Embed Size (px)
Citation preview
FcoDSgP
AR
E
SC
Co
0
CC
d
71
Impact of preoperative anxiolytic onsurgical site infection in patientsundergoing abdominal hysterectomy
Rosa Levandovski, MSH,a Maria Beatriz Cardoso Ferreira, MD, PhD,a,b Maria Paz Loayza Hidalgo, MD, PhD,a,cCassio Alves Konrath, MD,d Daniel Lemons da Silva, MD,d and Wolnei Caumo, MD, PhDa,b,d
Porto Alegre, Brazil
romine,gy Depa
ervire (orto
ddreal,
-mai
upplınic
onflf the
196
opyont
oi:1
8
Background: An increased anxiety may be associated with a higher risk of surgical site infection (SSI), but there is little objectivedata on the effect of preoperative anxiolytic interventions on SSI. To address this issue, we evaluated the effects of preoperativediazepam on postoperative SSI following abdominal hysterectomy.Methods: This randomized, double-blinded, placebo-controlled study included 130 patients, American Society of Anesthesiologistphysical status 1 or 2. Patients were randomly assigned to receive either oral diazepam 10 mg (n 5 65) or placebo (n 565) the nightbefore and 1 hour prior to surgery. The assessment instruments were the Visual Analogue Scale and the State-Trait Anxiety Inventory.SSI was diagnosed according to the criteria of the Centers for Disease Control and Prevention with standard follow-up of 30 days.Results: The relative risk (RR) was 1.79 (95% confidence interval [CI]: 1.31-2.43), and the number of patients that needed to betreated was 5.2 (95% CI: 2.74-50.76) to prevent 1 additional SSI. The RR for SSI in placebo-treated patients with high postoperativeanxiety was 1.65 (95% CI: 1.07-2.56).Conclusion: Diazepam-treated patients showed lower postoperative anxiety and lower incidence of SSI up to 30 days after surgerycompared with placebo in patients undergoing abdominal hysterectomy. (Am J Infect Control 2008;36:718-26.)
Wound infections are among the most commonserious complications of anesthesia and surgery.1
Surgical wound infections can prolong hospitalization2
and substantially increase the cost of care.1,3 The firstfew hours after tissue is contaminated by bacteria con-stitute a critical period during which wound infection isestablished,4 but infections are typically not detecteduntil some days after surgery. The factors that influ-ence the incidence of surgical wound infection includesite and complexity of surgery,3 patient’s underlyingillness, smoking, obesity,5,6 blood transfusion, hyper-glycemia,7 presence or absence of hypovolemia,8
patient’s temperature during surgery,2 use or not of
the Postgraduate Program in Medical Sciences,a School of Medi-Universidade Federal do Rio Grande do Sul (UFRGS); Pharmacol-epartment,b Instituto de Ciencias Basicas da Saude of UFRGS;
rtment of Psychiatry,c School of Medicine, UFRGS; and Anesthesiace and Perioperative Medicine,d Hospital de Clınicas de Porto Ale-HCPA)/Universidade Federal do Rio Grande do Sul (UFRGS),Alegre, Brazil.
ess correspondence to Wolnei Caumo, MD, PhD, Coronel Corte295 Bairro Petropolis CEP: 90630-008, Porto Alegre, RS, Brazil.
orted by The Postgraduate Research Group (GPPG) at Hospital deas de Porto Alegre.
icts of interest: There were no financial relationshipa between anyauthors or any commercial interest in the outcome of this study.
-6553/$34.00
right ª 2008 by the Association for Professionals in Infectionrol and Epidemiology, Inc.
0.1016/j.ajic.2007.12.010
prophylactic antibiotics,9 and degree to which painand anxiety are controlled.10
Studies published in the health-psychology literaturesuggest that increased preoperative anxiety is associatedwith poor postoperative behavioral and clinical recov-ery.11,12 Furthermore, numerous reports indicate thatpreoperative psychologic interventions aimed at reduc-ing preoperative anxiety may also result in improvedpostoperative behavioral and clinical recovery.12-14
Additionally, several reports indicate that increased per-ioperative anxiety and uncontrolled postoperative painmay increase the perioperative neuroendocrine stressresponse.14,15 Thus, a possible hypothesis is that factorssuch as anxiety and pain can stimulate sympatheticvasoconstriction16,17 and thereby increase the risk ofpostoperative infection.18 Accordingly, one can hypoth-esize that the use of preoperative benzodiazepine canreduce postoperative anxiety and enhance the postoper-ative clinical recovery process involved in warding offinfection. To test this hypothesis, we conducted arandomized, double-blind, placebo-controlled trial toevaluate the effect of preoperative diazepam on the post-operative incidence of surgical site infection (SSI) duringthe first 30 days following abdominal hysterectomy.
