr KN, Carey TS. A simple and valid tooldistinguished efficacy from effectiveness studies.J Clin Epidemiol 2006; 59: 1040–48.

14 Klepstad P, Loge JH, Borchgrevink PC,Mendoza TR, Cleeland CS, Kaasa S. TheNorwegian brief pain inventory questionnaire:translation and validation in cancer pain patients.J Pain Symptom Manage 2002; 24: 517–25.

15 Albert HB, Jensen AM, Dahl D, Rasmussen MN.[Criteria validation of the Roland Morrisquestionnaire. A Danish translation of the inter-national scale for the assessment of functional




level in patients with low back pain and sciatica].Ugeskr Laeger 2003; 165: 1875–80.

16 Roland M, Fairbank J. The Roland-MorrisDisability Questionnaire and the OswestryDisability Questionnaire. Spine 2000; 25: 3115–24.

17 Simmonds MJ, Olson SL, Jones S et al.Psychometric characteristics and clinical useful-ness of physical performance tests in patients withlow back pain. Spine 1998; 23: 2412–21.

18 Sintonen H. The 15D instrument of health-relatedquality of life: properties and applications.Ann Med 2001; 33: 328–36.

19 Sintonen H. An approach to measuring andvaluing health states. Soc Sci Med [Med Econ]1981; 15: 55–65.

20 Kehlet H, Wilmore DW. Multimodal strategies toimprove surgical outcome. Am J Surg 2002; 183:630–41.

21 Healy WL, Ayers ME, Iorio R, Patch DA,Appleby D, Pfeifer BA. Impact of a clinicalpathway and implant standardization on total hiparthroplasty: a clinical and economic study ofshort-term patient outcome. J Arthroplasty 1998;13: 266–76.

22 Scranton PE, Jr. The cost effectiveness of stream-lined care pathways and product standardizationin total knee arthroplasty. J Arthroplasty 1999;14: 182–86.

23 Kehlet H. Future perspectives and research initia-tives in fast-track surgery. Langenbecks Arch Surg2006; 391: 495–98.

24 Gonzalez-Castro A, Shetty A, Nagendar K,Greenough CG. Day-case conventional discect-omy: a randomised controlled trial. Eur Spine J2002; 11: 67–70.

25 Brox JI, Reikeras O, Nygaard O et al. Lumbarinstrumented fusion compared with cognitiveintervention and exercises in patients with chronicback pain after previous surgery for disc hernia-tion: a prospective randomized controlled study.Pain 2006; 122: 145–55.

26 Brox JI, Sorensen R, Friis A et al. Randomizedclinical trial of lumbar instrumented fusion andcognitive intervention and exercises in patientswith chronic low back pain and disc degeneration.Spine 2003; 28: 1913–21.

27 Rooks DS, Huang J, Bierbaum BE et al. Effect ofpreoperative exercise on measures of functionalstatus in men and women undergoing total hipand knee arthroplasty. Arthritis Rheum 2006; 55:700–708.

28 Petersen MK, Madsen C, Andersen NT,Soballe K. Efficacy of multimodal optimization ofmobilization and nutrition in patients undergoinghip replacement: a randomized clinical trial.Acta Anaesthesiol Scand 2006; 50: 712–17.

29 Rodgers JA, Garvin KL, Walker CW,Morford D, Urban J, Bedard J. Preoperativephysical therapy in primary total knee arthro-plasty. J Arthroplasty 1998; 13: 414–21.

30 Ackerman IN, Bennell KL. Does pre-operativephysiotherapy improve outcomes from lower limbjoint replacement surgery? A systematic review.Aust J Physiother 2004; 50: 25–30.

31 Silvasti M, Pitkanen M. Patient-controlledepidural analgesia versus continuous epiduralanalgesia after total knee arthroplasty. ActaAnaesthesiol Scand 2001; 45: 471–76.

32 van der Vyver M, Halpern S, Joseph G. Patient-controlled epidural analgesia versus continuousinfusion for labour analgesia: a meta-analysis. BrJ Anaesth 2002; 89: 459–65.

33 Andersen B. Methodological errors in medicalresearch. Oxford, Blackwell Scientific, 1990.

34 Per Rotbøll Nielsen, Asa Rudin, Mads U.Werner.Prediction of postoperative pain. Curr AnaesthCrit Care 2007; 18: 157–65.

35 Kalkman CJ, Visser K, Moen J, Bonsel GJ,Grobbee DE, Moons KG. Preoperative predic-tion of severe postoperative pain. Pain 2003; 105:415–23.

36 Nielsen PR, Andreasen J, Asmussen A, TønnesenH. Costs and quality of life for fast tracking spinesurgery. BMC Health Serv Res 2008; 8: 209.

37 Christensen FB, Laurberg I, Bunger CE.Importance of the back-cafe concept to rehabilita-tion after lumbar spinal fusion: a randomizedclinical study with a 2-year follow-up. Spine 2003;28: 2561–69.

38 Soegaard R, Christensen FB, Lauerberg I,Bunger CE. Lumbar spinal fusion patients’demands to the primary health sector: evaluationof three rehabilitation protocols. A prospectiverandomized study. Eur Spine J 2006; 15: 648–56.

39 Nuelle DG, Mann K. Minimal incision protocolsfor anesthesia, pain management, and physicaltherapy with standard incisions in hip and kneearthroplasties: the effect on early outcomes.J Arthroplasty 2007; 22: 20–25.

40 Roland M, Morris R. A study of the naturalhistory of back pain. Part I: development of areliable and sensitive measure of disability inlow-back pain. Spine 1983; 8: 141–44.




Documents

Prehabilitation and early rehabilitation after spinal surgery: randomized clinical trial