Surgical Site Infections in Orthopedic Surgery: The Effect of Mupirocin Nasal Ointment in a Double-Blind, Randomized, Placebo-Controlled Study

Transcription

1 MAJOR ARTICLE Surgical Site Infections in Orthopedic Surgery: The Effect of Mupirocin Nasal Ointment in a Double-Blind, Randomized, Placebo-Controlled Study M. D. Kalmeijer, 1 H. Coertjens, 2 P. M. van Nieuwland-Bollen, 2 D. Bogaers-Hofman, 2 G. A. J. de Baere, 3 A. Stuurman, 1 A. van Belkum, 4 and J. A. J. W. Kluytmans 2 Departments of 1 Pharmacy, 2 Clinical Microbiology and Infection Prevention, and 3 Orthopedic Surgery, Amphia Hospital, Breda, and 4 Department of Medical Microbiology and Infectious Diseases, Erasmus University Medical Center Rotterdam, The Netherlands The objective of this study was to determine whether use of mupirocin nasal ointment for perioperativeeradication of Staphylococcus aureus nasal carriage is effective in preventing the development of surgical site infections (SSIs). A randomized, double-blind, placebo-controlled design was used. Either mupirocin or placebo nasal ointment was applied twice daily to 614 assessable patients from the day of admission to the hospital until the day of surgery. A total of 315 and 299 patients were randomized to receive mupirocin and placebo, respectively. Eradication of nasal carriage was significantly more effective in the mupirocin (eradication rate, 83.5% versus 27.8%). In the mupirocin, the rate of endogenous S. aureus infections was 5 times lower than in the placebo (0.3% and 1.7%, respectively; relative risk, 0.19; 95% confidence interval, ). Mupirocin nasal ointment did not reduce the SSI rate (by S. aureus) or the duration of hospital stay. Surgical site infections (SSIs) are important complications of orthopedic procedures that involve prosthetic implants. A recent national surveillance study showed that, in The Netherlands, the mean rates of infection in orthopedic surgery that involved artificial implants ranged from 1.5% for total knee implantation to 6.8% for femur head replacement [1]. These infections are associated with discomfort for the patient and with considerable prolongation of hospital stay and increased cost of care [2 7]. When a deep SSI of a pros- Received 21 August 2001; revised 1 February 2002; electronically published 15 July Mupirocin nasal ointment and the ingredients for the placebo nasal ointment were provided by GlaxoSmithKline. Reprints or correspondence: M. D. Kalmeijer, Dept. of Pharmacy, Academic Medical Center, University of Amsterdam, Postbus 22660, 1100 DD Amsterdam, The Netherlands (m.d.kalmeijer@amc.uva.nl). Clinical Infectious Diseases 2002; 35: by the Infectious Diseases Society of America. All rights reserved /2002/ $15.00 thetic implant develops, removal of the prosthesis is often required. Several risk factors for the development of SSIs have been identified. The 1999 guidelines of the Centers for Disease Control and Prevention (CDC) for the prevention of SSI included, for the first time, preoperative nasal carriage of Staphylococcus aureus as a risk factor for SSI [8]. The first studies to have suggested a relationship between nasal carriage of S. aureus and SSI were performed in the late 1950s and early 1960s. These observations were further confirmed by several more recent studies [9]. These studies showed that preoperative nasal carriage of S. aureus is associated with an increased risk of developing an SSI with the same strain. In a study of patients undergoing orthopedic surgery with prosthetic implants, nasal carriage of S. aureus was the most important independent risk factor for developing SSI [10]. Compared with persons who did not have nasal carriage, those with carriage had an RR of developing an SSI of 8.9 (95% CI, ). Preventive S. aureus Eradication CID 2002:35 (15 August) 353

2 These findings led to the hypothesis that perioperative elimination of nasal carriage could lower the infection rate. Mupirocin nasal ointment is highly effective in eliminating S. aureus [11, 12]. In a study involving cardiothoracic surgery [13], perioperative elimination of nasal carriage with this ointment resulted in a significantly lower SSI rate. However, this study used a historical control. In the present randomized, double-blind, placebo-controlled study, we evaluated the effectiveness of perioperative mupirocin nasal ointment in the prevention of SSI in orthopedic surgery with artificial implant material. METHODS Selection of patients. The study was performed from January 1997 through July 1999 at the Department of Orthopedic Surgery at the Ignatius Hospital, Breda, The Netherlands. All patients undergoing elective orthopedic surgery during which prosthetic implant material was used (i.e., hip, knee, or back surgery) were eligible for the study, as were patients undergoing