Obstetrics & Gynecology:
Obesity as an Independent Risk Factor for Infectious Morbidity in Patients Who Undergo Cesarean Delivery
Myles, Thomas D. MD; Gooch, Jason MD; Santolaya, Joaquin MD, PhD
Department of Obstetrics and Gynecology, Texas Tech Health Sciences Center at Amarillo, Amarillo, Texas.
Address reprint requests to: Thomas D. Myles, MD, St. Louis University, 6420 Clayton Road, Suite 559, St. Louis, MO 63117; E‐mail: firstname.lastname@example.org.
Received January 28, 2002. Received in revised form April 11, 2002. Accepted May 2, 2002.
OBJECTIVE: Our purpose was to evaluate obesity (body mass index greater than 30.0) as an independent risk factor for infectious morbidity in women having elective or non‐elective cesarean deliveries.
METHODS: Charts of 611 patients undergoing cesarean were reviewed. After exclusion of those with pre‐existing chorioamnionitis, 574 cases were separated into two groups (elective or nonelective cesarean) and then subdivided based on the presence or not of postdelivery infectious morbidity. Estimated blood loss, operative time, number of vaginal examinations, labor length, use of internal monitors, body mass index (BMI), and obesity (BMI greater than 30.0) were then recorded. Student t test, χ2, multivariate analysis, and receiver operating characteristics curves were used where appropriate (significance: P < .05).
RESULTS: The mean gestational age at delivery was 38.3 weeks. Three hundred sixty patients had nonelective cesareans, and 214 had elective cesareans. Prophylactic antibiotics were used for 86.6% of the nonelective group and 75.2% of the elective group. In the nonelective group and after multivariate analysis, significant risk factors for postoperative infections were as follows: labor length (18.4 hours versus 10.9, P < .003), number of vaginal examinations (6.1 versus 4.5, P < .001), BMI (36.6 versus 32.3, P < .001), and obesity (81.8% versus 57.3%, P < .001). For the elective group, a higher BMI (38.9 versus 32.2, P < .003), and black race (63.2% versus 11.5%, P < .001) were found to be significant.
CONCLUSION: Our data suggest that obesity is a independent risk factor for postcesarean infectious morbidity and endomyometritis, even if the cesarean is elective and prophylactic antibiotics are given.
Postoperative infectious morbidity often prolongs hospitalization and increases expenses, owing to the need for additional diagnostic studies and treatment. Postoperative infectious morbidity in women undergoing cesarean delivery can be reduced by allowing for spontaneous separation of the placenta,1,2 not closing the visceral peritoneum,3 as well as by the use of a closed suction drain in the morbidly obese women.4 There is no true consensus as to whom to administer antibiotics5; however, prophylactic antibiotics reduce the incidence of infectious morbidity after nonelective cesarean deliveries and may also be beneficial to patients undergoing elective cesarean.6–9
A variety of risk factors for infectious morbidity with nonelective cesarean delivery include the following: pre‐existing infections, such as gonorrhea, chlamydia, and bacterial vaginosis; length of labor; length of membrane rupture; number of vaginal examinations10; presence of meconium11,12; primary indication for surgery; type of anesthesia; total operative time; operator experience; and procedure‐related blood loss.10 The effect of internal fetal monitoring is controversial.10,13 Maternal obesity (250–300 lbs) or increased body mass index (BMI; kg/ m2) are also risk factors for postcesarean infectious morbidity (nonelective cases)14–16; however, it was not clear in our literature review (MEDLINE, 1966–April 2002) if prophylactic antibiotics could neutralize the risk of infectious morbidity in obese women undergoing elective cesarean. Descriptions of risk factors for infectious morbidity in patients undergoing an elective cesarean are limited, based on our MEDLINE review.
In this study we sought to determine whether obesity was an independent risk factor for infection in women having both elective and nonelective caesareans. We also sought to evaluate for additional risk factors for both of these cesarean categories.
