Reduction of Surgical Site Infection Rates Associated With Active Surveillance
Objective. To evaluate whether surgical site infection (SSI) rates decrease in surgical departments as a result of performing active SSI surveillance.
Design. Retrospective multiple logistic regression analyses.
Setting. A group of 130 surgical departments of German hospitals participating in the Krankenhaus Infektions Surveillance System (KISS).
Methods. Data for 19 categories of operative procedures performed between January 1997 and June 2004 were included (119,114 operations). Active SSI surveillance was performed according to National Nosocomial Infections Surveillance system (NNIS) methods and definitions. Departments' SSI rates were calculated individually for each year of surveillance and for each operative procedure category, taking into account when the individual departments had begun their surveillance activities. Multiple logistic regression analyses on a single operation basis were carried out with stepwise variable selection to predict outcomes for patients with SSI. The variables included were as follows: the department's year of participation, NNIS risk index variables, patients' age and sex, and the hospitals' structural characteristics, such as yearly operation frequency, number of beds, and academic status.
Results. For 14 of 19 operative procedure categories analyzed, there was a tendency toward lower SSI rates that was associated with increasing duration of SSI surveillance. In multiple logistic regression analyses of pooled data for all operative procedures, the departments’ participation in the surveillance system was a significant independent protective factor. Compared with the surveillance year 1, the SSI risk decreased in year 2 (odds ratio, 0.84; 95% confidence interval, 0.77‐0.93) and in year 3 (odds ratio, 0.75; 95% confidence interval, 0.68‐0.82), and there was no change in year 4.
Conclusion. The SSI incidence was reduced by one quarter as a result of the surveillance‐induced infection control efforts, which indicates the usefulness of a voluntary surveillance system.
Received August 8, 2005; accepted November 10, 2005; electronically published November 21, 2006.
Surgical site infections (SSIs) are a significant public health problem and have a major impact on the cost of health care.1‐3 Surveillance of nosocomial infections is undertaken to reduce the burden of such infections by identifying infection problems in hospitals and encouraging infection control practices.4‐8 Clear evidence of a successful reduction in the infection rate may be difficult to produce. First, on the level of the individual hospital, the absolute number of SSIs per year in a single operative procedure category may be low; therefore, it is difficult to prove the advantageous effects of infection control measures. Second, at the level of multiple hospitals, the effect of regression to the mean as a possible confounder should be taken into account: hospitals with relatively high infection rates in the first surveillance period are likely to have lower rates by chance in the next surveillance period, even in the absence of any intervention. Analysis of pooled data from a surveillance system that includes all participants, regression to the mean is less likely than in analysis of data from hospitals with initially high infection rates. The aim of this study was to evaluate whether SSI rates decrease in surgical departments as a result of performance of active SSI surveillance.
Methods
Study Hospitals
All study participants were voluntary participants in the German Krankenhaus Infektions Surveillance System (KISS). The Centers for Disease Control and Prevention definitions for SSI and risk stratification are applied according to National Nosocomial Infections Surveillance system (NNIS) guidelines.9,10 Postdischarge surveillance is strongly recommended, although a “gold standard” method is not available. Participation in KISS is also explicitly recommended by the German state authorities. Data are handled confidentially and rendered anonymously before being published by the national reference center.11,12 Most of the participating hospitals perform ongoing surveillance. Data were included from surgical departments that had at least 4 years of participation in KISS and at least 30 operations per year registered for the operative procedure category in question. The surgical departments included started participating in KISS between 1997 and 2000; the operations were performed between January 1997 and June 2004.
Statistical Analyses
We decided to evaluate the effect of surveillance by analyzing the pooled data of all the participants together, rather than by analyzing data from individual hospitals. First, for a univariate analysis, the SSI rates of all participating departments were calculated for each year of surveillance and for each operative procedure category, taking into account the beginning of participation by each department. Second, for a multivariate analysis based on all the operations performed, the variables available for KISS were considered to be possible confounders. These variables were patient sex and age, the NNIS risk index categories (American Society of Anesthesiologists score, wound class, and duration of operation), fully endoscopic performance of operations, the departments’ yearly frequency of the operative procedure in question (obtained from the KISS database), and the number of beds in the hospital and its academic status (which were obtained from the participants through a questionnaire). For each operative procedure category, we performed separate multiple logistic regression analyses based on the level of operation with stepwise variable selection (forward and backward alternately) to predict the outcome with respect to SSI (ie, SSI occurred or did not occur). The significance level for the stepwise selection procedures was .05, taking into account the large sample sizes.