MATERIALS AND METHODS
Study population
Following ethics committee approval and written in-formed consent, 130 patients, American Society of An-esthesiologist classification 1 or 2, aged 19 to 60 years,
Levandovski et al December 2008 719
scheduled to undergo total abdominal hysterectomy,were enrolled in the randomized, double-blind, pla-cebo-controlled study. The patients were undergoingelective abdominal hysterectomy for myomatosis inthe Gynecology Service, and all operations were per-formed by a third-year resident supervised by one ofthe attending gynecologists. All patients were admittedto the hospital 1 day before surgery, and those patientswho did not develop complications were dischargedon day 3 postsurgery, on any day of the week. Patientswith contraindications to regional anesthesia; mentalimpairment; chronic pain; or a history of congestiveheart failure, valvular heart disease, renal or hepatic dis-ease, psychotropic drug use, and language or communi-cation difficulties were excluded. In addition, patientswith a body mass index (BMI) higher than 35 kg/m2
and those with a history of psychiatric disorder and/orpositive screening for minor psychiatric disorders(scores, $8) on the World Health Organization’s (WHO)Self-Reporting Questionnaire (SRQ-20) were excluded.19
The SRQ-20 measures somatic symptoms, depressivemood, depressive thoughts, and decreased energy.
Randomization and interventions
The treatment allocation method used wasadvanced simple randomization without blocking orstratification. Before the recruitment phase of thestudy, 160 envelopes containing all protocol materialswere prepared and numbered sequentially. A randomnumber was used to assign each consecutively num-bered envelope to receive either 10 mg oral diazepamor placebo the night before (10 PM) and 1 hour priorto surgery, and the envelopes were grouped so thateach had an independent 50% probability of beingincluded in either group. A sheet indicating the allo-cated treatment was then placed in the envelope, andthe envelopes were sealed. Throughout the course ofthe study, the sealed envelopes were removed andopened sequentially by a pharmacy technician, whodelivered the tablets of diazepam 10 mg or placeboonly after prospective patients had been screenedand had consented to participation. No other preoper-ative medication was given. During the entire protocoltime line, blinding and randomization were undertakenby 2 investigators who were not involved in thepatient’s evaluation. Other individuals involved in thepatient’s care were unaware of the treatment group towhich the patient belonged.
Outcome measures
Our major outcome (SSI) was either of 2 distinct cri-teria for diagnosis. The first criterion, as in previousstudies,2,20 was that wounds were considered likely tobe infected when pus could be expressed from the
incision or aspirated from a loculated mass within thewound. The second criterion was based on the Centersfor Disease Control and Prevention (CDC)’s National Nos-ocomial Infections Surveillance (NNIS) system, modifiedin 1999,5 in which SSI is classified as being either inci-sional (superficial ordeep) ororgan space with a diagnos-tic period up to 30 days. At all times, the gynecologistexaminer was blinded to the aim of the study, preopera-tive interventions, and all other measures.
Instruments and assessment
The data were collected in the following multiplestandardized phases: (1) during the period of preopera-tive evaluation, a structured questionnaire was used tocollect information concerning demographic charac-teristics, clinical state, and anxiety level; (2) the surgicalprocedures were assessed using the anesthesiologist’srecords; (3) during the patients’ stay at the hospital,the gynecologists examined the surgical wound dailyfor the presence of SSI; (4) after discharge, all patientswere examined 2 times, at 1 and 4 weeks of follow-up, carried out by a gynecologist in the same hospital,in the outpatient service. Data regarding drug records,medical and diagnostic examinations, reinterventions,other medical procedures, and hospital readmissionfor treatment of surgical complications were allcollected from the patient’s records. Two independent,trained data collectors followed a standard method bywhich they reviewed all clinical and laboratoryinformation available in the hospital. Urinary and res-piratory infections were defined by clinical and labora-tory criteria. The gynecologist in charge of dischargewas blinded to the aim of the study and all other mea-sures. The length of hospital stay was defined as thenumber of days spent in the hospital.
Assessment of pain and psychologic state
The day prior to surgery, the same anesthesiologist,who provided patients with information on the periop-erative course and instructed them regarding the use ofthe patient-controlled analgesia pump, saw all patients.Moreover, each patient underwent psychologic testingand pain evaluation.
The pain was measured by a 100-mm Visual AnalogScale (VAS), by which scores ranged from no pain (zero)to worst possible pain (100 mm).21 Incisional pain atrest was defined as the average of pain ratings obtainedat 6 and 24 hours following surgery. A 30-mm cut-offpoint was used to classify the patients into 2 groups:(1) absence of pain or mild pain (scores equal to orless than 30 mm); and (2) moderate, intense, or worstpossible pain (scores greater than 30 mm).21 Incisionalpain on coughing, deep breathing, or movement as-sessed using yes or no type questions and the patient’s
720 Vol. 36 No. 10 Levandovski et al
desire for additional analgesic were assessed at 6 and 24hours after surgery. The presence of incisional pain dur-ing coughing or movement and the desire for additionalanalgesic in any assessment were the criteria used toclassify the patients as presenting or not with pain oncoughing or movement and desiring or not additionalanalgesia. Sedation was assessed by the following crite-ria: zero, completely alert; 1, dozes now and then but an-swers immediately; 2, often asleep but arousable; 3,asleep and not arousable. A sedation level equal to orhigher than 2 in any assessment was used as the cut-off point to classify patients as follows: (1) mild sedationlevel (,2) or (2) moderate to intense sedation level ($2).
