Postdischarge surveillance following cesarean section: The incidence of surgical site infection and associated factors Meire Celeste Cardoso Del Monte, RN,a and Aarao Mendes Pinto Neto, MD, PhDb ˜ ˜ Sao Paulo, Brazil Background: The rate of surgical site infections (SSI) and their associated risk factors was identiﬁed by performing postdischarge surveillance following cesarean section at a public university teaching hospital in Brazil. Methods: The study was conducted at the Center for Women’s Integrated Health Care in Brazil between May 2008 and March 2009. Women were contacted by telephone 15 and 30 days after cesarean section. During hospitalization, a form was completed on factors associated with post-cesarean SSI. The x2 test and Fisher exact test were used to analyze categorical variables and the Mann-Whitney test for numerical variables. Relative risks (RR) and their respective 95% conﬁdence intervals (95% CI) were calculated for factors associated with SSI. P values , .05 were considered signiﬁcant. Results: The ﬁnal sample consisted of 187 women. SSI was detected in 44 cases (23.5%). In 42 of 44 women (95%), SSI appeared following discharge from hospital, becoming evident within the ﬁrst 15 days following surgery. Number of prenatal consultations #7 (RR, 2.09; 95% CI: 1.26-3.48) and hypertension (RR, 2.07; 95% CI: 1.25-3.43) were signiﬁcantly associated with SSI in the bivariate analysis. In the multivariate analysis, only hypertension (RR, 2.47; 95% CI: 1.21-5.04) remained signiﬁcant. Conclusion: Postdischarge surveillance is essential for ensuring accurate estimates of post-cesarean section SSI. A 15-day postdischarge follow-up was shown to be sufﬁcient. Hypertension was a factor associated with SSI. Key Words: Postdischarge surveillance; cesarean section; surgical site infection; hospital infection; infection control nurse. Copyright ª 2010 by the Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved. (Am J Infect Control 2010;38:467-72.) Surgical site infections (SSI) are common postoperative their statistics to those patients in whom infectioncomplications, constituting a major clinical problem in becomes apparent prior to discharge or those who re-terms of morbidity and mortality,1 duration of hospitali- turn spontaneously to the hospital for treatment. Vari-zation, and hospital costs.2,3 ous studies have been published showing a consensus Most SSIs only become apparent after the patient is on the need to perform postdischarge surveillance ofdischarged from hospital.4-13 Different studies that patients submitted to C-sections to obtain moreincluded postdischarge surveillance have reported accurate statistics on the frequency of SSI.7-13infection rates varying from 27.6%4 to 84%,5 particu- In Brazil, Couto et al (1998)12 reported post-C-sectionlarly following surgeries such as cesarean sections SSI rates of 1.6% when surveillance was limited to(C-sections) for which the hospitalization period is brief.6 hospitalized patients and 9.6% when postdischarge sur- Few Brazilian hospitals conduct surveillance after veillance was implemented. Another study carried outthe patient has been discharged from hospital, limiting ˜ in a maternity hospital in the state of Sao Paulo found a post-C-section SSI rate of 1.2% when surveillance From the Department of Infection Controla and Department of was limited to hospitalized patients versus 14.4% Obstetrics and Gynecology,b Women’s Hospital, School of Medical when these rates were obtained from postdischarge sur- ˜ Sciences, University of Campinas (UNICAMP), Campinas, Sao Paulo, veillance.13 The study was conducted at the Center for Brazil. Women’s Integrated Health Care, a tertiary, 142-bed, Address correspondence to Meire Celeste Cardoso Del Monte, RN, university teaching and public hospital situated in the ´ ˜ Rua Dr. Liraucio Gomes, 257 Cambuı, 13.024-490 Campinas, Sao Paulo, Brazil. E-mail: email@example.com. ˜ state of Sao Paulo, Brazil. A mean of 100 C-sections are performed in this hospital monthly. This study was car- Conﬂicts of interest: None to report. ried out between May 2008 and March 2009. There has 0196-6553/$36.00 been a hospital infection control committee (HICC) in Copyright ª 2010 by the Association for Professionals in Infection this institute since 1986, and this committee currently Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved. recommends prophylactic antibiotic therapy following all C-section deliveries. This protocol determines the doi:10.1016/j.ajic.2009.10.008 use of cefazolin, 2 g intravenously, after the umbilical 467
