Cesarean delivery is one of the most common surgical procedures performed in the United States. Approximately 1.37 million cesarean deliveries were performed in 2007, accounting for 31.8% of all U.S. births.1 Infectious morbidity, consisting primarily of endometritis and wound infection,2 occurs in 5–10% of cesarean deliveries.3–5 The risk of postpartum infection is nearly five-fold higher after cesarean delivery compared with vaginal births.6,7 With the continued rise in the cesarean delivery rate, postpartum infectious morbidity will only become a more considerable problem unless more aggressive and systematic measures are taken to reduce this preventable morbidity.
As the skin is a major source of pathogens, skin antisepsis can contribute greatly to the reduction of postoperative infectious morbidity. Recent studies have tried to determine which antiseptic surgical site preparation agents—including iodophors, alcohol-containing products, or chlorhexidine gluconate8—might be most effective in decreasing postoperative infectious morbidity.7,9–13 However, the most effective and efficient strategy to achieve skin antisepsis remains controversial: the Centers for Disease Control and Prevention has not issued recommendations regarding which antiseptic should be used preoperatively to prevent postoperative infection,8 and a 2004 Cochrane Review summarized that further research is required to answer this question appropriately.14
Traditionally, a single povidone-iodine agent (10% solution paint) has been the antiseptic surgical site preparation agent of choice at our institution. Povidone-iodine exerts a bacteriostatic effect through an oxidation mechanism while present on the skin. It has a broad spectrum of activity and provides coverage against Gram-positive bacteria that colonize the skin.8 Use of this skin preparation agent results in postoperative reduction of indigenous skin microflora compared with the prescrub level in both postscrub and postoperative bacterial cultures.15 In contrast, the protocol at some local hospitals is to administer two povidone-iodine agents (scrub and paint) in the operating room immediately before cesarean delivery in an effort to obtain skin antisepsis and to prevent postoperative infectious complications. There are no data to suggest that using two forms of the same antiseptic agent decreases postsurgical infectious complications compared with a single agent alone. However, anecdotally, the rate of postoperative infectious complications appears lower at these local institutions despite treating similar patient populations.
To this end, the objective of our study was to estimate whether use of two povidone-iodine agents (a 13% solution scrub followed by a 10% solution paint) was associated with a decreased rate of postcesarean delivery infectious complications compared with one povidone-iodine agent (10% solution paint) alone. We hypothesized that the addition of povidone-iodine scrub, before entering the operating room, to the standard paint in the operating room would be associated with a decreased rate of postcesarean delivery infectious complications compared with standard paint alone.
MATERIALS AND METHODS
We performed a retrospective cohort study of all women who underwent cesarean delivery over 27 months at a single urban academic medical center. This study (protocol #811021) was approved by the institutional review board (#4) at the University of Pennsylvania.
Before November 2008, all patients who underwent cesarean delivery at our institution had preoperative skin preparation with povidone-iodine topical paint (10% solution [PAINT]) in the operating room immediately before skin incision. In November 2008, a new preoperative skin preparation protocol was initiated as a quality improvement initiative by the Obstetrics Clinical Effectiveness and Quality Improvement committee at the Hospital of the University of Pennsylvania. The Obstetrics Clinical Effectiveness and Quality Improvement committee was established to maximize the quality of patient care. Interventions aimed at the reduction of infectious complications were recognized as an important initial focus of the committee, and it was suggested that modification of the current cesarean delivery skin preparation might successfully reduce postoperative infectious complications. The addition of the povidone-iodine topical scrub was the only change made to the cesarean delivery skin preparation protocol, and the protocol was the only change implemented by the committee to reduce postoperative infectious complications.
The protocol called for a 3-minute povidone-iodine topical scrub (13% solution) to be administered with the sponge side of a surgical scrub brush to the area 4 cm below the anticipated incision up to the umbilicus of all cesarean delivery patients before entering the operating room. All patients were scrubbed by a registered nurse at the time of admission if they were scheduled for cesarean delivery. The scrub occurred in the Labor and Delivery room if concern for need of cesarean delivery arose during labor, as determined by the physician caring for the patient. The protocol then called for this scrub to be followed by a povidone-iodine topical solution paint (10% solution), which was applied to the surgical site in the operating room immediately before incision (SCRUB+PAINT). This protocol was implemented through simultaneous and universal education of all Labor and Delivery staff, including physicians and nurses. Specifically, the Obstetrics nursing leadership met with all Labor and Delivery nurses and instructed them on the details of the cesarean delivery preoperative skin preparation protocol, including the timing of administration and the specific technique for applying the povidone-iodine topical scrub (13% solution). Physicians were educated on the details of the protocol change at departmental grand rounds conferences and also received electronic messages regarding the change in protocol. During the study time period, the protocol was intermittently reinforced by the head of Obstetrics Clinical Effectiveness and Quality Improvement committee to ensure continued compliance.