To measure pre- and postoperative anxiety, we usedthe State-Trait Anxiety Inventory (STAI)22 validatedfor the Brazilian population: state-anxiety (transitoryanxiety that varies according to the situation) andtrait-anxiety (stable personality disposition reflectinggeneral level of fearfulness). The postoperative anxietywas defined as the average of the scores obtained at 2time points: 6 and 24 hours following surgery.13 Thecut-off point to classify high anxiety or low anxietywas the highest quartile so that individuals with a scoreabove the highest quartile were classified as highlyanxious and those with a score equal to or below thehighest quartile as mildly anxious. The highest quartilefor postoperative state-anxiety was 45.5. The stratifica-tion procedure was based on Spielberger and Diaz-Guerrero’s manual.22 To ensure that the evaluatorswere blinded to preoperative psychologic state, thephysicians involved in the preoperative evaluationdid not perform the postoperative assessment.
Anesthesia technique
All patients had standard monitoring prior to epidu-ral anesthesia. An isotonic saline solution (10 mL � kg-1),cefazolin (2 g), and intravenous (IV) fentanyl (100 mg)were administered. Afterward, all patients had anextradural catheter inserted at lumbar segments L2/L3 or L3/L4. An injection of 15 to 20 mL ropivacaine(10 mg � mL-1) was given epidurally. If there were signsof inadequate analgesia, additional 5-mL doses wereused. During surgery, all patients received propofolcontinuously to maintain sedation and supplementaryoxygen by an intranasal catheter. A vesicourethral cath-eter was routinely maintained until the first postopera-tive day. In the operating room, the patient’s skin wasdisinfected with alcoholic chlorhexidine. At the endof surgery, sedation was stopped, and the extraduralcatheter was removed.
Postoperative analgesia
The PCA device (APM; Abbott, Chicago, IL) was avail-able as soon as the patient arrived in the postanesthesia
care unit. The PCA was programmed to deliver a 2-mgmorphine bolus with a lockout period of 10 minutes.The maximum dose that could be requested during a4-hour period was 30 mg. If the pain was not relieved,an additional dose of 20 mg/day of tenoxicam was ad-ministered intravenously. The patients were connectedto PCA for the entire period of 72 hours after surgery.Analgesia in all patients at 72 hours after surgery wasmaintained with codeine 30 mg plus acetaminophen500 mg every 6 hours and 75 mg sodium diclofenacthree times per day. If pain was still not relieved,1000 mg dipyrone was given every 6 hours.
Statistical analysis
For the initial power analysis, the number of sub-jects in each study group was determined based onprevious studies involving the incidence of SSI accord-ing to the American NISS risk index23 and the effect ofinterventions to attenuate the preoperative stress re-sponse on postoperative outcomes.14 To determinethe power analysis, an infection rate of 2.5% was ex-pected in the diazepam-treated patients, which corre-sponds to a prorating incidence of SSI reported in alarge cohort study in patients having undergone ab-dominal hysterectomy.23 However, the rate used inthe placebo group was the incidence of postoperativeinfection reported in placebo-treated patients in anearlier study that assessed the effect of interventionsto attenuate the preoperative stress response on post-operative outcomes, which was 16%.14 The analysisindicated that a study of 130 subjects (65 per group)was required to detect a 15% difference betweengroups with a power of 80% and a set at 0.05.24 Con-sidering that the sample size calculation is just an esti-mation, we thus decided to calculate the power of thisanalysis based on the incidence of SSI that we found inthis study, that is, 25.8% in the placebo-treated and6.6% in the diazepam-treated patients, which resultedin a power of 89.72% (with a 2-tailed a level of 0.05).
The differences between groups for continuing datawere examined by the t test for independent samples,and categorical data were examined by x2 or Fisherexact tests. Multivariate logistic regression was usedfollowing the stepwise forward procedure to minimizethe effect of the differences between groups of otherfactors showing significant imbalance in the univariateanalysis, except for preoperative anxiolytic interven-tion.25 The variables related to pain were included with-out considering the statistical significance of univariateanalysis because a previous study suggested that post-operative pain10 could be linked to SSI.14 To analyzethe effect of morphine doses used in the first 24 hoursafter surgery, the cut-off points were established usingthe mean (60 mg � kg-1). For all analyses, statistical
Fig 1. Flow diagram of study, including number ofpatients at each point.
Levandovski et al December 2008 721
significance was set at P , .05, 2-tailed. Data were ana-lyzed using SPSS version 11.0 (SPSS, Inc, Chicago, IL).The magnitude of the effect of preoperative diazepamwas demonstrated by the number of patients whoneeded to be treated (NNT) to avoid 1 additional SSI.
RESULTS
Figure 1 is a flow diagram of the 130 women eligibleto participate in the trial. Patients’ characteristics andperioperative variables are summarized in Tables 1 to3. Demographic and morphometric characteristicswere similar in patients assigned to diazepam andplacebo. However, the diazepam-treated patients dem-onstrated a level of anxiety significantly lower thanthat of placebo-treated patients throughout the first24 hours after surgery (Table 3). In the placebo group,59.7% (40/62) showed high state-anxiety postopera-tively, compared with 21.4% (27/61) in the diazepamgroup. In the placebo group, the relative risk (RR) was1.52 (95% confidence interval [CI]: 1.04-2.22) and theNNT 5.15 (95% CI: 2.74-42.55) to observe 1 additionalpatient with high anxiety in the postoperative period.