468 Cardoso Del Monte and Pinto Neto American Journal of Infection Control August 2010cord has been clamped or 900 mg of clindamycin for Review Board of the institute prior to initiation of thepatients who are allergic to penicillin. study. If the woman agreed to participate, she then A recent systematic review of the methods used to signed the informed consent form, was given a copy,identify SSI following discharge from hospital and was admitted to the study.concluded that existing studies on the subject have so Diagnosis of SSI was deﬁned according to the criteriafar failed to identify a valid, reliable method for identi- standardized by the Centers for Disease Control and Pre-fying such infections postdischarge. On a local level, vention, Atlanta, GA, determining superﬁcial incisionalthe method used to identify postdischarge SSI is likely SSI, deep incisional SSI, or organ/space SSI.15 The inves-to be dependent on existing resources, on the objective tigator made a questionnaire, based in this criteria, fromof surveillance, and on the nature of the data routinely which questions about purulent discharge, identiﬁcationavailable.14 In this study, the researcher contacted the of an isolated organism, signs and symptoms of infectionpatients by telephone following discharge and estab- (fever, pain or tenderness, localized swelling, redness,lished the occurrence of a SSI by having the patients heat), an abscess or other evidence of infection involvingconﬁrm the presence or absence of symptoms associ- the deep incision, or diagnosis of SSI by attending physi-ated with a SSI. Therefore, the methodology used to cian were applied by telephone interview.determine postdischarge SSI should be considered as The following data were obtained from the patient’sa potential limitation of the study. chart: age, gestational age, whether the patient had The objectives of the present study were to evaluate undergone prenatal care and the number of prenatalthe incidence of SSI following discharge of the patient consultations she had attended, weight and height,from hospital through the use of telephone interviews parity, presence of community infection, rupturedand to identify factors associated with the presence of membranes at admission and the time of membranethese infections in a tertiary Brazilian university rupture, presence of diabetes or arterial hypertension,teaching hospital that is a regional referral hospital for duration of labor, indication for C-section, surgicalmaternal and child health care. wound classiﬁcation (clean, clean contaminated, con- taminated and dirty/infected), use of general anesthe-METHODS sia or other type of anesthesia, the American Society of Anesthesiologists (ASA) physical status classiﬁcation This is an observational, longitudinal, cohort study score, whether the C-section represented an emer-carried out using data collected from patient charts gency or elective surgery, duration of the surgery,and from interviews with patients with the objective volume of intrasurgical blood loss as calculated byof identifying post-C-section SSI. Sample size calcula- the anesthesiologist, whether any other procedurestion was based on the SSI rate of 0.7% registered by were carried out, and whether there was compliancethe HICC in 2006, established according to the epide- with the institutional protocol of prophylactic antibi-miologic surveillance of patients up to the time of their otic therapy. The intraoperative nursing chart wasdischarge from hospital, and the estimated SSI used to detect whether electrocauterization was per-incidence of around 23.5% obtained in a previously formed. To evaluate obesity, body mass index (BMI)conducted pilot study that included postdischarge was calculated for each patient and classiﬁed accordingsurveillance. A signiﬁcance level of 5% and a sampling to the following categories: BMI ,20, underweight; BMIerror of 8% were adopted; therefore, a minimum 20 to 24.9, ideal weight; BMI 25 to 29.9, overweight;sample of 108 women was required, with an estimated BMI .30, obese. C-sections were classiﬁed accordingconﬁdence interval (CI) of 15.5% to 31%. to the risk of SSI as 0, 1, 2, or 3 in accordance with The inclusion criteria consisted of having been the NNIS system.16submitted to a C-section after April 2008 and having A postdischarge telephone interview was conducteda telephone for contact. Exclusion criteria consisted by the investigator herself or by a student nurse specif-of death in the immediate postoperative period or the ically trained for this task. A structured questionnairepresence of any form of impairment that would ham- was used, and the estimated duration of the interviewper the patient’s ability to consent to participate in was 5 minutes. The questions were speciﬁcally de-the proposed study or affect postdischarge telephone signed to identify any signs of a SSI following the pa-contact. The criterion for discontinuation consisted of tient’s discharge from hospital. An initial contact wasbeing unable to contact the woman within the 30-day made 15 days after the C-section and a second contactfollow-up period, except if she had already reported 30 days after the surgery. A maximum of 3 attemptsan SSI at ﬁrst contact. The investigator approached were made to establish contact on both occasions.women who had undergone a C-section while they Patients who could not be contacted for the second in-were still in hospital and read them the informed con- terview were discontinued from the study unless an SSIsent form, which had been approved by the Internal was identiﬁed at the ﬁrst interview.