One distinct model was developed in which data for all operative procedures were included. For this model, the surgical specialization (eg, abdominal surgery or obstetrics) was included as a variable. Developed logistic regression model equations were used to determine the odds of acquiring SSI, depending on the patients’ risk factors. Each regression coefficient could be interpreted as an adjusted odds ratio (OR) for acquiring SSI, if such risk factors were present; this is why this method only generates significant odds ratios at the .05 level. All analyses were performed using SAS statistical software for Windows, version 6.12 (SAS Institute).
Results
The criteria for inclusion of a hospital in this study (4 years of participation, with each department performing at least 30 operations per procedure category) were met by 130 departments from 86 hospitals (from a total of 285 departments at 150 hospitals that participated in the SSI module). The pool of data analyzed consisted of 191,114 operations and 3,241 SSIs. The 75th percentile for the duration of the operations (in minutes) was taken as the cutoff for the NNIS risk index calculation. The cutoff times were the same for the data included in this study as those used for the complete KISS data. In the raw data stratified according to the surveillance year, 14 of 19 operative procedures showed a tendency toward lower SSI rates that was associated with increasing duration of SSI surveillance (Table 1).
Logistic regression models were developed for each procedure category separately, and 1 model for the pooled data (Table 2). These models were able to control for possible changes in the patients’ SSI risk or in the hospitals’ characteristics over time: each regression coefficient could be interpreted as an adjusted OR for acquiring SSI, if such factors were present. Among the variables included in the model, the NNIS risk index factors of age and sex were the most significant. Risk values for variables that described the structure of the surgical department (number of operations performed in any one category, number of beds in the hospital, and hospital’s academic status) were, for some operative procedures, significantly different from 1. A tendency toward a higher SSI risk in hospitals that had more than the mean number of beds (370) was observed for some procedures and was also observed in analysis of pooled data.
In multivariate analysis, the model for all procedures yielded the participation of the departments in the surveillance system as a significant independent protective factor. The ORs for SSI, compared with surveillance year 1, were as follows: for year 2, it was 0.84 (95% confidence interval [CI], 0.77‐0.93); for year 3, it was 0.75 (95% CI, 0.68‐0.82); and for year 4, it was 0.75 (95% CI, 0.68‐0.83). Some procedure‐specific models showed significant protection against SSI: for coronary artery bypass from year 2 on, and for cesarean section and hip prosthesis arthroplasty from year 3 on.
The median time between surgery and recording of an infection was 9 days for the pooled data on all operative procedures, ranging from 6 days (for cesarean section) to 14 days (for hip prosthesis arthroplasty). No differences in these intervals from surveillance year 1 to year 4 were observed.
Discussion
The effectiveness of SSI surveillance was demonstrated in a group of hospitals that had long‐term experience (4 years or more) with SSI surveillance. A considerable reduction of 25% in the SSI rate was observed after 2 years of surveillance. The trend toward reduction was observed for most, but not all, of the operative procedures. Statistical significance at the .05 level was attained for pooled data and for operative procedures frequently performed, such as cesarean section, hip prosthesis arthroplasty, and coronary artery bypass. The reduction in the SSI rates for nephrectomy and prostatectomy was impressive but not statistically significant because of low sample sizes.
Although this study pooled data from the surveillance system, its power to detect reductions in outcomes that have a low prevalence, , such as SSI after clean surgery, was limited. Hence, no further reduction in SSI rates was shown for operative procedures that had low initial SSI rates (arthroscopic knee surgery, thyroidectomy, laparoscopic cholecystectomy, knee prosthesis arthroplasty, and herniorrhaphy).
The higher the SSI incidence, the better the power of our approach for detection of differences over time. However, even for operative procedures with a relatively high incidence of SSI, such as colon surgery, the power may not be sufficient for detecting a reduction in the SSI rate: for example, if the rate is 5.3 SSIs per 100 operations and a reduction by 20% is considered as worthy to be detected, 6,350 operations would be required in each group to obtain a power of 0.8 (
). However, for some operative procedures with high initial SSI rates, such as appendectomy, colon surgery, and vascular surgery, no tendency toward a reduction in the SSI rate was observed. This fact cannot be explained with the KISS data and requires further investigation.