The overall incidence of SSI was 15.20% (95% CI:10.23-21.75), in which 6.1% had the diagnosis beforedischarge and 9.1% had the diagnosis after discharge.The superficial SSI rate was 8.6% (95% CI: 4.89-14.10), the deep incision infection rate was 4.2%(95% CI: 1.87-8.89), and the organ/cavity infection fre-quency was 2.4% (95% CI: 0.78-6.48). The placebo-treated group showed an SSI incidence of 25.8% (16/62), compared with 6.6% (4/61) in the diazepam group.In the placebo-treated patients, the RR to SSI was 1.79(95% CI: 1.31-2.43) and the NNT 5.2 (95% CI: 2.74-50.76). Subgroup analysis demonstrated that theincidence of SSI in highly anxious placebo-treatedwomen was 47.6% and 10% in diazepam treated andthat the RR to SSI in placebo treated was 1.65 (95%CI: 1.07-2.56). The incidence of SSI in mildly anxiousplacebo-treated patients was 14.6% and 3% in diaze-pam-treated patients, and the RR to SSI in placebo-
treated patients was 1.58 (95% CI: 0.93-2.67). Therewas no case of mortality, sepsis, or major cardiovascu-lar complications in either group. The length of hospi-tal stay was significantly longer in placebo-treatedcompared with diazepam-treated patients (Table 3),and this difference was even larger when comparingthose patients who developed a SSI compared withpatients without SSI (7.25 6 2.73 days vs 4.66 6 0.9days; t 5 4.19, P 5 .00, respectively).
The multivariate analysis summarized in Table 4confirmed the effect of preoperative anxiolytic inter-vention on SSI infection after controlling for dose oflocal anesthetic used during the surgery, morphine con-sumption during the first 24 hours after surgery, type ofsurgical incision, and postoperative incisional pain atrest and during coughing or movement (Table 4).
DISCUSSION
Patients treated with diazepam preoperatively had alower risk of high state-anxiety postoperatively andlower incidence of SSI. Furthermore, it was demon-strated that patients who were highly anxious postop-eratively had a higher incidence of SSI up to 30 daysafter surgery. These findings are supported by a previ-ous study in patients undergoing minor outpatient sur-gery, in which midazolam-treated patients showedlower rates of postoperative infection compared withplacebo-treated patients.14 Additionally, the presentfindings may be supported by the immunologic resultsof a previous randomized clinical trial, in which pla-cebo-treated patients exhibited a greater postoperativeB-cell response compared with midazolam-treatedpatients, suggesting that anxiety may influence thefunction of circulating immune cells in the immediatepostoperative period.26 In addition, we can hypothe-size that diazepam used preoperatively itself could re-duce the risk of infection. This hypothesis may besupported by evidence from studies in animals, possi-bly by diverse mechanisms such as an interaction ofbenzodiazepine with the benzodiazepine receptors(both central and peripheral) that form part of a molec-ular network that mediates the effects of stress andanxiety on immune function. The pathways by whichendogenous and exogenous ligands of benzodiazepinereceptors may modulate the immune system involvethe attenuation of behavioral stress responses inducedby corticotrophin releasing hormone (CRH), responsi-ble for immune suppression.27 Additionally, diazepammay decrease the release of tumor necrosis factor-a,interleukin-1, and interleukin-6 by macrophages.28 Inaddition, its action on peripheral receptors inhibits ap-optosis of neutrophils, which is essential for maintain-ing immune homeostasis and limiting host tissuedamage by promoting resolution of the inflammatory
Table 1. Characteristics of the study sample
Group
Characteristic Placebo (n 5 62) Diazepam (n 5 61) P value
Formal education, yr* 6.19 6 3.77 6.54 6 3.38 .59
Age, yr* 45.29 6 5.87 44.54 6 7.18 .52
Weight, kg* 67.66 6 11.79 69.44 6 10.19 .37
Body mass index* 26.45 6 3.95 27.43 6 4.55 .21
Smoking status, yes/noy 15/47 10/51 .20
Diabetes, yes/noy 5/57 4/57 .51
Alcohol consumption, yes/noy 4/58 6/55 .36
ASA status, 1/2y 24/38 27/34 .33
Hemoglobin preoperative, mg 11.17 6 2.59 11.8161.70 .14
Number of previous surgeries* 1.58 6.1.54 1.59 6 1.27 .97
Preoperative pain reported on Visual Analogue Scale* 0.96 6 1.80 1.42 6 2.11 .20
Trait-anxiety* 40.95 6 11.13 40.84 6 8.73 .83
Preoperative state-anxiety* 40.05 6 10.10 39.98 6 8.92 .88
NOTE. Values are means (SD) or frequencies (n 5 123).
*Unpaired t test to compare means 6 standard deviation.yFisher exact test or x2 test to compare frequencies.
Table 2. Clinical variables measured during intraoperative periods
Group
Characteristic Placebo (n 5 62) Diazepam (n 5 61) P value
Type of abdominal hysterectomy*
Hysterectomy/ Hysterectomy plus ophoorectomy 25/37 28/33 .33
Type of incision*
Vertical/ pfannenstiel 33/29 25/35 .14
Bleeding, mLy 625.85 6 535.84 551.84 6 449.35 .40
Red cell transfusion, No. of patients (mean 6 SD)y 8 (531.75 6 150.73) 10 (488.60 6 194.22) .48
Duration of surgery, miny 154.27 6 39.69 149.05 6 37.84 .20
Dose of ropivacaine (mg � kg-1)y 158.83 6 32.86 136.74 6 35.08 .00z
Intraoperative fentanyl dose (mg � kg-1)y 2.23 6 2.12 2.39 6 1.82 .65
Anesthetic or surgical complications, yes/no* 0/62 0/61 —
NOTE. Values are means (SD) or frequencies (n 5 123).
