Downward Trends in Surgical Site and Urinary Tract Infections After Cesarean Delivery in a French Surveillance Network, 1997‐2003
Objective. To evaluate whether the adjusted rates of surgical site infection (SSI) and urinary tract infection (UTI) after cesarean delivery decrease in maternity units that perform active healthcare‐associated infection surveillance.
Design. Trend analysis by means of multiple logistic regression.
Setting. A total of 80 maternity units participating in the Mater Sud‐Est surveillance network.
Patients. A total of 37,074 cesarean deliveries were included in the surveillance from January 1, 1997, through December 31, 2003.
Methods. We used a logistic regression model to estimate risk‐adjusted post–cesarean delivery infection odds ratios. The variables included were the maternity units’ annual rate of operative procedures, the level of dispensed neonatal care, the year of delivery, maternal risk factors, and the characteristics of cesarean delivery. The trend of risk‐adjusted odds ratios for SSI and UTI during the study period was studied by linear regression.
Results. The crude rates of SSI and UTI after cesarean delivery were 1.5% (571 of 37,074 patients) and 1.8% (685 of 37,074 patients), respectively. During the study period, the decrease in SSI and UTI adjusted odds ratios was statistically significant (
[
] and 
[
], respectively).
Conclusion. Reductions of 48% in the SSI rate and 52% in the UTI rate were observed in the maternity units. These unbiased trends could be related to progress in preventive practices as a result of the increased dissemination of national standards and a collaborative surveillance with benchmarking of rates.
Received May 29, 2007; accepted November 13, 2007; electronically published February 1, 2008.
Surgical site infections (SSIs) and urinary tract infections (UTIs) are the 2 infections that most commonly occur after cesarean delivery. Various surveillance systems from Europe and the United States have reported rates of SSI after cesarean delivery (hereafter, “postcesarean SSI”) of 1.6‐5.38 per 100 deliveries.1‐3 Fewer data are available regarding epidemiology for postpartum UTI, but all published data show rates of 3‐4 UTIs per 100 deliveries.4‐6 These rates are too high, considering the fact that most of the women are young and healthy at the time of delivery. The Mater Sud‐Est Study Group (Mater Network) is a healthcare‐associated infection (HAI) continuous surveillance network of maternity units located in southeast France.7,8 Since 1997, the first year of surveillance, one of the aims of the Mater Network has been to help staff members of maternity units decrease the HAI rate by a variety of means, including surveillance and regular feedback of results, annual network meetings, and distribution of half‐yearly network magazines of posters. The present study was formulated to evaluate the impact of this surveillance on the postcesarean SSI and UTI rates.
Methods
Principles of the Mater Network
Participation in the Mater Network is voluntary, but the surveillance should be performed for at least 4 months. All ward‐specific data are kept confidential. Surveillance is performed by an infection control practitioner or a dedicated midwife. The surveillance was initiated in 1997 with 25 maternity units. In 2003, 53 units participated in the surveillance. During the study period (1997‐2003), 80 maternity units participated for at least 4 months in the Mater Network. The surveillance included cesarean deliveries and vaginal deliveries, but because the subject of this article is the decrease of infection after cesarean delivery, we do not describe the surveillance of vaginal deliveries.
