Surgical site infections lead to longer, costlier hospitalizations and increased mortality.1 They occur in 3–15% of patients after cesarean delivery and obesity is a known risk factor associated with the diagnosis.2,3 Antibiotic prophylaxis is one of the Centers for Disease Control and Prevention–recommended measures to prevent surgical site infections.4
Few studies have explored increased antibiotic dosing and tissues levels in obese patients.5–7 Edmiston et al5 demonstrated that patients having undergone gastric bypass who were administered 2 g cefazolin had significantly lower antibiotic tissue concentrations at the time of wound closure than the therapeutic breakpoint for susceptibility to cefazolin. In contrast, Ho et al6 demonstrated that nonpregnant obese patients undergoing elective surgery who were administered 2 compared with 3 g cefazolin maintained serum concentrations of cefazolin above the minimum inhibitory concentration for more than 3 hours regardless of body mass index (BMI, calculated as weight (kg)/[height (m)]2). Limited data exist in pregnancy, but they suggest that 2 g may not be sufficient to provide adequate antimicrobial coverage in obese patients.8 Importantly, all these studies were pharmacokinetic in nature; there are no data to suggest that increased antibiotic dosing affects clinical outcomes such as wound infection.
Hence, our primary objective was to elucidate whether 3 g cefazolin was associated with a reduced risk of developing an obstetric surgical site infection compared with 2 g cefazolin. Answering this question would help to evaluate whether an increase in the amount of cefazolin given preoperatively would benefit our patients. Secondary objectives included identifying other risk factors associated with surgical site infections in a morbidly obese obstetric population.
MATERIALS AND METHODS
This was a multicenter retrospective cohort study among a population of morbidly obese patients who underwent cesarean delivery at Johns Hopkins Hospital and Duke University from 2008 to 2013. The dose of preoperative cefazolin was increased from 2 g to 3 g for morbidly obese women at both institutions; however, the timing and implementation differed. In 2007, the Johns Hopkins Hospital Obstetric Committee recommended the increased dose for gravid women who weighed 300 pounds or more at the time of delivery. In 2010, Duke instituted a policy that all patients with BMIs 40 or higher at the time of delivery should receive the increased dose.
Institutional review board approval and a data use agreement contract were obtained for the chart reviews from both institutions. At Johns Hopkins Hospital, a list of patients weighing 250 pounds or more and clinical data were extracted from the electronic medical records. Although the recommended weight cutoff for increased dosing was 300 pounds at Johns Hopkins Hospital, during data abstraction it was noted that many patients who weighed at least 290 pounds received the higher dose. Therefore, we included all patients who weighed at least 290 pounds in the study at both sites. At Duke, a list of eligible patients were identified from the anesthesia medical record followed by review of their electronic medical records. Patients were excluded if they had multiple gestations, vaginal deliveries, penicillin or cephalosporin allergy, human immunodeficiency virus or acquired immunodeficiency syndrome, suspected chorioamnionitis, or antibiotic type not recorded in the medical record. Data were collected on patient ethnicity, height, weight, BMI, prior surgeries, chronic medical conditions, number of pregnancies, outcome of pregnancies, social history, name and dose of preoperative antibiotic used, group B Streptococcus status, type and number of doses of group B Streptococcus prophylaxis administered, laboring, duration of labor, use of internal monitors (eg, fetal scalp electrode or intrauterine pressure catheter), rupture of membranes and duration, type of cesarean delivery (ie, planned before admission [scheduled], performed after trial of labor [unscheduled], and performed immediately after decision [emergent]), surgery details (eg, incision type, wound closure, estimated blood loss), and postoperative variables (eg, length of stay, fever [defined as temperature 38°C or higher], antibiotic administration or prescription, readmission, wound complications). Medical record review included all available hospitalizations and outpatient clinic follow-up through 30 days postcesarean delivery.