We identified all women who underwent cesarean delivery from October 1, 2007, through December 31, 2009, using the standard diagnosis related group codes 370 and 371. Women were divided into two groups according to the timing of their cesarean delivery relative to the initiation of the SCRUB+PAINT intervention. Specifically, the preintervention group, exposed to PAINT alone, consisted of all patients with a cesarean delivery between October 1, 2007, and September 30, 2008. The postintervention group, exposed to SCRUB+PAINT, consisted of all patients with a cesarean delivery between January 1, 2009, and December 31, 2009. Women who underwent cesarean delivery in the month before, during, or after initiation of SCRUB+PAINT (October 1, 2008–December 31, 2008) were excluded to allow for a “washout” period around the time of implementation of the intervention.
Outcomes and confounding variables were identified using International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) coding. The primary outcome was major puerperal infection (ICD-9-CM 670.xx), which included endometritis, sepsis, cellulitis, peritonitis, and salpingitis. The secondary outcome was a composite outcome of major puerperal infection and infectious wound complications (ICD-9-CM 670.xx and 674.xx). Outcomes were measured up to 30 days after surgery for all patients admitted to the hospital. Our final models controlled for the following confounding variables based on identification by ICD-9-CM codes: diabetes (pregestational and gestational diabetes), obesity, chorioamnionitis, cefazolin prophylaxis, previous cesarean delivery, maternal race, month of delivery, failed induction of labor, and failed trial of labor.
Statistical analyses were performed using Stata 10.1. Categorical data were compared across the treatment time periods (preintervention and postintervention) using χ2 tests. Means of continuous data were compared using Student t tests. Poisson regression was used to estimate monthly infection rates and to test for differences in infection rates by treatment (PAINT compared with SCRUB+PAINT) after controlling for confounders and any variations in the rate of infections over time.16
The infection rate per month is defined as the number of infectious complications, including both major puerperal infection and a composite of major puerperal infections and infectious wound complications, divided by the number of cesarean deliveries that occurred during the month. The Poisson regression models a log transformation of the monthly infection rate and allows for the inclusion of multiple covariates to examine and test whether these covariates influence the rates over time. The quantity is a rate because it has a specific unit of time associated with it. Rates typically follow a Poisson distribution. If the infection rate were stable over time so that there were a constant number of infectious complications divided by total cesarean deliveries performed each month, then a logistic regression could be used to approximate the Poisson rate model. However, we could not assume that the infection rate over time was constant, because the primary objective of the study was to investigate whether a change in protocol resulted in a change in the infection rates.
Finally, the data were evaluated for systematic changes in the infection rates over time to rule out the potential for secular trends as an explanation for reduced infection rates after the SCRUB+PAINT intervention was implemented. To accomplish this, we defined a quantitative time variable, in units of month from study intervention, which was coded with values −12 to −1 for times in the preintervention time period before the protocol change and values 1 to 12 for the postintervention time period after the change. This time variable was added to the model as a quantitative variable to test for a linear trend. Including a squared version of the variable allowed for assessment of a quadratic trend. We fit these models for the entire study period and individually for both the preintervention and postintervention time periods. No significant trends over time were detected. Further, to accurately estimate the difference in the infection rate with the protocol change we included indicator variables for calendar month. By doing this, the incident rate ratio for protocol change is estimated as a weighted average of observed monthly ratios comparing each calendar month under the old protocol to the same calendar month under the new protocol (ie, January to January). Therefore the incident rate ratio associated with the change in protocol is adjusted for any inherent variability over time.
Data from 2,143 patients (PAINT: n=1,139; SCRUB+PAINT: n=1,004) were included in the analysis. The number of cesarean deliveries performed during the study period ranged from 60 to 112 per month.
No significant differences in demographic variables existed between the preintervention and postintervention groups with respect to the diagnosis of chorioamnionitis, pregestational diabetes, and previous cesarean delivery. However, women in the postintervention group (SCRUB+PAINT) were more likely to be obese, be African American, receive cefazolin prophylaxis, and have a cesarean delivery for failed induction of labor than women in the preintervention group (PAINT) (Table 1).