The surveillance form incorporated a description of the maternity unit: the category of the maternity unit (level 1, no pediatric unit within the hospital; level 2, at least 1 neonatal or pediatric unit within the hospital; level 3, at least 1 neonatal intensive care unit within the hospital9), number of cesarean deliveries per year, number of vaginal deliveries per year, and number of childbirths per year. It also included the individual characteristics of the mother (age, parity, date of cesarean delivery, date of maternity discharge), pregnancy risk factors (UTI during pregnancy, UTI at admission to the maternity unit, labor pyrexia, premature ruptured membranes [12 hours or more before admission], blood loss of more than 800 mL), the characteristics of cesarean delivery (elective cesarean delivery, first cesarean delivery, use of antibiotic prophylaxis, use of general anesthetic, or use of urethral catheterization), and HAI data (type of infection, date of diagnosis). The 2 HAIs that we investigated in this study are UTI and SSI. Because data on superficial SSI and deep SSI were not collected from 1997 through 2000, we only present the global rate of SSI. The number of days of urethral catheterization was not collected, because the duration of the catheterization never exceeded 24 hours after a cesarean delivery. All risk factors were selected after a careful review of the relevant literature, which was performed at the beginning of the surveillance and regularly updated.6,10‐12 The definitions of the infections are based on Centers for Disease Control and Prevention definitions and on French definitions written by the French National Council of Infection Control.13,14 Postdischarge surveillance was not routinely performed, but a SSI or UTI that became evident after discharge was included if the infection occurred within 30 days after the cesarean delivery. Every 4 months, the data were sent to the coordinating center, which controlled the quality of the data. Each maternity unit received quarterly and annually their own results and the network results.15 Because the UTI and SSI standardized incidence ratios (risk‐adjusted infection incidence rates) were calculated with a logistic regression model, each unit could compare their own infection rates with the rates of the network units found in the annual network report.
Statistical Analysis
Data management and analysis were performed with the software Epi‐Info version 6.04b (Centers for Disease Control and Prevention) and SPSS software package (SPSS) 13.0 for Windows. The χ2 and Student t tests were used for discrete and continuous variables, when appropriate. The variables age, number of deliveries per year, number of vaginal deliveries per year, and number of cesarean deliveries per year were divided into quartiles for upfront analyses and analyzed as discrete variables. The crude trends in SSI and UTI rates during the study period were analyzed with a χ2 test for trend.
The logistic regression equation was used to determine the adjusted odds of acquiring an SSI or a UTI, depending on the patient's risk factors, on the maternity unit's characteristics, and on the year of childbirth. Variables that were likely to be associated with the outcome (P value of .2 or less in univariate analysis) or considered potential confounders in the literature and the variable “year of childbirth” were included in the multivariate logistic regression model. P less than .05 was considered statistically significant.
To analyze the temporal trend of the adjusted odds ratios (ORs) for SSI and for UTI, we sought to determine a correlation between these ORs and the year of delivery. Pearson and Spearman tests were used to analyze the significance of the correlation.
Results
Patient Population
During the study period, 37,074 cesarean deliveries were performed in 80 maternity units (information for some characteristics was not available for all women). Of these 37,074 women, 685 (1.8%) had a postcesarean UTI, and 571 (1.5%) had a postcesarean SSI. Few of the women were younger than 22 years (1,550 [4.2%]) or older than 39 years (1,748 [4.7%]). Most women were 22‐39 years old (33,733 [91.1%]), were nulliparous (17,964 [48.5%]), and underwent a cesarean delivery for the first time (24,509 [53.9%]). UTI during the pregnancy or UTI at admission were not frequent risk factors (1,146 [0.5%] and 683 [2.0%] women, respectively). Women with premature ruptured membranes and/or pyrexia during labor accounted for 1,264 (3.4%) and 1,031 (2.8%) of the patients, respectively.
Estimated blood loss of more than 800 mL occurred in 1,753 women (4.7%). More than half of the women (19,936 [53.8%]) underwent a planned cesarean delivery, and 5,137 women (13.9%) underwent a cesarean delivery under general anesthetic. Most of the women received urinary catheterization (35,770 [96.5%]) and antimicrobial prophylaxis (29,961 [81.7%]).
Almost a quarter of the women (8,691 [23.4%]) had given birth in a level 3 maternity unit, 16,513 (44.5%) in a level 2 maternity unit, and 11,817 (31.9%) in a level 1 maternity unit. Most of these cesarean deliveries (15,289 [41.2%]) occurred in a busy maternity unit (ie, more than 103 childbirths per month), a quarter (9,489 [25.6%]) in a maternity unit of average activity (ie, 69‐102 childbirths per month), 7,589 (20.5%) in a small maternity unit (ie, 50‐68 childbirths per month), and 4,707 (12.7%) in an even smaller maternity unit (ie, fewer than 50 childbirths per month).