a revision operation of the same type. Every patient was included only once. Patients with an active infection at the moment of inclusion and who had received antibiotic treatment in the previous 24 h were excluded. The ethics committee of the hospital approved the study protocol, and appropriate written informed consent was obtained from all patients. Study medication. Mupirocin nasal ointment (Glaxo- SmithKline; lot /96G04) contains 2.15% weight/weight mupirocin calcium in a soft, white ointment base consisting of paraffin and a mixture of glycerin esters (Softisan 649). The placebo ointment was produced in the hospital pharmacy from paraffin (Bufa; lot 96H16GR) and Softisan (GlaxoSmithKline; lot ), according to the GlaxoSmithKline protocol. Both active and placebo ointment were placed in 5-g tubes and numbered according to an individual randomization list by an independent pharmacist. The appearance of both ointments was identical. Treatment. Generally, patients were admitted to the hospital the day before surgery was performed. On that day, if necessary, depilatory cream was used to remove hair. A nasal swab was done of both nares for culture. Thereafter, therapy with mupirocin or placebo nasal ointment was administered according to the next number on the randomization list. The ointment was applied to both nares twice per day. The first dose of ointment was applied by one of the investigators. Treatment was continued by the nursing staff until the day of surgery. Before surgery, 2 doses of nasal ointment were administered. In persons with nasal carriage, a posttreatment sample was obtained for culture 3 5 days after surgery. Cefamandole was given as perioperative antibiotic prophylaxis to all patients according to the following schedule: 2 g at min before surgery, followed by 2 more doses of 1 g each at 8 h and 16 h after surgery. If patients were allergic to cephalosporins, clindamycin was given according to the following schedule: 600 mg at min before surgery, followed by 2 more doses of 600 mg each at 8 h and 16 h after surgery. Nasal culture. Swabs of both nares were obtained with a Dacron swab. The swab was inoculated on a blood agar plate and on a mannitol salt agar plate. The material was spread into quadrants with a sterile loop [14]. The amount of growth was scored semiquantitatively. Growth of S. aureus limited to the first sector on the blood agar plate was scored as level 1 colonization. Growth up to the second, third, and fourth sector were scored as level 2, 3, and 4 colonization, respectively. The S. aureus strains isolated from nose swabs and wound specimens were typed by means of pulsed-field gel electrophoresis (PFGE) according to a protocol that has been described elsewhere [15]. DNA was embedded in agarose blocks and digested with the restriction enzyme SmaI (Boehringer Mannheim). PFGE was performed in 1% Seakem agarose gels (FMC Bioproducts) buffered in 0.5 Tris-borate-EDTA at 14 C. Electrophoresis (BioRad CHEF Mapper) was performed for 22 h (ramping, s) at an angle of 120 at 6 V/cm. Banding patterns were scored and interpreted according to the criteria of Tenover et al. [16]. Variables and infection surveillance. The following variables were recorded: age, sex, body mass index (weight/ [length] 2 ; in kg/m 2 ), kind of procedure performed, date of the admission to the hospital, date of surgery, date of discharge, total length of hospital stay, postoperative length of hospital stay, duration of the operation, surgeon (numbered 1 to 7), underlying disease (malignancy or diabetes mellitus), use of immunosuppressive drugs, American Society of Anesthesiologists score, method of anesthesia (general or spinal/ epidural), level of growth for the first nasal swab, and level of growth for the posttreatment nasal swab. The medical records of all patients were studied for the development of SSI on the bases of criteria of the CDC [17]. The development of SSI was observed until 1 month after surgery. After discharge from the hospital, follow-up studies were performed over the phone by use of a standardized questionnaire that was based on CDC criteria. Patients were asked whether they had experienced any of the following symptoms: purulent drainage from the incision, pain and tenderness, and redness or heat; in addition, they were asked whether an antibiotic had been prescribed and whether a specimen of the wound had been obtained for culture. If one of the answers indicated an infection, the patient was seen at the outpatient department by an orthopedic surgeon for further evaluation. We checked all patients to see whether they were readmitted to the hospital and whether this readmission was for wound infection. For SSI, the date of onset and pathogens involved were recorded. 354 CID 2002:35 (15 August) Kalmeijer et al.