www.ajicjournal.org Cardoso Del Monte and Pinto Neto 469Vol. 38 No. 6Table 1. Infections based on hospital surveillance and 204 women admitted to the studyinfection postdischarge, 2008-2009 17 were discontinued 187 final sample – follow Forms of surveillance up 30 days after surgery In-hospital Post-discharge 44 SSI 143 no SSI n % n % P value 42 postdischarge 02 while in hospitalNo infection 185 98.9 145 77.54 95% 5%Infection present 2 1.06 42 22.45 ,.0001Total 187 100 187 100 39 in 1st contact 03 in 2nd contact 2 93% 07%NOTE. McNemar x test. In 4 patients in whom the infection became evident within the first 15 days, symptoms persisted 30 days after surgery Statistical analysis was performed using the Statisti- Fig 1. Selected sample.cal Analysis System, a statistical software program, ver-sion 9.02 (SAS Institute, Cary, NC). The McNemar x2 testand Fisher exact test were used, as applicable, for cate-gorical variables, and the Mann-Whitney test for numer- calculated from the date of the woman’s last menstrualical variables. Cox proportional hazards regression period or by ultrasonography when this date wasmodel was used to identify the risk factors associated unavailable, was 38.8 weeks (range, 27.2-45.4 weeks).with the occurrence of SSI. Relative risks and their Thirteen women in the sample had a community infec-respective 95% conﬁdence intervals were calculated, tion: 4 cases of diverse infections, 3 cases of urinaryP values ,.05 being considered statistically signiﬁcant. infections, 2 cases of chorioamnionitis, 2 cases of hep- atitis C, 1 case of HIV, and 1 case of HIV and concomi-RESULTS tant hepatitis C. There was no statistically signiﬁcant difference in the SSI rate in these patients. In the study A total of 204 women were admitted to the study. Of sample, 185 women (98.9%) had undergone prenatalthese, 17 were discontinued principally because of care, 130 (70.3%) of whom had attended more thandifﬁculty in establishing contact; therefore, the ﬁnal 7 prenatal consultations. With respect to BMI, 88sample consisted of 187 women. Of these, 44 (23.5%) women (47.05%) were classiﬁed as obese, 77developed a SSI. In 42 of 44 women (95.4%), the SSI ap- (41.17%) as overweight, and 17 (9%) as being of idealpeared after the patient was discharged from hospital weight or underweight. In 5 cases (2.7%), this informa-(Table 1). In 39 of 42 cases (93.9%), the infection was tion was unavailable. With respect to parity, 96 patientsdiagnosed during the ﬁrst telephone contact with the (51.3%) were nulliparas and 49 (26.2%) primiparas,patient (15 days after the C-section), whereas, in 3 whereas 26 patients (13.9%) had already had 2 chil-cases (7.1%), diagnosis was only made at the second dren, and 16 patients (8.5%) had given birth 3 orcontact (30 days after the C-section) (Fig 1). In 4 of 44 more times previously. In 98 cases (52.4%), the womenpatients (9%) in whom the infection became evident had intact membranes when surgery began; whereas,within the ﬁrst 15 days, symptoms persisted 30 days in 60 patients (32.1%), membrane rupture occurredafter surgery. In 3 cases, the patient was readmitted 12 hours or less prior to surgery; and, in 29 patientsto hospital for treatment. Infections classiﬁed as super- (15.5%), membranes ruptured more than 12 hoursﬁcial incisional SSI involving the skin and subcutane- prior to C-section. In 147 cases (78.6%), the womenous region occurred in 43 patients (97.7% of cases). went into labor in hospital, and, in these cases, laborMost of these cases were treated with 500 mg of ﬁrst- lasted for a mean of 14.6 hours (median, 9.3 hours;generation cephalosporin every 6 hours for 7 to 10 range, 1-76.3 hours). The principal characteristics ofdays, as prescribed by physicians at basic health care the study population are summarized in Table 2. Theunits. In one case, the infection, classiﬁed as an or- primary reasons for indicating C-section were asgan/space SSI, progressed in the form of an abscess follows: fetal distress in 47 cases (25.1%), repeat C-sec-that began in the pelvis and extended to the right hypo- tion in 26 cases (13.9%), functional dystocia in 26chondrium, and a subtotal hysterectomy had to be per- cases (13.9%), cephalopelvic disproportion in 22 casesformed. In this case, the patient required prolonged (11.8%), and breech presentation in 19 cases (10.2%).hospitalization for 20 days, including 6 days in an Of the women in this study, 40 were hypertensiveintensive care unit. This treatment was carried out at (21.4%), and 14 were diabetic (7.5%) individuals. Re-the Center for Women’s Integrated Health Care. garding the potential for contamination during surgery, The mean age of the women in this study was 27.3 only 2 cases were classiﬁed as infected following detec-years (range, 13-44 years). Mean gestational age, tion of chorioamnionitis during the C-section. The