MATERIALS AND METHODS
This retrospective, institutional review board–approved study was conducted at Northwest Texas Hospital in Amarillo. The hospital charts and outpatient records for all patients (n = 611) who underwent cesarean between January 1, 2000 and August 2001 were reviewed for evidence of infectious morbidity related to surgery. Thirty‐seven patients with documented pre‐existing chorioamnionitis were excluded from further analysis. Basic demographic information, including age, race, gravidity, parity, gestational age at delivery, BMI, tobacco or illicit drug use, insurance type, and the presence or absence of diabetes (including 50‐g glucola results) was collected from the remaining 574 charts. Evidence of postoperative infectious morbidity requiring antimicrobial therapy or interventional therapy (wound debridement) and occurring 24 hours or more after surgery was then documented. Febrile morbidity was defined as two (or more) temperature readings of 38C or more obtained 4 or more hours apart. Evidence for morbidity included wound infection, endomyometritis, urinary tract infection, pneumonia, septic pelvic thrombophlebitis, fever of unknown origin, and readmission for postoperative infection. Discharge records, diagnostic studies, and clinical evaluations were used as evidence for each of the entities, where appropriate. Risk factors for infections investigated included the previously mentioned variables, as well as the indication(s) for surgery, number of digital examinations, length of labor, length of membrane rupture, duration of surgery, type of anesthesia, bladder catheterization, presence of meconium, use of antibiotic prophylaxis, type of skin and uterine incision, use of internal fetal or uterine pressure monitors, estimated blood loss at surgery and postpartum hemorrhage (defined as estimated blood loss greater than 1000 mL), and professional level of primary surgeon. Very few patients had subcutaneous drains, so this could not be evaluated. Newborn records were evaluated for Apgar scores, birth weight, umbilical artery cord pH, neonatal intensive care unit (NICU) admissions, and neonatal sepsis. Maternal obesity was defined as a BMI greater than 30.0. Prophylactic antibiotics were given at the discretion of the attending physicians; the timing of antibiotic administration was also at their discretion. The exact time of antibiotic administration was not uniformly documented. Patients with known group B streptococci or group B streptococci risk factors were given antibiotics prior to delivery.
Cesarean patients were separated into two categories: elective (n = 214) (patients who had reassuring fetal heart tones and did not have active labor or rupture of membranes prior to surgery) and nonelective (n = 360) (everyone else). Comparisons between the various risk factors collected and the occurrence of some type of infectious morbidity were made. Statistical comparisons were made using Student t test, χ2 test of association, and multivariate analysis where appropriate. Significance was set at P < .05. SPSS 9.0 (SPSS Inc., Chicago, IL) was used for all statistical comparisons. Relative risks with 95% confidence intervals were determined where appropriate. Receiver operating characteristic curves were investigated as well for the significant factors (after multivariate analysis). Cut points that allowed us to obtain the most optimal sensitivities and specificities for each were determined. Confidence intervals were then calculated.
Table 1 contains the demographic data for both groups. The mean gestational age at delivery was 38.3 weeks. Patients who were in the elective cesarean group had the greater number of previous cesarean deliveries and were also more likely to be white. Table 2 contains the various risk factors for postcesarean infection. Patients in the nonelective cesarean group had a greater number of digital examinations, meconium‐stained amniotic fluid, and usage of antibiotic prophylaxis (82.2%), whereas those in the elective cesarean section group had a greater number of vertical skin incisions and operations performed by attending physicians. All cesareans were included. The most common indications for cesarean were cephalopelvic disproportion, fetal indication, malpresentation, and elective repeat cesarean.
Of the 574 patients studied, a total of 85 patients (14.8%) were found to have infectious morbidity requiring treatment: 67 patients (14.1%) developed endomyometritis; 20 (3.5%) wound infections; nine had documented urinary tract infections; four patients had septic pelvic thrombophlebitis; and two patients developed pneumonia. Some patients developed more than one type of infection. Most cases of infectious morbidity were in the nonelective cesarean group (n = 66 nonelective, n = 19 for elective group) (18.3% versus 8.9%, P < .001; relative risk [RR] = 0.4 [95% confidence interval (CI) 0.3, 0.7]). Twelve cases of endomyometritis occurred in the elective group, whereas 55 were in the nonelective group (P < .001; RR = 0.3 [95% CI 0.2, 0.6]). Two wound infections, one urinary tract infection, and one case of septic pelvic thrombophlebitis occurred in the elective cesarean group. One case of pneumonia occurred in each group. No significant differences were seen between cesarean groups with respect to these subgroups of infection.