*Fisher exact test or x2 test to compare frequencies.yUnpaired t test to compare means 6 standard deviation.zP , .05.
722 Vol. 36 No. 10 Levandovski et al
response.29 Thus, these effects may influence the in-flammatory response to surgery and explain the effectshown in diazepam-treated patients. Although we didobserve this clinical phenomenon, our data only per-mit us to hypothesize this cause-effect relationshipbecause the diazepam effect on apoptosis was not di-rectly assessed. In addition, if we consider that periop-erative stress is determined by pre- and postoperativeanxiety, it is possible that placebo-treated patientshad higher rates of postoperative infection because ofincreased anxiety levels throughout the postoperativeperiod, whereas the prolonged anxiolytic effect of diaz-epam attenuated the anxiety in those who received thebenzodiazepine.30
This hypothesis can explain the contrast betweenour findings and the previous investigation that as-sessed the effects of preoperative midazolam on post-operative anxiety in patients undergoing abdominal
hysterectomy,14 in which midazolam had a minimalimpact on the clinical recovery process.26 Possibly,this discrepancy may be attributed to a short midazo-lam half-life, leaving the patients without the benefitsof anxiolysis during most of the postoperative period.Therefore, because the anxiety level is related to thetype and the extension of surgery and because thepurpose of this investigation was to determine whetherpreoperative anxiolytic could reduce the incidence ofSSI in patients after abdominal hysterectomy, we chosediazepam, which could extend its effect throughout thepostoperative period. However, further studies areneeded to elucidate the possible mechanism impli-cated in this response and to clarify the issue aboutthe impact of different pharmacologic and nonpharma-cologic anxiolytic interventions on postoperative SSI.
In addition, there may be additional mechanisms toexplain our SSI findings. First, one mechanism may be
Table 3. Clinical variables measured during the postoperative period
Group
Characteristic Placebo (n 5 62) Diazepam (n 5 61) P value
Hemoglobin postoperative, mg 9.69 6 2.17 9.94 6 1.95 .67
Oxygen in the Postoperative Care Unit, yes/no* 60/02 59/02 .68
Time of urinary catheterization after surgery, days 1.16 6 0.79 1.31 6 0.88 .34
Urinary infection, yes/no* 12/50 10/51 .42
Pulmonary infection, yes/no* 02/60 01/60 .50
Postoperative time to eliminate first flatus, daysy 1.27 6 0.51 1.36 6 0.60 .39
Sedation level in any assessment in the first 24 h after surgery, $2/,2 16/46 17/44 .48
Mean of postoperative state-anxiety 41.91 6 8.88 38.13 6 6.94 .01z
Incisional pain on VAS 6 h after surgeryy 4.83 6 3.07 5.13 6 2.74 .20
Incisional pain on VAS 24 h after surgeryy 1.73 6 1.68 2.17 6 1.82 .16
Incisional pain on coughing or deep breathing 6 h after surgery, yes/no* 34/28 29/32 .27
Incisional pain on coughing or deep breathing 24 h after surgery, yes/no* 39/23 38/23 .54
Some incisional pain at rest but no desire for additional analgesic, yes/no* 17/45 21/40 .26
Time to delivery of the first dose of morphine after surgery, miny 69.77 6 88.39 58.97 6 82.71 .48
Morphine consumption during the first 24 h after surgery, mgy 66.39 6 23.98 57.65 6 23.92 .04z
Morphine consumption during the first 24 h after surgery, .60/#60 mg 41/21 21/40 .00z
Morphine consumption during the first 72 h after surgery, mg 87.62 6 34.16 83.48 6 35.48 .51
Number of antiinflammatory doses used during the first 72 h after surgeryy 5.08 6 3.68 5.66 6 3.93 .41
Surgical site infection up to 30 days after surgery, yes/no 16/46 4/57 .00z
Length of hospital stay, daysy 5.61 6 1.84 4.54 6 1.24 .00z
NOTE. Values are means (SD) or frequencies (n 5 123).
*Fisher exact test or x2 test to compare frequencies.yUnpaired t test to compare means 6 standard deviation.zP , .05.
Table 4. Results of stepwise forward logistic regression analysis of the adjusted effect of preoperative anxiolyticintervention on SSI
Beta SE P value OR* 95% CI
Preoperative placebo use 1.38 0.622 .03 3.99 (1.18-13.53)
Incisional pain on coughing or movement 1.347 0.583 .02 3.84 (1.22-12.06)
Morphine consumption .60 mg during the first 24 h after surgery 1.165 0.591 .04 3.20 (1.01-10.21)
NOTE. N 5 123.
*Adjusted by type of surgical incision, postoperative incisional pain at rest, and local anesthetic dose.