The primary outcome was incidence of surgical site infections, which included superficial, deep, and organ space (ie, endometritis) infections. Surgical site infections were defined by the Centers for Disease Control and Prevention criteria (Box 1).9 At Johns Hopkins Hospital, an experienced infection control public health nurse in collaboration with the Obstetric Surgical Site Infection Committee chair reviewed all charts, which contained inpatient and outpatient visits, from women who delivered by cesarean and identified surgical site infection cases. At Duke, the study investigators identified surgical site infection cases from inpatient and outpatient records. The secondary outcomes were length of stay more than 4 days, readmission, or any infectious morbidity, which included surgical site infections, pneumonia, urinary tract infection, or receipt of postoperative antibiotics without an identified source of infection. Based on baseline surgical site infection data from Johns Hopkins Hospital, 15% of the patients who receive 2 g cefazolin would be diagnosed with a surgical site infection. We estimated that we needed a sample size of 141 patients in each group for a 67% reduction in baseline risk for surgical site infections (15% to 5%, equivalent to odds ratio [OR] 0.30) among those women given 3 g cefazolin (two-tailed α=0.05, β=0.20).
Box 1 Centers for Disease Control and Prevention Criteria for Surgical Site Infection Cited Here...
Superficial incisional surgical site infections
- Infection occurs within 30 days after the operative procedure and involves only skin and subcutaneous tissue of the incision and patient has at least one of the following:
- a. Purulent drainage from the superficial incision.
- b. Organisms isolated from culture of fluid or tissue from the superficial incision.
- c. Superficial incision that is deliberately opened and patient has at least one of the following signs or symptoms: pain or tenderness, localized swelling, redness, or heat.
- d. Diagnosis of a superficial incisional surgical site infection by the physician.
Deep incisional surgical site infections
- Infection occurs within 30 or 90 days after the operative procedure and involves deep soft tissues of the incision (eg, fascial and muscle layers) and patient has at least one of the following:
- a. Purulent drainage from the deep incision.
- b. a deep incision that spontaneously dehisces or is deliberately opened by a physician and patient has at least one of the following signs or symptoms: fever (higher than 38°C), localized pain, or tenderness.
- c. An abscess or other evidence of infection involving the deep incision that is detected on direct examination, during invasive procedure, or by histopathologic examination or imaging test.
Organ or space surgical site infections
- Infection occurs within 30 or 90 days after the operative procedure and infection involves any part of the body deeper than the fascial or muscle layers that is opened or manipulated during the operative procedure and patient has at least one of the following:
- a. Purulent drainage from a drain that is placed into the organ or space.
- b. Organisms isolated from an aseptically obtained culture of fluid or tissue in the organ or space.
- c. An abscess or other evidence of infection involving the organ or space that is detected on gross anatomical or histopathologic examination or imaging test.
Continuous variables were assessed for normality in distribution by box-and-whisker plot and Shapiro-Wilk test and were summarized with mean±standard deviation or median and interquartile range. Student's t test was used to compare normal variables. Wilcoxon rank-sum and Kruskal-Wallis tests were used to compare nonnormal variables. Binary or categorical variables were compared with χ2 test or Fisher's exact test. A priori, antibiotic dose and variables with P<.1 in univariable analyses were included in the multiple logistic regression models. Model accuracy was evaluated with Hosmer-Lemeshow goodness-of-fit test and area under the receiver operating characteristic curve. Collinearity was evaluated by variance inflation factor. The results of regression analyses were expressed as an adjusted OR and 95% confidence interval (CI). All the reported P values were based on two-sided tests and P<.05 was considered statistically significant. Missing data were not included in the analyses. Data analyses were performed with Stata 13.1.