From 1997 to 2003, the crude SSI rate decreased steadily from 2.5% (74 of 2,938) to 1.3% (81 of 6,436) (
), and the UTI rate decreased from 2.8% (83 of 2,938) to 1.4% (88 of 6,436) (
). During the study period, the mean length of stay of women who underwent a cesarean delivery decreased steadily from 7.3 days in 1997 to 6.5 days in 2003 (
).
Univariate and Multivariate Analysis
Factors associated with the risk of developing postcesarean SSI or UTI are listed in Tables 1 and 2, respectively. Some typical risk factors associated with postcesarean SSI or UTI were not independent risk factors: antimicrobial prophylaxis, primary cesarean delivery, nulliparous status, greater or lesser maternal age, premature ruptured membranes, and UTI during pregnancy. Multiple logistic regression found 5 different predictors that were independently associated with postcesarean SSI or UTI: the year of delivery, the number of deliveries per month in the maternity unit, the number of cesarean deliveries per month in the maternity unit, and the maternity unit level of neonatal care (Tables 3 and 4). Planned cesarean delivery was independently associated with SSI, whereas UTI at admission, blood loss of more than 800 mL, and primary cesarean delivery were independently associated with UTI.
Regression Analysis
During the study period, we found that the adjusted ORs of acquiring a postcesarean SSI or UTI decreased significantly, using the first surveillance year as a reference. For a postcesarean SSI, the Pearson correlation coefficient (R) was −0.823 (
), and the Spearman rank correlation coefficient (ρ) was −0.786 (
). For a postcesarean UTI, the Pearson correlation coefficient (R) was −0.906 (
), and the Spearman rank correlation coefficient (ρ) was −0.926 (
) (Figures 1 and 2).
Figure 1. Evolution of odds ratios for postcesarean surgical site infection during the Mater Network surveillance study from 1997 to 2003.
Figure 2. Evolution of odds ratios for postcesarean urinary tract infection during the Mater Network surveillance study from 1997 to 2003.
Discussion
We found that, during the study period, there was a decrease in the adjusted rates of postcesarean SSI and UTI. Although the infection rates at the beginning were not high, we achieved a significant reduction of 48% in the SSI rate (from 2.5% to 1.3%) and 52% in the UTI rate (from 2.9% to 1.4%).
The present study has some limitations. During the last few years, changes in the characteristics of mothers and in obstetric practices15,16 have certainly influenced the primary cesarean delivery rate and may have influenced the evolution of the HAI rate. The decrease in postcesarean UTI and SSI rates may also be partially attributable to a steady reduction in the hospital stay of mothers from 7.3 to 6.5 days during the study period. This may also be a methodologic bias attributable to the lack of formal postdischarge surveillance in this network. Couto et al.17 report a 6‐fold multiplication of the SSI rate when SSI detected by postdischarge surveillance was included (from 1.6% to 9.6%). In other studies, SSI rates of 42%‐71% were detected after discharge.18,19
The lack of postdischarge surveillance may explain why the HAI rates we report are lower than the infection rates found in the literature. We detail, during a 7‐year surveillance period, a postcesarean SSI rate of 1.5%, whereas other studies report rates of 2.17%‐11.2%.3,6,18,20 The Mater Network postcesarean UTI rate is also far below the 6% rate found in the literature.21,22 The Mater Network postcesarean SSI rate compares with the network rates that lack formal postdischarge surveillance. In 2006, the German national surveillance system (Krankenhaus Infektions Surveillance System) reported a postcesarean SSI rate of 1.9%, during a 3‐year surveillance period.2 Couto et al.17 identified, by means of an in‐hospital surveillance, a postcesarean SSI rate of 1.6%. However, comparison of rates is limited by the variety of definitions of HAI used. Unlike other surveillance studies, ours did not include endometritis in the SSI rate and did not include asymptomatic bacteriuria in the reported UTI rate.3,22,23
The 80 units that participated in the surveillance study were not included for the entire period. A standardization method lessens the influence of individual units on the network results. Like other investigators, we chose to use logistic regression as the risk adjustment method.19,24 Logistic regression modeling has 2 advantages, compared with direct standardization: adjusted infection ORs are directly obtained, and it does not matter how large the included population is. Some confounding variables were not controlled in this study: the mother’s body mass index, whether the mother had diabetes mellitus, and the American Society of Anesthesiologists (ASA) score.3,25‐27 Those variables were not included in our surveillance because of the difficulty in collecting them. The ASA score is common in surgical surveillance networks, but the Mater Network also included vaginal deliveries (80% of all included deliveries) and therefore had to incorporate other major risk factors. Moreover, among the included population, the ASA score is almost always equal to 1 and therefore is not a discriminating factor.