Table 3 shows the risk factors that showed significant differences in patients having a nonelective cesarean depending on the development or not of an infection. Variables tested but not found to have any significant differences include length of membrane rupture, type of anesthesia, length of bladder catheterization, presence of meconium‐stained amniotic fluid, use of prophylactic antibiotics, type of skin incision, type of uterine incision, estimated blood loss, postpartum hemorrhage, Apgar scores, birth weight, umbilical artery cord pH, NICU admission, maternal smoking, drug abuse, insurance type, presence of maternal diabetes, maternal age, and gestational age at delivery. Nonwhite ethnicity, obesity, greater BMI, longer time in labor, greater number of digital examinations, professional level of the operator, and primary cesarean were all more frequently present in those developing infection. The RR for an obese patient to develop an infection postcesarean was 3.0 (95% CI 1.6, 5.8). Obese patients were more likely to develop endomyometritis (15.9% versus 5.0%, P < .001; RR = 3.3 [95% CI 1.6, 6.9]). Though not significant, 75% of the wound infections also occurred in the obese patient group. In addition, when evaluating only patients who received prophylactic antibiotics (n = 265), obese patients still had the greater risk of developing endomyometritis (14.5% versus 4.4%, P < .001; RR = 3.9 [95% CI 1.7, 9.0]). The risk for developing any infection also remained higher even when antibiotic prophylaxis was provided for the obese patient (23.4% versus 8.5%, P < .001; RR = 3.3 [95% CI 1.6, 6.8]). No differences were seen in infection rates between obese and nonobese patients who were not given antibiotic prophylaxis (n = 47) (23.1% versus 14.2%, RR = 1.8 [95% CI 0.4, 8.3]).
Patients undergoing a cesarean for cephalopelvic disproportion or failure to dilate (n = 134) (as opposed to all other indications) had a higher incidence of infectious morbidity (27.1% versus 11.9%, P < .001; RR = 2.7 [95% CI 1.5, 4.9]). The incidence of infection in patients undergoing cesarean for cephalopelvic disproportion (n = 90) was even higher than in those who had other cesarean indications (34.4% versus 10.8%, P < .001; RR = 4.3 [95% CI 1.4, 13.6]).
Only BMI, maternal obesity, length of labor, and the mean number of digital cervical examinations remained significant as potential risk factors of postcesarean infectious morbidity after multivariate analysis. Receiver operating characteristic (ROC) curves were determined for BMI, length of labor, and the number of digital examinations. For BMI, the area under the curve was 0.678 (0.602–0.753, P < .001). For a BMI greater than 30, the sensitivity was 85.1%, and the specificity was 55.8%. With respect to prediction of infectious morbidity, the ROC curve for labor length suggested a cut point of 10 hours for increased risk for infectious morbidity (sensitivity 67.3%, specificity 50.0%). The area under the curve was 0.663 (0.579–0.747, P < .001). A cut point of four or more examinations was associated with an increased risk for infection (sensitivity 72.7%, specificity 55.4%). The area under the curve for digital examination and the occurrence of infectious morbidity was 0.614 (0.529–0.699, P < .001).
Table 4 contains the significant risk factors for infectious morbidity for those patients having an elective cesarean. Other variables tested but not showing significant difference included number of digital examinations, gestational age at delivery, type of skin incision, maternal tobacco usage, maternal insurance type, maternal drug abuse, maternal diabetes, type of anesthesia, total operative time, length of bladder catheterization, presence of meconium‐stained amniotic fluid, estimated blood loss, incidence of postpartum hemorrhage, birth weight, Apgar scores, umbilical artery cord pH, and NICU admission. The increased risks for infectious morbidity associated with maternal race (black) and BMI remained after multivariate analysis. No risk was seen with respect to professional level of the primary surgeon when one controls for the use of antibiotic prophylaxis. The use of prophylactic antibiotics reduced the risk for postoperative infection in the residents' cases. If one controls only for obesity, more postcesarean infections were seen with the residents' patients. This risk could be race‐related, as black patients had higher rates of infection. The relative risk for postoperative morbidity in an obese patient having an elective cesarean is 1.6 (95% CI 1.2, 2.0). For the obese patient not given prophylactic antibiotics, the risk is increased significantly to 2.5 (95% CI 1.3, 4.9). The ROC curve for BMI and the occurrence of some type of infectious morbidity was determined. For a BMI greater than 30.0, the sensitivity was 90.0%, and the specificity was 58.0%. The area under the curve was 0.737 (0.613–8.60, P < .001).