470 Cardoso Del Monte and Pinto Neto American Journal of Infection Control August 2010Table 2. Selected characteristics of the study sampleVariables No. Mean 6SD Median Minimum MaximumAge, yr 187 27.38 6.85 27 13 44Gestational age (wk) 185 38.81 2.39 39 27.71 45.43BMI (kg/m2) 182 31.04 6.4 29.75 18.91 58.14Number of prenatal consultations 184 9.06 2.81 9 2 16Parity 187 0.83 1.07 0 0 5Time of membrane rupture* 89 13.71 29.24 6.42 0.58 264Duration of labor* 147 14.6 14.6 9.3 1 76.3Duration of surgery* 179 1.25 0.41 1.17 0.58 3.5SD, standard deviation.*Hours.Table 3. Variables signiﬁcantly associated with surgical site infection in the bivariate analysisVariables Category SSI, % P value RR 95% CINumber of prenatal consultations #7 37.0 .0047* 2.09 1.26-3.48 .7 17.7Arterial hypertension Yes 40 .0065* 2.07 1.25-3.43 No 19.3CI, conﬁdence interval; RR, relative risk; SSI, surgical site infection.*P , .05, x2 test.most commonly used type of anesthesia was a spinal in SSI between surgeries classiﬁed as 1 and 2; however,block in 107 cases (57.2%), followed by an epidural this calculation could not be made for a risk score ofin 56 cases (29.9%) and spinal block plus epidural in 3 because there was only 1 patient with this score. In20 cases (10.7%). Only 4 patients were submitted to the bivariate analysis, factors found to be signiﬁcantlygeneral anesthesia (2.1%). According to the ASA phys- associated with SSI were the number of prenatalical status classiﬁcation, 82 women were classiﬁed as consultations and the presence of hypertensionASA 1 and 82 as ASA 2, constituting 87.7% of the sam- (Table 3); however, only arterial hypertension remainedple. In addition, 16 women (8.6%) were classiﬁed as signiﬁcant in the multivariate analysis (Table 4).ASA 3 and 1 (0.5%) as ASA 4. In 6 patients, ASA classi-ﬁcation was not recorded. Of the 187 C-sections per- DISCUSSIONformed, 128 (68.4%) were elective surgeries, and 53(28.3%) were emergency. In 6 cases, there was no This study clearly shows that neglecting to performrecord of whether the surgery had been elective or postdischarge surveillance of women undergoing to aemergency. The mean duration of surgery was 1 hour C-section leads to under-notiﬁcation of SSI with respect25 minutes (range, 34.8 minutes to 3.5 hours). In the to this type of surgery. The fact that most of the casesmajority of cases in which the duration of surgery were detected after the patient was discharged fromwas longer than the mean, it was found to have been hospital is probably due to the brief period of hospitali-associated with another procedure (tubal ligation, zation (72 hours) associated with this type of surgerytotal abdominal hysterectomy, or repair of bladder during which time the infection is not yet apparent.laceration). With respect to intrasurgical blood loss, This ﬁnding is consistent with data published by other107 patients (57.2%) lost #700 mL of blood, whereas investigators in the international literature, citing67 patients (35.8%) lost more than 700 mL. In 13 cases postdischarge infection rates following C-sections that(6.9%), there was no record of blood loss. Electrocau- varied depending on the type of postdischarge surveil-terization was used during surgery in 112 cases lance implemented.5-14 It is important to say that the cri-(60%). Compliance with the institute’s protocol for pro- teria used for diagnosis for SSI in these studies was thatphylactic antibiotic therapy was conﬁrmed in 181 from the Centers for Disease Control and Prevention.cases (96.8%). With respect to the NNIS basic SSI risk A study carried out by Ferraz et al17 in 1995, involv-index, 116 cases were classiﬁed as 2 and 70 as 1, ing outpatients who had recently been submitted towhereas only 1 patient was classiﬁed as 3 and none surgery in Brazil, reported a postdischarge SSI rate ofas 0. There was no statistically signiﬁcant difference 91.4% following C-section. An interesting ﬁnding of
www.ajicjournal.org Cardoso Del Monte and Pinto Neto 471Vol. 38 No. 6Table 4. Variable signiﬁcantly associated with surgical site Multiple factors contribute to post-C-section SSI. Ininfection in multivariate analysis the present study, factors classically associated with post-C-section SSI such as obesity, premature ruptureVariable P value RR 95% CI of membranes, diabetes, and emergency C-sectionsPresence of arterial .01* 2.47 1.21-5.04 were not found to be signiﬁcantly associated.19 hypertension With respect to the number of prenatal consultationsCI, conﬁdence interval; RR, relative risk. attended by the patient, similar ﬁndings have been*P , .05, Cox proportional hazards regression with stepwise variable selection. reported by Killian et al (2001),19 whose study showed that having attended