Levandovski et al December 2008 723
similar to that responsible for the finding reported instudies involving preoperative psychologic interven-tions.31 Given that preoperative anxiety is associatedwith a surge of stress hormones,31 if we are able to pre-vent or decrease this surge, we may in fact change theonset of the entire perioperative neuroendocrinologicstress response. This in turn could result in an overalldecrease in the global hormonal stress response. Sec-ond, previous reports indicate that multiple periopera-tive factors, including anxiety, hypovolemia, pain, andcold, all stimulate sympathetic vasoconstriction16,17
and, therefore, probably increase the risk of infec-tion.18 Of interest in this regard are the multiple studiesin the psychosomatic field that have reported an in-creased incidence of various infectious processes inpatients who were more anxious and felt more stress.32
Finally, even though the mechanism involved in thisresponse is not clear, the magnitude of clinical effect
of diazepam in the incidence of SSI is undoubtedlyimportant because the NNT relative to placebo for 1 ad-ditional patient to show SSI was 5.15, and, to the best ofour knowledge, this is the first clinical trial specificallydesigned to assess the effect of preoperative anxiolyticmedication on the incidence of SSI, which extendssome literature data.
In addition, in the present study, patients withincisional pain during coughing or movement had ahigher incidence of SSI, even though they did notwish additional analgesic (Table 3). This suggests thatpain control was satisfactory because some pain levelduring coughing or movement is expected with opi-oid-based analgesia,33 even though the patient’s painat rest was controlled. In the preset study, those pa-tients who acquired SSI had higher morphine con-sumption during the first 24 hours following surgery,but this difference is not statically significant when
724 Vol. 36 No. 10 Levandovski et al
we consider drug consumption for the entire period ofthe first 72 hours after surgery. Although the reason forthis finding is not clear, one hypothetic explanation isthat SSI may stimulate a heightened pain responseand thereby contribute to the continuity of the cycleof pain-morphine consumption and a greater require-ment for opioid analgesic. Alternatively, according toearlier reports, the immunosuppressive effects of mod-erate to large morphine doses administered intrave-nously could explain the present finding.34 Finally, itis possible that the higher incidence of SSI in the pla-cebo group indicates a requirement for more morphineto control pain because of increased anxiety levels.Further investigations are needed to clarify this issue.
Possibly, the higher doses of ropivacaine used in theplacebo-treated patients was also determined by thepatient’s anxiety level because no patients needed toswitch from epidural to general anesthesia because ofinsufficient anesthesia. Thus, as the duration of surgerywas similar between treatment groups, one hypothesisto explain this finding was that placebo-treated pa-tients were more anxious during surgery, which in-duced the anesthetist to use higher doses of localanesthetic according to previous studies.35 This hy-pothesis is supported by previous studies that showedthat the patient’s anxiety was the strongest trigger forprovider-directed medication.35
Although a widely used and validated scoringsystem36 was employed by examiners to establish thediagnosis of SSI directly, there were limitations in con-firming SSI using culture because the gynecologistsometimes had difficulty in collecting the sampleadequately in detriment of the small quantity of puru-lent exudate and because in other cases the drainage ofthe purulent exudate occurred only at 1 or 2 points. Tostandardize the clinical criteria to define the diagnosisof SSI, in addition to using the CDC criteria, the exam-iners could rely on at least 1 of the following criteria:drainage of pus or purulent exudates at least at 1 point,separation of deep tissues with purulent exudates, anddrainage of pus during local anesthesia or in the de-bridement of the wound during general anesthesia.
A global incidence of SSI of 15.2% was found, whichis equivalent to the highest incidence of SSI observed incontaminated surgery in our institution in the last 5years. This rate is more than 4 times the rate observedin potentially contaminated surgery in the same insti-tution in the last 5 years, which oscillated between1.99 and 4.38. Still, this rate is within the limits of stud-ies published in the last decade for patients undergoingabdominal hysterectomy (0%-27%).37,38 However, theincidence of SSI observed in the diazepam-treated pa-tients is in agreement with the incidence of infectionin abdominal hysterectomy reported by the Nosoco-mial Infection National Surveillance Service for British
hospitals, which included more than 60,000 surgicalprocedures.23 On the other hand, the incidence of SSIin placebo-treated patients was excessively high, eventhough the groups were closely matched for psycho-logic and clinical characteristics and despite that allpatients received standardized perioperative care (anti-biotic prophylaxis and anesthetic and analgesic proce-dures). This discrepancy may be explained by severalreasons. First, because the groups were homogenouson baseline and they received the same perioperativemanagement, except for preoperative medication, itis unlikely that factors other than the use of diazepamaccounted for the decreased rate of SSI in diazepam-treated patients. Second, the absence of any anxiolyticmedication prior to and during the surgery in placebo-treated individuals makes management of this groupdifferent from that of other patients treated in ourinstitution because all patients undergo an anesthesiol-ogist’s assessment and they routinely receive benzodi-azepine preoperatively. Furthermore, most of thepatients undergoing surgery in our hospital receivebenzodiazepine during the surgery. Thus, these differ-ences do not permit us to compare the incidence ofSSI in placebo-treated patients with the incidence ob-served in the same hospital prior to this protocol. Addi-tionally, because this study is a clinical trial that wasspecifically designed to assess the effect of preopera-tive medication on SSI, it does allow us to standardizethe perioperative management, the diagnosis protocolof SSI, and the follow-up period. Thus, this may in-crease the possibility of detecting SSI compared withprevious observational studies. Furthermore, we as-sessed the incidence of SSI diagnosed during the hospi-talization period up to the first 30 days after surgery.This may explain in part the higher incidence of SSI de-tected in the present investigation. However, even afterconducting this protocol carefully and adjusting our es-timates as much as possible for the confounding effectof changes in patients and procedure-related charac-teristics over time, some unmeasured, confounding re-sidual effects may persist, but it is unlikely that it wouldchange the direction of the conclusion.