3 Table 1. Baseline characteristics of 614 patients assessed for surgical site infections after orthopedic surgery performed with artificial implant material. Characteristic Mupirocin (n p 315) Placebo (n p 299) Female sex 211 (67.0) 196 (65.6) Age, years Height, m Weight, kg Body mass index, kg/m Surgical procedure Total hip 128 (40.6) 133 (44.5) Revision total hip 26 (8.3) 21 (7.0) Total knee 97 (30.8) 81 (27.1) Revision total knee 4 (1.3) 8 (2.7) Back 50 (15.9) 50 (16.7) Other 10 (3.2) 6 (2.0) Duration of surgery, min Underlying disease Malignancy 2 (0.6) 1 (0.3) Insulin-dependent diabetes 3 (1.0) 1 (0.3) Immunosuppression 5 (1.6) 6 (2.0) ASA score, % of patients NOTE. Data are no. (%) of patients or mean value SD, unless other- wise indicated. ASA, American Society of Anesthesiologists. Of the 627 included patients, 13 could not be evaluated. Seven of these 13 patients did not undergo surgery, and 6 patients had to be excluded for the following reasons: 2 patients had an active infection that was detected after inclusion, 2 patients were included for the second time, and, in 2 patients, the respective study medications were possibly mixed up. None of these patients developed an SSI. The remaining 614 patients were included in the intentto-treat analysis. Of these 614 patients, 43 (7.0%) underwent surgery that did not involve the use of prosthetic material, which resulted in 571 patients (93.0%) fulfilling the inclusion criteria. In the intent-to-treat analysis, 315 patients were randomized to the mupirocin and 299 to the placebo. Both s were comparable with respect to sex, age, body mass index, type of surgical procedure, duration of surgery, underlying disease, immunosuppressive therapy received, American Society of Anesthesiologists score, and surgeon (table 1). Also, no differences were found in the method of anesthesia and the distribution of operations among the surgeons (data not shown). All patients received 2 doses of nasal ointment before undergoing surgery. The mupirocin and placebo s were also comparable with respect to the preoperative nasal carriage rates: 30.3% in the mupirocin and 28.8% in the placebo had nasal carriage of S. aureus (table 2). After treatment, 15 of the 95 patients who were initially carrying S. aureus in the mupirocin and 61 of 86 in the placebo were still carrying S. aureus ( P!.05). In 12 patients, no nasal swab was taken to Outcome. The primary outcome was the SSI rate. Secondary outcomes were the rate of SSIs due to S. aureus, the rate of endogenous S. aureus SSIs, and the duration of hospital stay. An endogenous SSI was defined as an SSI with a S. aureus isolate that was identical on PFGE analysis to the nasal isolate of that patient. Statistical analysis. The sample size was calculated by using the results of a previous surveillance study conducted in this department [10]. On the basis of an incidence of SSI of 6.6%, a reduction of 75%, and a power of 80%, with a significance level of 95%, 300 patients were needed for each treatment. An intent-to-treat analysis was performed. Results were analyzed using SPSS software (SPSS). Differences between s were tested by use of Student s t test, Fisher s exact test, or Wilcoxon rank sum test, as appropriate. P!.05 was considered to be statistically significant. RESULTS During the study period, 692 patients were eligible for inclusion in the study; 65 patients (9.4%) were unwilling to participate. Table 2. Baseline and posttreatment rates of nasal carriage of Staphylococcus aureus in 614 patients assessed for surgical site infections after orthopedic surgery performed with artificial implant material. Carriage Mupirocin (n p 315) Placebo (n p 299) Nasal carriage at baseline, no. (%) of patients 95 (30.3) a 86 (28.8) Level of nasal carriage at baseline, % of patients b Nasal carriage rate after treatment, c no. (%) of patients 15 (16.5) d 61 (78.2) e a No initial culture results were available for 1 patient. b See Methods. c Only patients who had nasal carriage of S. aureus before treatment are included. d The results of posttreatment cultures were not available for 4 patients. e The results of posttreatment cultures were not available for 8 patients. Preventive S. aureus Eradication CID 2002:35 (15 August) 355