Overall, SCRUB+PAINT was associated with a decrease in the risk of major puerperal infection (3.4 compared with 5.4 of 100 cesarean deliveries, P=.03) and composite wound infection (5.5 compared with 7.8 of 100 cesarean deliveries, P=.03) compared with PAINT alone. Postoperative length of hospital stay was also significantly shorter (3 compared with 4 days, P=.003) in the SCRUB+PAINT group.
Poisson regression was performed to adjust for potential confounders known to be associated with infectious risk 3,4,6,17–20 including diabetes (pregestational and gestational diabetes), obesity, chorioamnionitis, cefazolin prophylaxis, history of cesarean delivery, maternal race, month of delivery, failed induction of labor, and failed trial of labor. Controlling for these biologically plausible confounders, SCRUB+PAINT was associated with a 38% reduction in major puerperal infection (incident rate ratio 0.62 [0.42–0.93], P=.02) and a 31% reduction in composite wound infection (incident rate ratio 0.69 [0.50–0.96], P=.03) compared with PAINT alone after controlling for confounders and fluctuations in the rates of infection over time (Table 2).
Secular trend analyses revealed that there was no significant decrease in the number of infections over time before initiating SCRUB+PAINT (P=.42) (Fig. 1).
We conclude that a more rigorous preoperative skin-preparation technique involving a 3-minute povidone-iodine scrub (13% solution) followed by a povidone-iodine topical paint (10% solution) is associated with a decrease in the rate of postcesarean delivery infectious complications compared with povidone-iodine topical paint (10% solution) alone.
To our knowledge, there are only two published studies that attempt to address which preoperative skin preparation is optimal to reduce the rate of postcesarean delivery infectious complications.9,10 Neither study found a significant reduction in infectious complications between treatment algorithms. A third general surgery study comparing povidone-iodine scrub followed by a povidone-iodine paint to povidone-iodine paint alone found no difference in rates of wound infection.21
Our study involves a much larger sample size than previous studies,9,10,21 and it is the only study to have been performed in the past 15 years that specifically describes rates of postcesarean delivery infectious complications when comparing two different preoperative skin preparation methods. Our large sample size allowed us to control for several important confounders that are known to be associated with infectious risk.3,4,6,17–20 Deriving all data from patients at a single institution over a fairly brief time period minimized variability in practice pattern between patients in each study group. There were also no changes in the surveillance of or in the definition of infections during the study time period. Finally, although ICD-9-CM coding was performed retrospectively, this coding was done by trained hospital personnel thereby minimizing the chances of coding errors and inconsistencies. There were no known changes in coding education or definitions during or before the study time period. The benefit of using ICD-9-CM code data is that it allows for the evaluation of a large number of patients over a short period of time to determine the effect of the intervention soon after its implementation.
Our study was not without limitations. It was an observational study and not an interventional trial with randomization to each study group. Second, we assumed that the SCRUB+PAINT intervention occurred for all cesarean deliveries after November 2008. This intervention was primarily nursing-executed and whether or not the patient received both SCRUB and PAINT is not specifically documented. Therefore, it is possible that some patients may not have been exposed to this intervention. However, we performed the analyses as intention to treat. Thus, if some SCRUB+PAINT patients were actually exposed to PAINT only, our results would be biased toward the null, further strengthening our significant findings.
Third, we identified study outcomes and potential confounders using ICD-9-CM codes. As a result, some outcomes and confounders may have been misclassified. Specifically, undercoding may result in lack of ascertainment of some confounders. However, there were no changes in coding or, in the definitions of, postcesarean delivery infectious complications or specific confounders during the study time period. Therefore, any misclassification is nondifferential and likely similar in both study groups. The benefits of using such a data set, with the large number of patients included, outweigh the limitations with regard to the specific aims of this study.
Another limitation is that only infections diagnosed during the initial hospital admission or during a readmission in the first 30 days postoperatively were included. Although our methodology allows us to capture the majority of the most concerning postoperative infectious complications, patients with infections who presented to an outpatient clinic, the emergency department, or another institution were not included. Evaluating the effect of our protocol change on the incidence of infectious complications diagnosed and managed in the outpatient setting is an important topic for further investigation.
Finally, although there is some variability in the primary outcome from month to month, we found no evidence of a secular trend preceding the intervention that accounts for the decrease in infectious complications after the protocol was implemented.
As the nationwide cesarean delivery rate continues to rise, identifying methods through which postcesarean delivery infectious complication rates can be reduced is critical. Given our findings, we recommend implementation of a preoperative skin-preparation technique involving a 3-minute povidone-iodine (13% solution) scrub followed by a povidone-iodine (10% solution) paint over preparation with povidone-iodine (10% solution) paint alone. Future studies must continue to evaluate whether other interventions can even further reduce postcesarean delivery infectious related morbidity.