Based on our MEDLINE search of literature from 1966 to April 2002, our finding of an increased risk for infectious morbidity in obese patient regardless of elective or nonelective nature of cesarean had not been previously reported. Knowledge that the risk of infectious morbidity is increased 2–3 times in obese patients undergoing cesarean is important. Our study has confirmed the reports of both Tran and colleagues14 and Perlow and Morgan,15 who each noted an increased incidence of postcesarean infectious morbidity in the nonelective delivered obese patient. Chauhan et al17 noted more infectious morbidity in patients of 300 or more pounds who had a failed vaginal birth after cesarean delivery attempt when compared with a similar group having an elective repeat cesarean delivery. Their numbers were limited, and no control group was described. Our postcesarean infectious morbidity rate of 14.8% in obese patients is within the 12.4%–23.3% range previously reported.14 Our study also confirms that, in spite of the use of prophylactic antibiotics, obese patients have higher rates of infectious morbidity. This raises clinical questions related to current prophylactic doses, the time of administration, and perhaps the need for a more prolonged administration of antibiotics. Because obesity was the most important risk factor for postoperative infection identified in this study, and because it is one of the few risk factors that cannot be avoided when a patient presents in labor, education to minimize excessive prenatal weight gain, recommendations for preconceptual weight loss, and studies investigating alternative prophylactic treatments are needed. The ROC curves generated in this study would seem to suggest using a BMI cut point of 30.0 to design these studies.
Knowledge of who is at increased risk for infectious morbidity should impact labor management. Early intervention with oxytocin has been shown to reduce infectious morbidity.18,19 The risk for increased intrauterine infections associated with more frequent digital examinations of the cervix has also been described.20,21 Steps to reduce the frequency of digital examinations of the cervix would also seem warranted, particularly in light of the work of Imseis and coworkers.22 They found that a single digital examination nearly doubled the number of different types of organisms cultured from the cervix. Our data suggest that attempts to both shorten the length of labor to less than 10 hours by administration of oxytocin and performing less than four digital examinations could decrease the risk of infection if the patient requires a cesarean.
We also noted an increased risk of infection in black women undergoing elective cesarean. Corrections for obesity and antibiotic usage did not eliminate this risk. This would seem to be consistent with a review of the literature by Fiscella23; however, the increased risk for infectious morbidity for blacks was lost after multivariate analysis in the nonelective cesarean group. A possible explanation for the higher infection rates seen in patients of this ethnicity may have to do with higher colonization rates for group B streptococci.24,25 Because less than 50% of our patients had group B streptococci screening performed, adequate comparisons could not be made for this factor; however, all patients were treated preoperatively if either a positive culture result or presence of group B streptococci risk factors were found in accordance with the recommendations of the American College of Obstetricians and Gynecologists.26 We did not make comparisons between groups for chlamydia, gonorrhea, or bacterial vaginosis. Patients found to have chlamydia or gonorrhea prenatally were treated at diagnosis. Test of cures were done after treatment. We were not able to locate any studies suggesting a different frequency of these infections in the obese patient; however, the higher postoperative infection rates seen in this study for the obese population would also suggest further evaluations for these pathogens.
We did not evaluate the method of placental delivery, owing to poor documentation. The same holds true for the management of the visceral peritoneum and timing of antibiotic administration. Very few subcutaneous drains were documented in the hospital records, so no analysis could be done for this variable. The importance of each of these infection‐preventing techniques necessitates further evaluation in the obese patient. Until such studies are made, one could consider making use of each of these techniques in order to further minimize the risk of postoperative infection in the obese patient.
1. Atkinson MW, Owen J, Wren A, Huth JC. The effect of manual removal of the placenta on post-cesarean endometritis. Obstet Gynecol 1996;87:99–102.
2. Lasley DS, Eblen A, Yancey MK, Duff P. The effect of placental removal method on the incidence of postcesarean infection. Am J Obstet Gynecol 1997;176:1250–4.
3. Nagele F, Karas H, Sptizer D, Staudach A, Karasegh S, Beck A, et al. Closure or nonclosure of the visceral peritoneum at cesarean delivery. Am J Obstet Gynecol 1996; 174:1366–70.
4. Allaire AD, Fisch J, McMahon MJ. Subcutaneous drain vs. suture in obese women undergoing cesarean delivery: A prospective, randomized trial. J Reprod Med 2000;45:327–31.