4 control for effectiveness of treatment. None of these patients developed an SSI or an SSI with S. aureus. Typing with PFGE revealed that, of the posttreatment cultures that yielded S. aureus in the placebo, 84% contained the original S. aureus strain. In the mupirocin, this rate was 80%. Almost all patients (99.3%) received cefamandole as antibiotic prophylaxis. Three patients received clindamycin, and 1 patient did not receive perioperative prophylaxis. This patient did not develop an SSI. The SSI rates in the mupirocin and placebo s were comparable (3.8% and 4.6%, respectively; table 3). Also the rates of SSI caused by S. aureus were not significantly different (mupirocin, 1.6%; placebo, 2.7%). Overall, 13 infections with S. aureus occurred (5 infections in the mupirocin and 8 in the placebo ). Of the 5 patients in the mupirocin infected with S. aureus, 2 had nasal carriage (figure 1). On the basis of the outcome of the PFGE analysis of pre- and posttreatment S. aureus isolates, it was concluded that 1 of these persons with nasal carriage developed an endogenous infection. Of the 8 patients infected with S. aureus in the placebo, 5 had nasal carriage. These 5 persons with nasal carriage all developed an endogenous infection. Endogenous S. aureus infections were 5 times less likely to occur in the mupirocin, but this difference was not statistically significant (RR, 0.19; 95% CI, ). No statistically significant differences were found between different types of surgery (data not shown). Each of the 7 exogenous strains was unique. All S. aureus strains in the study were susceptible to cefamandole and clindamycin, as was expected, because methicillin-resistant S. aureus is not endemic in The Netherlands. Also, all strains were susceptible to mupirocin. The mean ( SD) duration of hospital stay was comparable in both s: days in the placebo and days in the mupirocin. Five patients were readmitted to the hospital: 2 patients in the mupirocin (for repositioning) and 3 patients in the placebo (1 for Table 3. Surgical site infection (SSI) rates for 614 patients assessed for SSIs after orthopedic surgery performed with artificial implant material. Infection No. (%) of patients Mupirocin (n p 315) Placebo (n p 299) RR (95% CI) SSI 12 (3.8) 14 (4.7) Deep 0 (0) 1 (0.3) Superficial 12 (3.8) 13 (4.3) 0.81 ( ) Staphylococcus aureus SSI 5 (1.6) 8 (2.7) 0.59 ( ) Endogenous S. aureus SSI 1 (0.3) 5 (1.7) 0.19 ( ) removing the prosthesis as a result of a deep infection and 2 for repositioning). In addition to the intent-to-treat analysis, an analysis was performed of the 571 patients who complied with the inclusion criteria. The results of this analysis were not significantly different from the results of the intent-to-treat analysis. Also, all patients not included in the study during the study period underwent infection surveillance. No higher infection rate was found in this of patients. DISCUSSION The initial nasal carriage rate of S. aureus of 30% found in this study population falls within the range of nasal carriage rates reported in other studies [9]. Treatment resulted in an elimination rate of 83.5% among mupirocin recipients and 21.8% among placebo recipients. In an analysis of 6 clinical trials, Doebbeling et al. [11] reported a range of elimination rates of 68% 100% (mean, 91%) for patients who received mupirocin twice per day for 5 days, and they reported a range of elimination rates of 1% 13% (mean, 6%) for patients who received the same regimen with placebo. Compared with these results, the short treatment course in the present study was relatively effective. Because placebo is not effective in vitro against S. aureus, the reduction in the placebo is most likely a naturally occurring phenomenon that is the result of intermittent carriage, not an effect of placebo cream on S. aureus [9]. As can be concluded from table 2, the effectiveness of mupirocin, on the basis of the elimination of the original strain, was 86.8%, compared with 34.6% for the placebo ointment ( P!.05). This perioperative eradication, however, did not result in a significant difference in primary outcome, the overall rate of SSI. Also, in the secondary outcomes (rate of SSI due to S. aureus, rate of endogenous S. aureus infection, and the duration of hospital stay), no significant differences were found, although all were lower in the mupirocin. Therefore, the conclusion could be that perioperative treatment with mupirocin nasal ointment is not effective in SSI prevention in orthopedic surgery with artificial implant material. This is not in accordance with the results of earlier intervention studies of nasally administered mupirocin. In another study involving orthopedic surgery, the rate of SSI was reduced significantly, from 2.7% to 1.3%. The rate of S. aureus SSI decreased nonsignificantly, from 1.1% to 0.7% [18]. In a study involving cardiothoracic surgery, the SSI rate was significantly reduced, from 7.3% to 2.8% [13]. In both studies, a historical control was used. It is well known that the use of historical control s may lead to false conclusions. This may explain the discrepancy with our present data. 356 CID 2002:35 (15 August) Kalmeijer et al.