There were 335 women included in the cohort with a median absolute weight of 310 pounds (interquartile range 299–333, range 290–439 pounds) and median BMI of 51.3 (interquartile range 47.8–55.6, range 39.3–74.4) (Table 1). Seventy-one percent of women were black and multiparous. One hundred nine women (32.5%) had diabetes, and among these women, 51 (46.8%) had pre-existing diabetes and 58 (53.2%) had gestational diabetes. One hundred thirty-eight women (41.2%) were in labor at the time of their cesarean delivery (Table 2). One hundred sixty women (47.8%) received 3 g cefazolin, which was comparable at both study sites. A majority of cesarean deliveries (54.6%) was unscheduled, whereas approximately one third (35.2%) were scheduled; a minority (10.2%) was emergent. The most common indications for unscheduled cesarean delivery were arrest of labor or failure to progress. A Pfannenstiel skin incision was chosen in 89.0% of women and the incision was closed with staples 44.6% of the time.
Forty-four (13.1%) women were diagnosed with a surgical site infection. The proportion of women diagnosed with a surgical site infection who received 2 g cefazolin was 13.1% (23/175), which was comparable with those who received 3 g (13.1% [21/160], P=.996). The proportion of women diagnosed with a surgical site infection at Johns Hopkins Hospital and Duke was 14.3% and 12.4%, respectively (P=.61). From 2008 to 2013, there were significant fluctuations in the percentage of surgical site infections at Duke but not at Hopkins (P for trend <.001 and =.55, respectively) (Fig. 1A). Figure 1B illustrates the difference between total surgical site infections differentiating wound infection from endometritis over time. There were 33 (9.9%) patients with a superficial or deep wound infection, 13 (3.9%) patients with endometritis, and two (0.6%) with both.
Univariable analyses are presented in Table 3. There was no association between surgical site infections and any demographic characteristics or medical illness. Repeat cesarean delivery, maternal group B Streptococcus-positive status, longer surgery times, higher dose antibiotics, and vertical skin incision were not significantly associated with surgical site infections. Laboring, internal monitoring, duration of rupture of membranes, estimated blood loss greater than 1,500 mL, and staple closure were associated with surgical site infections. In the multivariable analysis that adjusted for labor, internal monitoring, staple closure, scheduled cesarean delivery, estimated blood loss greater than 1,500 mL, and higher dose antibiotics, estimated blood loss greater than 1,500 mL (adjusted OR 3.32, 1.32–8.37) and staple closure (adjusted OR 2.45, 1.19–5.02) remained associated with an increased risk for a surgical site infection, whereas 3 g cefazolin was not associated with a reduction in the risk for surgical site infections (adjusted OR 1.33, 0.64–2.74) (Table 4). The goodness-of-fit test (10 groups, P=.85) and area under the receiver operating characteristic curve (0.72) demonstrated model accuracy; variance inflation factors for the included variables (range 1.05–2.49) supported a low likelihood of collinearity. To assess the potential of overfitting, we did a sensitivity analysis by varying the number of covariates in the multiple logistic regression models with and without locking the 3 g cefazolin variable; however, the results did not differ and model accuracy remained unchanged (data not shown). Hence, we adopted the model in which cefazolin was locked because it was the major exposure of interest. Notably, although rupture of membranes duration was significantly associated with surgical site infections in univariable analysis, we did not include it in the final multivariable model as a result of its statistical and clinical correlation with other covariates such as labor, internal monitoring, scheduled cesarean delivery, and estimated blood loss greater than 1,500 mL (data not shown). In addition, rupture of membranes duration was missing for 23 patients (6.9%) and tended to be missing among those with surgical site infections (20 in surgical site infection group compared with three in the nonsurgical site infection group). Hence, the missingness was nonrandom.
There were 35 patients (10.5%) with a length of stay longer than 4 days (8% in the 2 g cefazolin group compared with 13% in the 3 g cefazolin group, P=.11), 18 patients (5.4%) who were readmitted (3.4% in the 2 g cefazolin group compared with 7.5% in the 3 g cefazolin group, P=.14), and 49 patients (14.6%) who had any infectious morbidity (13.1% in the 2 g cefazolin group compared with 16.3% in the 3 g cefazolin group, P=.42).