We believe that the reduction in SSI rates demonstrated with surveillance reflects a real reduction of SSIs. First, by performing multivariate analyses, we checked a number of possible confounding factors that might change over time, such as patients’ characteristics (ie, NNIS risk index) and the structural characteristics of the participating KISS hospitals. Second, participation in KISS is voluntary and anonymous, so hospitals have no motivation to deliver false or invalid data to the surveillance system. The reduction in recorded SSIs observed within our system is comparable to the results reported by the nosocomial infection surveillance network system in The Netherlands (Preventie van Ziekenhuisinfecties door Surveillance; PREZIES), which showed an overall reduction of one‐third from surveillance year 1 to surveillance year 4.13
Some limitations of this study may have influenced the results. A reduction in the number of SSIs recorded may have been the result of changes in healthcare practices that were unrelated to surveillance activities. This was not investigated in our study. The surveillance intensity in the participating hospitals has not been validated, and decreasing SSI rates may have been the result of less‐sensitive surveillance. An argument against this suspicion is that, with the increasing experience of the personnel performing active surveillance, the quality of the surveillance results may have become even better. As in all western countries, in Germany there is a tendency toward shorter hospital stays, especially after surgery. With shorter stays in hospital, it is conceivable that surveillance concentrated on the hospital stay overestimates the effects of surveillance. Postdischarge surveillance for all patients would be useful to obtain more‐accurate data.14‐16 Postdischarge surveillance is strongly recommended in KISS but is not mandatory, because, as in other countries, systematic postdischarge surveillance is not yet feasible in Germany. The number of infections detected by postdischarge surveillance is increasing in our system from year to year, and all these infections are included in this analysis. The stable time to infection we demonstrated in this study may possibly indicate that efforts to report postdischarge infections may have compensated for the shorter postoperative hospital stay. Another limitation of this study is that the individual patients’ hospital discharge dates were not yet systematically recorded by our surveillance system during the study period reported on here. This is why the influence of shorter hospital stays on the surveillance intensity, which might be compensated for by better postdischarge surveillance, cannot be quantified. We conclude that the reduction of one quarter in the number of SSIs shown in this study represents a real reduction in the incidence of SSI and is the result of surveillance‐induced infection control efforts in the participating surgical departments.
References
- 1. Coello R, Charlett A, Wilson J, Ward V, Pearson A, Borriello P. Adverse impact of surgical site infections in English hospitals. J Hosp Infect 2005; 60:93‐103.
- 2. Kirkland KB, Briggs JP, Trivette SL, Wilkinson WE, Sexton DJ. The impact of surgical‐site infections in the 1990s: attributable mortality, excess length of hospitalization, and extra costs. Infect Control Hosp Epidemiol 1999; 20:725‐730.
- 3. Merle V, Germain JM, Chamouni P, et al. Assessment of prolonged hospital stay attributable to surgical site infections using appropriateness evaluation protocol. Am J Infect Control 2000; 28:109‐115.
- 4. 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; 121:182‐205.
- 5. Harbarth S, Sax H, Gastmeier P. The preventable proportion of nosocomial infections: an overview of published reports. J Hosp Infect 2003; 54:258‐266.
- 6. Geubbels EL, Bakker HG, Houtman P, et al. Promoting quality through surveillance of surgical site infections: five prevention success stories. Am J Infect Control 2004; 32:424‐430.
- 7. Schneeberger PM, Smits MH, Zick RE, Wille JC. Surveillance as a starting point to reduce surgical‐site infection rates in elective orthopaedic surgery. J Hosp Infect 2002; 51:179‐184.
- 8. Condon RE, Schulte WJ, Malangoni MA, Anderson‐Teschendorf MJ. Effectiveness of a surgical wound surveillance program. Arch Surg 1983; 118:303‐307.
- 9. Horan TC, Emori TG. Definitions of key terms used in the NNIS System. Am J Infect Control 1997; 25:112‐116.
- 10. Brandt C, Hansen S, Sohr D, Daschner F, Rüden H, Gastmeier P. Finding a method for optimizing risk adjustment when comparing surgical‐site infection rates. Infect Control Hosp Epidemiol 2004; 25:313‐318.
- 11. Nosokomiale Infektionen: KISS‐Daten zu postoperativen Wundinfektionen. Epidemiol Bull 2003; 36:290‐292.
- 12. Nationales Referenzzentrumfür Surveillance von nosokomialen Infektionen Web site. Available at: http://www.nrz‐hygiene.de/english.htm. Accessed November 10, 2006.
- 13. Geubbels EL. Prevention of Surgical Site Infections Through Surveillance. Utrecht, the Netherlands: Universiteit Utrecht; 2002:151.
- 14. Byrne DJ, Lynch W, Napier A, Davey P, Malek M, Cuschieri A. Wound infection rates: the importance of definition and post‐discharge wound surveillance. J Hosp Infect 1994; 26:37‐43.
- 15. Sands K, Vineyard G, Platt R. Surgical site infections occurring after hospital discharge. J Infect Dis 1996; 173:963‐970.
- 16. Mitchell DH, Swift G, Gilbert GL. Surgical wound infection surveillance: the importance of infections that develop after hospital discharge. Aust N Z J Surg 1999; 69:117‐120.