5 Figure 1. Results of pulsed-field gel electrophoresis typing of Staphylococcus aureus strains The preliminary analysis of a prospective, randomized, placebo-controlled clinical trial in cardiothoracic, general, and neurological procedures shows that application of nasal mupirocin is associated with a significant decrease in the risk of nosocomial S. aureus infection [19]. Several aspects should be taken into consideration when interpreting the results of the present study. First, several observations indicate an effect of mupirocin. The rate of SSI caused by S. aureus was 2.7% in the placebo and 1.6% in the mupirocin, and the rate of endogenous S. aureus infection was 5 times lower in the mupirocin. Also, the RR of nasal carriage for the development of SSI was 2.6 (95% CI, ) in the placebo and 1.2 (95% CI, ) in the mupirocin. These differences were, however, not statistically significant. Colonization with S. aureus was found to be an independent risk factor for developing an SSI due to S. aureus [10]. On the bases of these results, we expected eradication of nasal carriage with mupirocin to lower the infection rate and, especially, to affect the rate of deep infection. In the present study, we found superficial SSI rates of 3.8% in the intervention and 4.3% in the placebo, which is comparable with the previous data (4.0%) [10]. Deep SSIs, however, almost disappeared: only 1 deep SSI was found in the placebo (0.3%), compared with 2.6% in the surveillance study ( P!.05). This unexpectedly strong decrease made it impossible to measure a difference between s with the current number of patients. It is well known that the pathogenesis of SSI and thereby the prevention of SSI depends on many different factors. It has been reported several times that SSI surveillance and feedback of the results lead to lower SSI rates [20 25]. If surgical personnel and nursing staff are informed about (high) SSI rates, they will improve infection-control measures as well as surgical technique. The relative contribution of the individual measures to the decrease of SSIs is unknown. This intervention study should be considered as an intensive means of surveillance with a strong effect on the deep SSI rate. No specific measures had been taken to prevent SSI, and the surgical team was the same during the entire period. In many hospitals, SSIs remain important complications in clean surgical procedures. Nasal carriage of S. aureus is now definitely considered to be a risk factor for the development of SSI. One study showed that nasal carriage also plays an important role in the development of S. aureus bacteremia [26]. The role of perioperative eradication of nasal carriage, however, remains unclear. Studies with historical control s have found that mupirocin has a statistically significant effect [13, 18]. However, in general, the use of historical control s is not accepted as a reliable method of research. The results of this study emphasize the pitfalls of this methodology. The significant decrease in the rate of deep SSI in the placebo compared with the control subjects of a previous surveillance study would have led to the unjustified conclusion that mupirocin is effective. Therefore, a randomized, doubleblind, placebo-controlled study was performed. In this trial, the intervention did not result in a statistically significant reduction in SSI rate. However, this does not necessary mean that perioperative eradication of nasal carriage is useless. The unexpected decrease in the SSI rate observed in the placebo resulted in insufficient power of the study, and, therefore, conclusions regarding the efficacy of mupirocin cannot be drawn. Although this study did not answer its initial question, 2 important lessons can be learned. First, paying serious attention to SSI by performing active surveillance in this case, through performing an intervention study is a highly effective way to prevent SSI. Considering the near disappearance of deep SSI, additional measures, such as the use of mupirocin nasal ointment, may have no additional effect. Second, future studies should seriously consider the effect of the study itself on the SSI rate in their power calculations. This effect may even be stronger than the effect of the intervention studied. Preventive S. aureus Eradication CID 2002:35 (15 August) 357