In our study, 3 g cefazolin administered to morbidly obese gravid women before cesarean delivery did not reduce the risk of surgical site infections compared with 2 g. The Clinical Practice Guidelines for Antimicrobial Prophylaxis in Surgery recommend that adults weighing more than 120 kg receive 3 g cefazolin before cesarean delivery.10 Additionally, the American College of Obstetricians and Gynecologists states that higher than 1 g cefazolin dosing can be considered for women with BMIs higher than 30 or an absolute weight more than 100 kg.11
Our findings are supported by cefazolin pharmacokinetic studies among obese women undergoing cesarean delivery. Stitely et al8 randomized women with BMIs higher than 35 to 2 g or 4 g cefazolin before cesarean delivery and found adequate tissue concentrations in both groups albeit significantly higher after 4 g. Similarly, Maggio et al12 randomized women with BMIs higher than 30 to 2 g or 3 g cefazolin before cesarean delivery and they did not find a difference in tissue concentrations. Neither study showed a difference in wound complications; however, they were not adequately powered. Because cefazolin is a concentration-independent antibiotic, the length of time above minimum inhibitory concentration, and not peak concentration, best describes its bactericidal activity, which might explain these results.13
There was a decrease in endometritis over the study time period. Because endometritis can be a subjective diagnosis, this decrease may be related to a change in the diagnostic criteria or the change in antibiotic dose. Conversely, wound infection cases appeared similar over time. Although there were trends with regard to incidence over time, we were not powered to assess the chronologic wound infection trend or the effect of 2 compared with 3 g cefazolin on endometritis and wound infection separately.
In our multivariable model, estimated blood loss greater than 1,500 mL and staple closure significantly increased the risk for surgical site infections, which has been demonstrated in other studies. Increased BMI in pregnant women is associated with higher estimated blood loss during surgery.14 To prevent reductions in serum antibiotic levels, redosing of cefazolin is recommended with excessive estimated blood losses and prolonged operative times.10 However, only a small minority of patients in our cohort was redosed, which could have affected surgical site infection risk. Moreover, the literature available on skin closure method is inconclusive. A Cochrane review found no difference in wound infection between staple and subcuticular skin closure.15 However, there were not enough data from the available studies to analyze the effect of BMI. In contrast, two recent randomized controlled trials among women with BMIs higher than 30 showed that subcuticular closure reduced the risk of postoperative wound complications.16,17 However, when wound complications were broken down to look at disruption compared with infection, there was no difference in infection. Therefore, although the data are mixed, our study suggests that there might be a true increase in infection among morbidly obese women with staple closures.
Our study has many strengths. This is one of the first studies to report clinical outcomes after 3 g preoperative cefazolin in morbidly obese women undergoing cesarean delivery. Although pharmacokinetics are important to consider in this population, without pharmacodynamic data, it is unclear how these pharmacokinetic parameters relate to clinical outcomes. Second, the study included 335 women who are morbidly or supermorbidly obese, which no other study has done in the past and has allowed us to examine clinical outcomes among a population at especially high risk of infection. Third, this was a multicenter study, which broadens the basis for generalization of the findings.
There are limitations. This was an observational study using data obtained from medical records, which are subject to information or misclassification bias. Next, the implementation of 3 g cefazolin differed between sites and there may have been other institutional changes that were not identified. Additionally, the weight cutoff of 290 pounds may have excluded some women who are classified as morbidly obese. Next, our initial estimate to detect a 67% reduction in surgical site infections was ambitious. However, to demonstrate a 25% or 50% decrease in surgical site infections from the baseline of 13%, a total sample size of 3,096 or 650 morbidly obese women, respectively, is required. Regardless, the incidence of surgical site infections was nearly identical in both antibiotic groups at 13.1%. A randomized trial to address this question would require an extraordinarily high number of morbidly obese women and may not be practical. Last, there may have been ascertainment bias and we missed patients who did not inform health care providers of a surgical site infection that was diagnosed at another hospital. By using only outpatient records to identify surgical site infections that occurred after discharge, there is the possibility that some infections were missed, if patients sought care outside of the Johns Hopkins Hospital and Duke systems. Therefore, the rates of surgical site infections may be higher than identified.