Furthermore, it was noted that hospitalization wasprolonged (by approximately 1 day) in the placebo-treated patients (Table 3). Although the difference inhospital stay between the treatment groups was only1 day, this difference was more than 3 days when wecompared the length of hospitalization of patientswith SSI patients with that of patients without SSI,thus indicating that most infections were substantial.This is an expected finding and similar to the prolonga-tion of hospitalization previously reported in infectedpatients.2,23 Finally, we must also emphasize that weexcluded from this investigation patients with Ameri-can Society of Anesthesiologists status higher than 2,
Levandovski et al December 2008 725
patients with a history of affective disorders or positivescreening for minor psychiatric disorders, and patientstaking any psychotropic medication. It is important topoint out that, although the homogenous populationof this study is methodologically advantageous, the is-sue of external validity arises. That is, hysterectomymay be associated with a specific psychologic and be-havioral state that is not common in other surgical pop-ulations. In addition, it is important to emphasize that,although diazepam was effective in attenuating anxietyand the risk of postoperative SSI with the anesthesiaand analgesia used in this investigation, these findingscannot be generalized to patients undergoing generalanesthesia or some other form of sedation, as well asother types of postoperative analgesia. Thus, the con-clusion of this study may be limited in its generaliza-tion to this surgical population alone.
In conclusion, diazepam-treated patients showedlower postoperative anxiety and lower incidence ofSSI up to 30 days after surgery compared with placeboin patients undergoing abdominal hysterectomy. Al-though the clinical effect of diazepam is undoubtedlyimportant, further studies are needed to elucidate themechanism underlying this response.
The authors thank Dr. A. Leyva for the English editing of the manuscript.
References
1. Bremmelgaard A, Raahave D, Beier-Holgersen R, Pedersen JV, Ander-
sen S, Sorensen AI. Computer-aided surveillance of surgical infections
and identification of risk factors. J Hosp Infect 1989;13:1-18.
2. Kurz A, Sessler DI, Lenhardt R. Perioperative normothermia to
reduce the incidence of surgical-wound infection and shorten hospital-
ization. Study of Wound Infection and Temperature Group. N Engl J
Med 1996;9(334):1209-15.
3. Haley RW, Culver DH, Morgan WM, White JW, Emori TG, Hooton
TM. Identifying patients at high risk of surgical wound infection: a sim-
ple multivariate index of patient susceptibility and wound contamina-
tion. Am J Epidemiol 1985;121:206-15.
4. Miles AA, Miles EM, Burke J. The value and duration of defence
reactions of the skin to the primary lodgement of bacteria. Br J Exp
Pathol 1957;38:79-96.
5. Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR. Guideline
for Prevention of Surgical Site Infection 1999. Centers for Disease
Control and Prevention (CDC) Hospital Infection Control Practices
Advisory Committee. Am J Infect Control 1999;27:97-132.
6. Hughes JM. Study on the efficacy of nosocomial infection control
(SENIC Project): results and implications for the future. Chemother-
apy 1988;34:553-61.
7. Cheadle WG. Risk factors for surgical site infection. Surg Infect 2006;
(Suppl 1)7:7–11.
8. Hartmann M, Jonsson K, Zederfeldt B. Effect of tissue perfusion and
oxygenation on accumulation of collagen in healing wounds: random-
ized study in patients after major abdominal operations. Eur J Surg
1992;158:521-6.
9. Classen DC, Evans RS, Pestotnik SL, Horn SD, Menlove RL, Burke JP.
The timing of prophylactic administration of antibiotics and the risk of
surgical-wound infection. N Engl J Med 1992;326:281-6.
10. Akcxa O, Melischek M, Scheck T, Hellwagner K, Arkilic CF, Kurz A,
et al. Postoperative pain and subcutaneous oxygen tension. Lancet
1999;354:41-2.
11. Wallace LM. Trait anxiety as a predictor of adjustment to and recov-
ery from surgery. Br. J Clin Psychol 1987;26:73-4.
12. Kiecolt-Glaser JK, Page GG, Marucha PT, MacCallum RC, Glaser R.
Psychological influences on surgical recovery: perspectives from psy-
choneuroimmunology. Am Psychol 1998;53:1209-18.
13. Johnston M, Vogele C. Benefits of psychological preparation for sur-
gery: a meta-analysis. Ann Behav Med 1993;15:245-56.
14. Kain ZN, Sevarino F, Pincus S, Alexander GM, Wang SM, Ayoub C,
et al. Attenuation of the preoperative stress response with midazo-
lam: effects on postoperative outcomes. Anesthesiology 2000;93:
141-7.
15. Zhou D, Kusnecov AW, Shurin MR, DePaoli M, Rabin BS. Exposure to
physical and psychological stressors elevates plasma interleukin 6: re-
lationship to the activation of hypothalamic-pituitary-adrenal axis. En-
docrinology 1993;133:2523-30.