fewer than 7 prenatal consulta-the present study that deserves particular mention is tions was a factor that signiﬁcantly increased the riskthat, in 93% of the women who developed an SSI, of post-C-section endometritis. More prenatal consulta-the infection became apparent in the ﬁrst 15 days fol- tions are a guarantee that primary prevention methodslowing C-section. This ﬁnding corroborates data from are instituted that will consequently avoid many perina-similar studies carried out in Brazil. Ferraz et al17 tal complications including postsurgical infection.reported that post-C-section SSI becomes evident after Women should be counseled to control their weighta mean of 8.6 days following surgery; Couto et al12 and blood pressure. The increased risk of infectionshowed that 95% of cases of post-C-section SSI following a C-section in hypertensive women may beoccurred between the tenth and ﬁfteenth days follow- explained by the chronic alteration in peripheral blooding surgery; and Dantas13 stated that 95% of these supply as a result of increased vascular resistance. Thisinfections became apparent within 14 days following ﬁnding was also reported in a study with a large samplesurgery. Therefore, a 30-day follow-up period may be size (19,416 C-sections) carried out to identify riskunnecessary for the detection of post C-section SSI. factors for early SSIs in C-sections (diagnosed prior to The SSI rate found in the present study (23.5%) is discharge from hospital).20 The present study alsohigher than rates found in other studies in which postdi- sought to evaluate electrocauterization, which isscharge surveillance was performed. Although no con- commonly used in this institute, as a factor associatedsensus has yet been reached on the best methodology with SSI; however, no such association was found.for implementing postdischarge surveillance, telephone In conclusion, it is our belief that accurate knowledgecontact would appear to represent a low-cost technique of infection rates will help develop and implement mea-that requires minimal resources and is widely used. Nev- sures for the prevention and control of these infections.ertheless, Whitby et al18 found a low positive predictive In view of the relevance of the data obtained, the postdi-value (28.7%) for diagnoses made according to patient scharge surveillance methodology used in the presenttelephone reports, although the negative predictive study will be incorporated by this institute’s HICC. Quar-value was high (98.2%) compared with diagnosis made terly data analyses will continue, and feedback will beby an infection control nurse through direct examina- provided to those responsible for the area with thetion of the surgical incision in the patient following objective of stimulating a review of current practicesdischarge from hospital. and reducing SSI rates. Postdischarge surveillance involving telephonecontact with patients and using a well-deﬁned scriptof questions has been shown to constitute a very useful Referencesinstrument for the detection of SSI. It is important to 1. Astagneau P, Rioux C, Golliot F, Brucker G. INCISO Network Study ¨emphasize that this model of surveillance was possible Group. Morbity and mortality associated with surgical site infections:in this particular setting because the majority of the results from the 1997-1999 INCISO surveillance. J Hosp Infect 2001;study population had a telephone. In addition, the 48:267-74.women were receptive to telephone contact and 2. Rios J, Murillo C, Carrasco G, Humet C. Increase in costs attributableshowed that they felt valued by the attention given to to surgical infection after appendicectomy and colectomy. Gac Sanit 2003;17:218-25.them by the investigator. Considering a workload of 3. Kasatpibal N, Thongpiyapoom S, Narong MN, Suwalak N, Jamulitrat S.8 hours/day, this type of surveillance requires approxi- Extra charge and extra length of postoperative stay attributable to sur-mately 5% of the time of an infection control nurse. To gical site infection in six selected operations. J Med Assoc Thai 2005;ensure even more accuracy in SSI rates, the nurse could 88:1083-91.use this methodology as a screening tool, requesting 4. Kasatpibal N, Jamulitrat S, Chongsuvivatwong V. Standardized incidence rates of surgical site infection: a multicenter study in Thailand. Am J Infectthe patients who suspect they have an infection to Control 2005;33:587-94.come to the hospital for evaluation because identiﬁca- 5. Ward VP, Charlett A, Fagan J, Crawshaw SC. Enhanced surgical site infec-tion of the signs and symptoms of infection by the tion surveillance following caesarean section: experience of a multicentrepatients themselves may lead to overestimation. collaborative post-discharge system. J Hosp Infect 2008;70:166-73.