1. Hamilton BE, Martin JA, Ventura SJ. Births: preliminary data for 2007. Natl Vital Stat Rep 2007;56:1–18.
2. Henderson E, Love EJ. Incidence of hospital-acquired infections associated with cesarean section. J Hosp Infect 1995;29:245–55.
3. Olsen MA, Butler AM, Willers DM, Devkota P, Gross GA, Fraser VJ. Risk factors for surgical site infection after low transverse cesarean section. Infect Control Hosp Epidemiol 2008;29:477–84.
4. Opoien HK, Valbo A, Grinde-Andersen A, Walberg M. Post-cesarean surgical site infections according to CDC standards: rates and risk factors. A prospective cohort study. Acta Obstet Gynecol Scand 2007;86:1097–102.
5. Yokoe DS, Christiansen CL, Johnson R, Sands KE, Livingston J, Shtatland ES, et al. Epidemiology of and surveillance for postpartum infections. Emerg Infect Dis 2001;7:837–41.
6. Leth RA, Moller JK, Thomsen RW, Uldbjerg N, Norgaard M. Risk of selected postpartum infections after cesarean section compared with vaginal birth: a five-year cohort study of 32,468 women. Acta Obstet Gynecol Scand 2009;1–8
7. Koroukian SM. Relative risk of postpartum complications in the Ohio Medicaid population: vaginal versus cesarean delivery. Med Care Res Rev 2004;61:203–24.
8. 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.
9. Lorenz RP, Botti JJ, Appelbaum PC, Bennett N. Skin preparation methods before cesarean section: a comparative study. J Reprod Med 1988;33:202–4.
10. Magann EF, Dodson MK, Ray MA, Harris RL, Martin JN Jr, Morrison JC. Preoperative skin preparation and intraoperative pelvic irrigation: impact on post-cesarean endometritis and wound infection. Obstet Gynecol 1993;81:922–5.
11. Brown TR, Ehrlich CE, Stehman FB, Golichowski AM, Madura JA, Eitzen HE. A clinical evaluation of chlorhexidine gluconate spray as compared with iodophor scrub for preoperative skin preparation. Surg Gynecol Obstet 1984;158:363–6.
12. Swenson BR, Hedrick TL, Metzger R, Bonatti H, Pruett TL, Sawyer RG. Effects of preoperative skin preparation on postoperative wound infection rates: a prospective study of 3 skin preparation protocols. Infect Control Hosp Epidemiol 2009;30:964–71.
13. Darouiche RO, Wall MJ Jr, Itani KM, Otterson MF, Webb AL, Carrick MM, et al. Chlorhexidine-alcohol versus povidone-iodine for surgical-site antisepsis. N Engl J Med 2010;362:18–26.
14. Edwards PS, Lipp A, Holmes A. Preoperative skin antiseptics for preventing surgical wound infections after clean surgery. The Cochrane Database of Systematic Reviews 2004, Issue 3. Art. No.: CD003949. DOI: 10.1002/14651858.CD003949.pub2.
15. Ritter MA, French ML, Eitzen HE, Gioe TJ. The antimicrobial effectiveness of operative-site preparative agents: a microbiological and clinical study. J Bone Joint Surg Am 1980;62:826–8.
16. Zou G. A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol 2004;159:702–6.
17. Schneid-Kofman N, Sheiner E, Levy A, Holcberg G. Risk factors for wound infection following cesarean deliveries. Int J Gynaecol Obstet 2005;90:10–5.
18. Chaim W, Bashiri A, Bar-David J, Shoham-Vardi I, Mazor M. Prevalence and clinical significance of postpartum endometritis and wound infection. Infect Dis Obstet Gynecol 2000;8:77–82.
19. Killian CA, Graffunder EM, Vinciguerra TJ, Venezia RA. Risk factors for surgical-site infections following cesarean section. Infect Control Hosp Epidemiol 2001;22:613–7.
20. Tran TS, Jamulitrat S, Chongsuvivatwong V, Geater A. Risk factors for postcesarean surgical site infection. Obstet Gynecol 2000;95:367–71.
21. Ellenhorn JD, Smith DD, Schwarz RE, Kawachi MH, Wilson TG, McGonigle KF, et al. Paint-only is equivalent to scrub-and-paint in preoperative preparation of abdominal surgery sites. J Am Coll Surg 2005;201:737–41.