6 References 1. Geubbels ELPE, van Dieten HEM, Mintjes-de Groot AJ, et al. PREZIES: prevention of hospital infections by surveillance. Component surgical site infections II, Report Bilthoven, The Netherlands: Rijksinstituut voor Volksgezondheid en Milieu, Utrecht, Kwaliteitsinstituut voor de Gezondheidszorg CBO, March Green JW, Wenzel RP. Postoperative wound infection: a controlled study of the increased duration of hospital stay and direct cost of hospitalization. Ann Surg 1977; 185: Coello R, Glenister H, Fereres J, et al. The cost of infection in surgical patients: a case-control study. J Hosp Infect 1993; 25: Kluytmans JAJW, Mouton JW, Jzerman EPF, et al. Nasal carriage of S. aureus as a major risk factor for wound infection after cardiac surgery. J Infect Dis 1995; 171: Ehrenkranz NJ, Meakins JL. Surgical infections. In: Bennet JV, Brachman PS, eds. Hospital infections. 3rd ed. Boston: Little, Brown, 1992: Kluytmans JAJW, Mouton JW, Maat APWM, et al. Surveillance of postoperative infections in thoracic surgery. J Hosp Infect 1994; 27: Petty W. Infections of skeletal prostheses. In: Bennet JV, Brachman PS, eds. Hospital infections. 3rd ed. Boston: Little, Brown, 1992: Mangram AJ, Horan TC, Pearson ML, et al. Guideline for prevention of surgical site infection, Hospital Infection Control Practices Advisory Committee. Infect Control Hosp Epidemiol 1999; 20: Kluytmans J, Van Belkum A, Verbrugh H. Nasal carriage of Staphylococcus aureus: epidemiology, underlying mechanisms, and associated risks. Clin Microbiol Rev 1997; 10: Kalmeijer MD, van Nieuwland-Bollen E, Bogaers-Hofman D, de Baere GA. Nasal carriage of Staphylococcus aureus is a major risk factor for surgical-site infections in orthopedic surgery. Infect Control Hosp Epidemiol 2000; 21: Doebbeling BN, Breneman DL, Neu HC, et al. Elimination of S. aureus nasal carriage in health care workers: analysis of six clinical trials with calcium mupirocin ointment. Clin Infect Dis 1993; 17: Fernandez C, Gaspar C, Torrellas A, et al. A double-blind, randomised, placebo-controlled clinical trial to evaluate the safety and efficacy of mupirocin calcium ointment for eliminating nasal carriage of S. aureus among hospital personnel. J Antimicrob Chemother 1995; 35: Kluytmans JAJW, Mouton JW, VandenBerg MFQ, et al. Reduction of surgical site infections in cardiothoracic surgery by elimination of nasal carriage of S. aureus. Infect Control Hosp Epidemiol 1996; 17: Koneman EW, Allen SD, Janda WM, et al. Color atlas and textbook of diagnostic microbiology. Philadelphia: Lippincott-Raven, VandenBergh MFQ, Yzerman EPF, van Belkum A, et al. Follow-up of Staphylococcus aureus nasal carriage after 8 years: redefining the persistent carrier state. J Clin Microbiol 1999; 37: Tenover F, Arbeit RD, Goering RV, et al. Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. J Clin Microbiol 1995; 33: Horan TC, Gaynes RP, Martone WJ, et al. CDC definitions of nosocomial surgical site infections, Infect Control Hosp Epidemiol 1992; 13: Gernaat-van der Sluis AJ, Hoogenboom-Verdegaal AMM, Edixhoven PJ, et al. Prophylactic mupirocin could reduce orthopedic wound infections. Acta Orthop Scand 1998; 69: Perl TM, Cullen JJ, Wenzel RP, et al. Intranasal mupirocin to prevent postoperative Staphyloccus aureus infections. N Engl J Med 2002; 346: Cruse PJE, Foord R. The epidemiology of wound infections: a 10-year prospective study of 62,939 wounds. Surg Clin North Am 1980; 60: Olson MM, Lee JT Jr. Continuous 10-year wound infections surveillance: results, advantages and unanswered questions. Arch Surg 1990; 125: Condon RE, Schulte WJ, Malangoni MA, et al. Effectiveness of surgical wound infection surveillance program. Arch Surg 1983; 118: Olson M, O Conner M, Schwartz ML. Surgical wound infections: a 5- year prospective study of 20,193 wounds at the Minneapolis VA medical center. Ann Surg 1984; 199: Mead PB, Pories SE, Hall P, et al. Decreasing the incidence of surgical wound infections: validation of a surveillance notification program. Arch Surg 1986; 121: Haley RW, Culver DH, White JW, et al. The efficacy of infection surveillance and control programs in preventing nosocomial infections in US hospitals. Am J Epidemiol 1985; 121: von Eiff C, Becker K, Machka K, et al. Nasal carriage as a source of Staphylococcus aureus bacteremia. Study Group. N Engl J Med 2001; 344: CID 2002:35 (15 August) Kalmeijer et al.