5. Pedersen TK, Blaaker J. Antiobiotic prophylaxis in cesarean section. Acta Obstet Gynecol Scand 1996;75:537–9.
6. Baker ER, D'Alton ME. Cesarean section birth and cesarean hysterectomy. Clin Obstet Gynecol 1994;37:806–15.
7. Spinnato JA, Youkilis B, Cook VD, Pietrantoni M, Clark AL, Gall SA. Antibiotic prophylaxis at cesarean delivery. J Matern Fetal Med 2000;9:348–50.
8. Chelmow D, Ruehli M, Huang E. Prophlactic use of antibiotics for nonlaboring patients undergoing cesarean delivery with intact membranes: A meta-analysis. Am J Obstet Gynecol 2001;184:656–61.
9. Jakobi P, Weissman A, Sigler E, Margolis K, Zimmer EZ. Post-cesarean section febrile morbidity antibiotic prophylaxis in low-risk patients. J Reprod Med 1994;39:707–10.
10. Yonekura ML. Risk factors for postcesarean endomyometritis. 1985;78:178–87.
11. Adair CD, Ernest JM, Sanchez-Ramos L, Burrus DR, Boles ML, Veille JC. Meconium-stained amniotic fluid-associated infectious morbidity: A randomized, double-blind trial of ampicillin-sulbactam prophylaxis. Obstet Gynecol 1996;88:216–20.
12. Piper JM, Newton ER, Berkus MD, Peairs WA. Meconium: A marker for peripartum infection. Obstet Gynecol 1998;91:741–5.
13. Gibbs RS, Listwa HM, Read JA. The effect of internal fetal monitoring on maternal infection following cesarean section. Obstet Gynecol 1976;48:653–58.
14. Tran TS, Jamulitrat S, Chongsuvivatwong V, Geater A. Risk factors for postcesarean surgical site infection. Obstet Gynecol 2000;95:367–71.
15. Perlow JH, Morgan MA. Massive maternal obesity and perioperative cesarean morbidity. Am J Obstet Gynecol 1994;170:560–5.
16. Martens MG, Kolrud BL, Faro S, Maccato M, Hammil H. Development of wound infection or separation after cesarean delivery. Prospective evaluation of 2,431 cases. J Reprod Med 1995;40:171–5.
17. Chauhan SP, Magann EF, Carroll CS, Barrilleaux PS, Scardo JA, Martin JN. Mode of delivery for the morbidly obese with prior cesarean delivery: Vaginal versus repeat cesarean section. Am J Obstet Gynecol 2001;185:349–54.
18. Hannah ME, Ohlsson A, Farine D, Hewson SA, Hodnett ED, Myher TL, et al. Induction of labor compared with expectant management for prelabor rupture of the membranes at term. N Engl J Med 1996;334:1005–10.
19. Hannah ME, Ahlsson A, Wang E, Matlow A, Foster GA, Willan AR, et al. Maternal colonization with group B streptococcus and prelabor rupture of membranes at term: The role of induction of labor. Am J Obstet Gynecol 1997;177:780–5.
20. Lewis DF, Major CA, Towers CV, Asrat T, Harding JA, Garite TJ. Effects of digital vaginal examinations on latency period in preterm premature rupture of membranes. Obstet Gynecol 1992;80:630–4.
21. Soper DE, Mayhall CG, Froggatt JW. Characterization and control of intraamniotic infection in an urban teaching hospital. Am J Obstet Gynecol 1996;175:304–10.
22. Imseis HM, Trout WC, Gabbe SG. The microbiologic effect of digital cervical examination. Am J Obstet Gynecol 1999;180:578–80.
23. Fiscella K. Race, perinatal outcome, and amniotic infection. Obstet Gynecol Surv 1995;51:60–6.
24. Reisner DP, Hass NJ, Zingheim RW, Williams MA, Luthy DA. Performance of a group B streptococcal prophylaxis protocol combining high-risk treatment and low-risk screening. Am J Obstet Gynecol 2000;182:1335–43.
25. Newton ER, Butler MC, Shain RN. Sexual behavior and vaginal colonization by group B streptococcus among minority women. Obstet Gynecol 1996;88:577–82.
26. Locksmith GJ, Clark P, Duff P. Maternal and neonatal infection rates with three different protocols for prevention of group B streptococcal disease. Am J Obstet Gynecol 1999;180:416–22.
© 2002 The American College of Obstetricians and Gynecologists