16. Derbyshire D, Smith G. Sympathoadrenal responses to anaesthesia
and surgery. Br J Anaesth 1984;56:725-39.
17. Halter J, Pflug A, Porte D. Mechanism of plasma catecholamine in-
creases during surgical stress in man. J Clin Endocrinol Metab 1977;
45:936-44.
18. Hopf HW, Hunt TK, West JM, Blomquist P, Goodson WH III, Jensen
JA, et al. Wound tissue oxygen tension predicts the risk of wound in-
fection in surgical patients. Arch Surg 1997;132:997-1004.
19. Mari JJ, Williams P. A validity study of a psychiatric screening question-
naire (SRQ-20) in primary care in the city of Sao Paulo. Br J Psychiatry
1986;148:23-6.
20. Greif R, Akca O, Horn EP, Kurz A, Sessler DI. Supplemental perioper-
ative oxygen to reduce the incidence of surgical-wound infection.
Outcomes Research Group. N Engl J Med 2000;342:161-7.
21. Collins SL, Moore RA, McQuay HJ. The visual analogue pain intensity
scale: what is moderate pain in millimeters? Pain 1997;72:95-7.
22. Biaggio AMB. A decade of research on State-Trait Anxiety in Brazil. In:
Spielberger CD, Diaz-Guerrero R, editors. Cross-cultural anxiety.
New York: Hemisphere; 1990. p. 157-67.
23. Coello R, Charlett A, Wilson J, Ward V, Pearson A, Borriello P. Ad-
verse impact of surgical site infections in English hospitals. J Hosp In-
fect 2005;60:93-103.
24. Browner WS, Black D, Newman TB, Hulley SB. Estimating sample size
and power. In: Hulley SB, Cummings SR, editors. Designing clinical re-
search. Baltimore: Lippincott Williams & Wilkins; 1988. p. 139-58.
25. Hosmer DW Jr, Lemeshow SM. Applied logistic regression. 2nd ed.
New York: John Wiley & Sons, Inc.; 1989.
26. Kain ZN, Sevarino FB, Rinder C, Pincus S, Alexander GM, Ivy M, et al.
Preoperative anxiolysis and postoperative recovery in women under-
going abdominal hysterectomy. Anesthesiology 2001;94:415-22.
27. Irwin M, Hauger R, Brown M. Central corticotropin-releasing hor-
mone activates the sympathetic nervous system and reduces immune
function: increased responsivity of the aged rat. Endocrinology 1992;
131:1047-53.
28. Zavala F, Taupin V, Descamps-Latscha B. In vivo treatment with benzo-
diazepines inhibits murine phagocyte oxidative metabolism and pro-
duction of interleukin 1, tumor necrosis factor and interleukin-6.
J Pharmacol Exp Ther 1990;255:442-50.
29. Goto Y, O’Malley C, Fanning NF, Wang J, Redmond HP, Shorten GD.
Benzodiazepines inhibit the rate of neutrophil apoptosis. Ir J Med Sci
2003;172:191-4.
30. Stoelting RK. Benzodiazepines. In: Stoelting RK, editor. Pharmacology
& physiology in anaesthetic practice. Philadelphia: Lipincott-Raven;
1999. p. 126-39.
31. Williams JG, Jones JR, Workhoven MN, Williams B. The psychological
control of preoperative anxiety. Psychophysiology 1975;12:50-4.
726 Vol. 36 No. 10 Levandovski et al
32. Biondi M, Zannino LG. Psychological stress, neuroimmunomodulation,
and susceptibility to infectious diseases in animals and man: a review.
Psychother Psychosom 1997;66:3-26.
33. Carr DB, Cousins MJ. Spinal route of analgesia opioids and future
potions. In: Cousins MJ, Bridenbaugh PO, editors. Neural blockade
in clinical anesthesia and management of pain. Philadelphia: Lippin-
cott-Raven; 1998. p. 915-83.
34. Beilin B, Shavit Y, Hart J, Mordashov B, Cohn S, Notti I, et al. Effects of
anesthesia based on large versus small doses of fentanyl on natural
killer cell cytotoxicity in the perioperative period. Anesth Analg
1996;82:492-7.
35. Schupp CJ, Berbaum K, Berbaum M, Lang EV. Pain and anxiety during
interventional radiologic procedures: effect of patients’ state anxiety at
baseline and modulation by nonpharmacologic analgesia adjuncts.
J Vasc Interv Radiol 2005;16:1585-92.
36. Gaynes RP, Culver DH, Horan TC, Edwards JR, Richards C, Tolson JS.
Surgical site infection (SSI) rates in the United States, 1992-1998: the
National Nosocomial Infections Surveillance System basic SSI risk in-
dex. Clin Infect Dis 2001;33:S69-77.
37. Kandula PK, Wenzel RP. Postoperative wound infection after total
adominal hysterectomy: a controlled study of the increased duration
of hospital stay and trends in postoperative wound infection. Am J In-
fect Control 1993;21:201-4.
38. Costa RJ, Krauss-Silva L. Systematic review and meta-analysis of anti-
biotic prophylaxis in abdominal Hysterectomy. Cad Saude Publica
2004;20:S175-89.