Nevertheless, the major risk factors (namely, greater or lesser maternal age, nulliparous status, anesthesia, urinary catheterization, pyrexia during labor, premature ruptured membranes, and primary cesarean delivery) have been studied as univariate factors and were included in the logistic regression if they were statistically significant.18,20,22,26,28,29
The level of dispensed neonatal care in the maternity unit, the maternity unit’s annual rate of operative procedure in question (ie, childbirth, cesarean delivery, and vaginal delivery), and the year of childbirth have also been included in the multivariate analysis. Our findings demonstrate that hospitals with a lower‐than‐average number of cesarean deliveries per month (fewer than 12) or with fewer than 50 vaginal and cesarean deliveries per month experience higher postcesarean SSI and UTI rates. Gould et al.30 demonstrate that hospitals with a lower‐than‐average cesarean delivery rate for low‐risk mothers experience higher levels of neonatal morbidity and therapeutic interventions. The level of dispensed neonatal care in the maternity unit was a major risk factor, independent of the annual rate of surgical procedures. We did not find this particular risk factor in the obstetrics literature, but it is well known that the dispensing practices of the institutions represent extrinsic patient risk factors for infection.31 Level 3 maternity units, which have to perform most of the high‐risk childbirths, are expected to have a patient population with higher intrinsic risk factors for nosocomial infection than do level 1 maternity units, without such a case mix. These results emphasize the need to include several maternity unit characteristics, in addition to maternal risk factors, in a standardization model, to facilitate a fairer comparison of HAI rates across institutions.
We believe that the reduction in SSI and UTI rates reported in this study reflects a real reduction in the HAI rate and an improvement in the quality of care in the maternity units. Surveillance has been shown to be a powerful instrument in the reduction of SSIs and other HAIs, especially when reference data are available for comparison.32‐34 Part of the decrease in HAIs reported in this study is certainly attributable to the regular feedback to the units; however, the significance of participating in a network, which would indicate a particular interest in HAI control and in specific measures for infection control, was not assessed.35 Indeed, if surveillance plays a role in the decrease of HAIs in the network units, it is certainly reinforced by the evolution of some obstetric practices and care guidelines. All these guidelines were regularly reported to the network units, by the annual network meetings and the half‐yearly network magazine.
Although maternity units that do not participate in surveillance may also experience a decrease in HAIs, this study suggests that continuous and noncontinuous involvement in a performance measurement exercise, with the reporting of results, contributes to a decrease in HAIs. Future research comparing a network of maternity units with a control group should help clarify the role surveillance networks play in decreasing HAIs.
Acknowledgments
We thank the maternity units of the Mater Sud‐Est Study Group: Ambérieu en Bugey, Alès, Annonay, Aubagne, Aubenas, Aurillac, Avignon, Beaune, Belley, Bourg‐en‐Bresse, Bourgoin, Brignoles, Cannes, Carcassonne, Carpentras, Châlon, Clermont‐Ferrand, Décines, Désertine, Digne, Dole, Draguignan, Écully, Fréjus, Givors, Grasse, Grenoble, Hyères, Issoire, La Ciotat, L’Arbresle, Lyon, Macon, Montbrison, Montélimar, Montpellier, Moulins, Narbonne, Nice, Nîmes, Orange, Perpignan, Pertuis, Pierre Bénite, Puyricard, Reims, Roanne, Roussillon, Sallanches, St Agrève, St Chamond, St Etienne, St Flour, Ste Foy‐les‐Lyon, St Jean de Maurienne, St Martin d’Hères, Tarare, Toulon, Tournon, Trinité, Valréas, Vaux‐en‐Velin, Vienne, Villeurbanne, Villefranche s/ Saône, Voiron. We also thank Mrs. Dorothy Russell for her attentive reviewing.