472 Cardoso Del Monte and Pinto Neto American Journal of Infection Control August 2010 6. Gravel-Tropper D, Oxley C, Memish Z, Garber GE. Underestimation 14. Petherick ES, Dalton JE, Moore PJ, Cullum N. Methods for identifying of surgical infection rates in obstetrics and gynecology. Am J Infect surgical wound infection after discharge from hospital: a systematic Control 1995;23:22-6. review. BMC Infect Dis 2006;6:170. 7. Creedy DK, Noy DL. Postdischarge surveillance after cesarean section. 15. Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR. Guideline Birth 2001;28:264-9. for prevention of surgical site infection, 1999. Hospital Infection 8. Barbut F, Carbonne B, Truchot F, Spielvogel C, Jannet D, Goderel I, et al. Control Practices Advisory Committee. Infect Control Hosp Epide- Surgical site infections after cesarean section: results of a ﬁve-year prospec- miol20):250-78. tive surveillance. J Gynecol Obstet Biol Reprod 2004;33(6 Pt 1):487-96. 16. Culver DH, Horan TC, Gaynes RP, Martone WJ, Jarvis WR, Emori 9. Mitt P, Lang K, Peri A, Maimets M. Surgical site infections following TG, et al. Surgical wound infection rates by wound class, operative cesarean section in an Estonian university hospital: postdischarge sur- procedure, and patient risk index. National Nosocomial Infections veillance and analysis of risk factors. Infect Control Hosp Epidemiol Surveillance System. Am J Med 1991;91:S152-7. 2005;26:449-54. 17. Ferraz EM, Ferraz AA, Coelho HS, Pereira Viana VP, Sobral SM, Vas-10. Johnson A, Young D, Reilly J. Caesarean section surgical site infection concelos MD, et al. Postdischarge surveillance for nosocomial wound surveillance. J Hosp Infect 2006;64:30-5. infection: does judicious monitoring ﬁnd cases? Am J Infect Control11. Reilly J, Allardice G, Bruce J, Hill R, McCoubrey J. Procedure-speciﬁc 1995;23:290-4. surgical site infection rates and postdischarge surveillance in Scotland. 18. Whitby M, McLaws ML, Collopy B, Looke DF, Doidge S, Henderson B, Infect Control Hosp Epidemiol 2006;27:1318-23. et al. Post-discharge surveillance: can patients reliably diagnose surgical12. Couto RC, Pedrosa TM, Nogueira JM, Gomes DL, Neto MF, Rezende wound infections? J Hosp Infect 2002;52:155-60. NA. Post-discharge surveillance and infection rates in obstetric patients. 19. Killian CA, Graffunder EM, Vinciguerra TJ, Venezia RA. Risk factors for Int J Gynaecol Obstet 1998;61:227-31. surgical-site infections following cesarean section. Infect Control Hosp ˆ x˜ ´ ´13. Dantas RHEA. Incidencia de infeccao do sıtio cirurgico de pacientes subme- Epidemiol 2001;22:613-7. ´ ˆ ˆ ´s tidas a parto cesarea; a importancia da vigilancia po alta [Tese -mestrado]. 20. Schneid-Kofman N, Sheiner E, Levy A, Holcberg G. Risk factors for ˜ ˜ Ribeirao Preto (SP): Universidade de Sao Paulo - Escola de Enfermagem de wound infection following cesarean deliveries. Int J Gynaecol Obstet ˜ Ribeirao Preto; 2001. 2005;90:10-5.