Other studies are needed to replicate our findings for external validation. Regardless, the absolute surgical site infection risk of 13% in our population is too high given the cost and morbidity associated with these infections. Additional methods such as alternative dosing regimens or postoperative negative pressure wound therapy to help reduce this risk should be investigated.
1. Kirkland KB, Briggs JP, Trivette SL, Wilkinson WE, Sexton DJ. The impact of surgical-site infections in the 1990s: attributable mortality, excess length of hospitalization, and extra costs. Infect Control Hosp Epidemiol 1999;20:725–30.
2. 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.
3. Tran TS, Jamulitrat S, Chongsuvivatwong V, Geater A. Risk factors for postcesarean surgical site infection. Obstet Gynecol 2000;95:367–71.
4. Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR. Guideline for prevention of surgical site infection, 1999. Hospital Infection Control Practices Advisory Committee. Infect Control Hosp Epidemiol 1999;20:250–78.
5. Edmiston CE, Krepel C, Kelly H, Larson J, Andris D, Hennen C, et al.. Perioperative antibiotic prophylaxis in the gastric bypass patient: do we achieve therapeutic levels? Surgery 2004;136:738–47.
6. Ho VP, Nicolau DP, Dakin GF, Pomp A, Rich BS, Towe CW, et al.. Cefazolin dosing for surgical prophylaxis in morbidly obese patients. Surg Infect (Larchmt) 2012;13:33–7.
7. Douglas A, Udy AA, Wallis SC, Jarrett P, Stuart J, Lassig-Smith M, et al.. Plasma and tissue pharmacokinetics of cefazolin in patients undergoing elective and semielective abdominal aortic aneurysm open repair surgery. Antimicrob Agents Chemother 2011;55:5238–42.
8. Stitely M, Sweet M, Slain D, Alons L, Holls W, Hochberg C, et al.. Plasma and tissue cefazolin concentrations in obese patients undergoing cesarean delivery and receiving differing preoperative doses of drug. Surg Infect (Larchmt) 2013;14:455–9.
10. Bratzler DW, Dellinger EP, Olsen KM, Perl TM, Auwaerter PG, Bolon MK, et al.. Clinical practice guidelines for antimicrobial prophylaxis in surgery. Am J Health Syst Pharm 2013;70:195–283.
11. Use of prophylactic antibiotics in labor and delivery. Practice Bulletin No. 120. American College of Obstetricians and Gynecologists. Obstet Gynecol 2011;117:1472–83.
12. Maggio LND, DaCosta M, Rouse DJ, Hughes BJ, Hughes BL. Cefazolin prophylaxis in obese women undergoing cesarean delivery. Obstet Gynecol 2015;125:1205–10.
13. Craig WA. Pharmacokinetic/pharmacodynamic parameters: Rationale for antibacterial dosing of mice and men. Clin Infect Dis 1998;26:1–10.
14. Conner SN, Verticchio JC, Tuuli MG, Odibo AO, Macones GA, Cahill AG. Maternal obesity and risk of postcesarean wound complications. Am J Perinatol 2014;31:299–304.
15. Mackeen AD, Berghella V, Larsen ML. Techniques and materials for skin closure in caesarean section. The Cochrane Database of Systematic Reviews 2012, Issue 11. Art. No.: CD003577. DOI: 10.1002/14651858.CD003577.pub3.
16. Figueroa D, Jauk VC, Szychowski JM, Garner R, Biggio JR, Andrews WW, et al.. Surgical staples compared with subcuticular suture for skin closure after cesarean delivery: a randomized controlled trial. Obstet Gynecol 2013;121:33–8.
© 2015 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
17. Mackeen AD, Khalifeh A, Fleisher J, Vogell A, Han C, Sendecki J, et al.. Suture compared with staple skin closure after cesarean delivery: a randomized controlled trial. Obstet Gynecol 2014;123:1169–75.