Financial support. IHRC‐NOSOQUAL, French Ministry of Health.
Potential conflicts of interest. All authors report no conflicts of interest relevant to this article.
References
- 1. Brandt C, Sohr D, Behnke M, Daschner F, Rüden H, Gastmeier P. Reduction of surgical site infection rates associated with active surveillance. Infect Control Hosp Epidemiol 2006; 27:1347‐1351.
- 2. Bärwolff S, Sohr D, Geffers C, et al. Reduction of surgical site infections after Caesarean delivery using surveillance. J Hosp Infect 2006; 64:156‐161.
- 3. Centers for Disease Control and Prevention. National Nosocomial Infections Surveillance (NNIS) system report, data summary from January 1992 through June 2004, issued October 2004. Am J Infect Control 2004; 32:470‐485.
- 4. Malavaud S, Bou‐Segonds E, Berrebi A, Castagno R, Assouline C, Connan L. Determination of nosocomial infection incidence in mothers and newborns during the early postpartum period [in French]. J Gynecol Obstet Biol Reprod (Paris) 2003; 32:169‐174.
- 5. Van Ham MA, Van Dongen PW, Mulder J. Maternal consequences of caesarean section: a retrospective study of intra‐operative and postoperative maternal complications of caesarean section during a 10‐year period. Eur J Obstet Gynecol Reprod Biol 1997; 74:1‐6.
- 6. Mead P, Hess S, Page S. Prevention and control of nosocomial infections in obstetrics and gynecology. In: Wenzel R, ed. Prevention and Control of Nosocomial Infections. 3rd ed. Baltimore, MD: Williams and Wilkins; 1997:995‐1016.
- 7. Tissot‐Guerraz F, Moussy L, Agniel F, et al. Hospital infection in the maternity department: 3 years of surveillance in 9,204 deliveries of which 1,333 were cesarean sections [in French]. J Gynecol Obstet Biol Reprod (Paris) 1990; 19:19‐24.
- 8. Girard R, Chemorin C, Lafarge J, et al. Diminution de l’incidence des infections nosocomiales au cours d’une surveillance en obstétrique. Med Hyg 1998; 56:1864‐1869.
- 9. Ministère de la Santé et de la Solidarité. Circulaire dh/eo 3 n°99‐402 du 8 juillet 1999 relative à l’application de l’arrêté du 1er avril 1999 fixant les indices de besoins nationaux afférents à la néonatologie et à la réanimation néonatale. Non paru au Journal Officiel 1999.
- 10. Donowitz L. Infection in the newborn. In: Wenzel R, ed. Prevention and Control of Nosocomial Infections. 2nd ed. Baltimore, MD: Williams and Wilkins; 1992:796‐811.
- 11. Hagglund L, Christensen K, Christensen P, Kamme C. Risk factors in cesarean section infection. Obstet Gynecol 1983; 62:145‐150.
- 12. Ott W. Primary cesarean section: factors related to postpartum infection. Obstet Gynecol 1981; 57:171‐176.
- 13. Garner J, Jarvis W, Emori T, Hogan T, Hughes J. CDC definitions for nosocomial infections. Am J Infect Control 1988; 16:128‐140.
- 14. French Ministry of Health and French National Council of Infection Control. 100 recommandations pour la surveillance et la prévention des infections nosocomiales. 2nd ed. Paris; 1999.
- 15. Vincent‐Boulétreau A, Caillat‐Vallet E, Dumas AM, et al. Changing medical practices and nosocomial infection rates in French maternity units from 1997 to 2000 [in French]. J Gynecol Obstet Biol Reprod (Paris) 2005; 34:128‐136.
- 16. Joseph KS, Young DC, Dodds L, et al. Changes in maternal characteristics and obstetric practice and recent increases in primary cesarean delivery. Obstet Gynecol 2003; 102:791‐800.
- 17. Couto RC, Pedrosa TM, Nogueira JM, Gomes DL, Neto MF, Rezende NA. Post‐discharge surveillance and infection rates in obstetric patients. Int J Gynaecol Obstet 1998; 61:227‐231.
- 18. Johnson A, Young D, Reilly J. Caesarean section surgical site infection surveillance. J Hosp Infect 2006; 64:30‐35.
- 19. Mitt P, Lang K, Aira P, Maimemts M. Surgical site infections following cesarean section in an Estonian university hospital: postdischarge surveillance and analysis of risk factors. Infect Control Hosp Epidemiol 2005; 26:449‐454.
- 20. Mah M, Pyper A, Oni G, Memish Z. Impact of antibiotic prophylaxis on wound infection after cesarean section in a situation of expected higher risk. Am J Infect Control 2001; 29:85‐88.
- 21. Senanayake H. Elective cesarean section without urethral catheterization. J Obstet Gynaecol Res 2005; 31:32‐37.
- 22. Leigh D, Sedgwick J, Dean R. Post‐operative urinary tract infection and wound infection in women undergoing caesarean section: a comparison of two study periods in 1985 and 1987. J Hosp Infect 1990; 15:107‐116.
- 23. Barbut F, Carbonne B, Truchot K, et al. Surgical site infections after cesarean section: results of a five‐year prospective surveillance [in French]. J Gynecol Obstet Biol Reprod (Paris) 2004; 33:487‐496.
- 24. Bailit J, Garrett J. Comparison of risk‐adjustment methodologies for cesarean delivery rates. Obstet Gynecol 2003; 102:45‐51.
- 25. Takoudes C, Weitzen S, Slocum J, Malee M. Risk of cesarean wound complication in diabetic gestations. Am J Obstet Gynecol 2004; 191:958‐963.
- 26. Tran T, Jamulitrat S, Chongsuvivatwong V, Geater A. Risk factors for postcesarean surgical site infection. Obstet Gynecol 2000; 95:367‐371.
- 27. Sarsam S, Elliott J, Lam K. Management of wound complication from cesarean delivery. Obstet Gynecol Surv 2005; 60:462‐473.
- 28. Schneid‐Kofman N, Sheiner E, Levy A, Holcberg G. Risk factors for wound infection following cesarean deliveries. Int J Gynaecol Obstet 2005; 90:10‐15.
- 29. Schwartz M, Wang C, Eckert L, Critchlow C. Risk factors for urinary tract infection in the postpartum period. Am J Obstet Gynecol 1999; 181:547‐553.
- 30. Gould JB, Danielsen B, Korst LM, et al. Cesarean delivery rates and neonatal morbidity in a low‐risk population. Obstet Gynecol 2004; 104:11‐19.
- 31. Centers for Disease Control and Prevention. Nosocomial infection rates for interhospital comparison: limitation and possible solutions. Infect Control Hosp Epidemiol 1991; 12:609‐621.
- 32. NNIS System. National nosocomial infection surveillance system report, data summary from January 1992–June 2001, issued August 2001. Am J Infect Control 2001; 29:404‐421.
- 33. 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.
- 34. Gastmeier P, Geffers C, Brandt C, et al. Effectiveness of a nationwide nosocomial infection surveillance system for reducing nosocomial infections. J Hosp Infect 2006; 64:16‐22.
- 35. Vanhems P, Dumas A, Berland M, et al. Hospital acquired infections in French maternity units: trend noted in 2‐year surveillance (1997‐1998). J Hosp Infect 2000; 45:334‐335.
-
Presented in part: 17th European Congress of Clinical Microbiology and Infectious Diseases; Munich, Germany; March 31–April 3, 2007. Oral session O15.