Objective
The purpose of this study was to compare surgical-site infection rates in obese women who had extended prophylactic antibiotic (EPA) vs standard prophylactic antibiotic.
Study Design
An electronic records-linkage system identified 145 obese women (body mass index, >30 kg/m 2 ) who underwent combined hysterectomy and panniculectomy from January 1, 2005, through December 31, 2008. The EPA cohort received standard antibiotics (cefazolin, 2 g) and continued oral antibiotic (ciprofloxacin) until removal of drains. Regression models were used to adjust for known confounders.
Results
The mean age was 56.0 ± 12.1 years, and mean body mass index was 42.6 ± 8.4 kg/m 2 (range, 30–86.4 kg/m 2 ). The EPA cohort experienced fewer surgical-site infections (6 [5.9%] vs 12 [27.9%]; P < .001; adjusted odds ratio, 0.16; 95% confidence interval, 0.04–0.51; P < .001), had lower probability of incision and drainage (3 [2.9%] vs 5 [11.6%]; P = .05), and required fewer infection-related admissions (5 [4.9%] vs 6 [13.9%]; P = .08).
Conclusion
Extended antibiotic prophylaxis can reduce surgical-site infections in obese women after combined hysterectomy and panniculectomy.
Obesity is a major public health problem in the United States. In 2004, the prevalence of obesity among women ≥20 years old was 33%, and 1 in every 14 women had a body mass index (BMI) ≥40 kg/m 2 . With this increase, gynecologic surgeons now face challenges that are associated with surgery in this high-risk cohort of women on a daily basis. Obesity limits access to the pelvis and increases associated morbidity after hysterectomy because of increased operative time, higher blood loss, increased number of wound infections, and febrile complications.
Panniculectomy was first described by Kelly in 1910. Although first performed as a plastic surgical procedure, panniculectomy has been used in gynecologic practice to increase accessibility to the pelvis in morbidly obese patients. The procedure is well-tolerated and associated with high patient satisfaction. Over the past 2 decades, combined panniculectomy and abdominal hysterectomy gradually gained popularity as a result of its advantage of better exposure with comparable operative time and blood loss to other incisions. Those advantages are of special importance for morbidly obese patients with cancer. In these patients, combined panniculectomy provides adequate exposure and improved lymph node yield.
Surgical-site infection (SSI) remains the most frequent complication after panniculectomy and ranges from 2.3–33.3%. Several modifications of the technique, which include the use of closed subcutaneous drains, resulted in decreased infection rate after the combined procedure. Prophylactic antibiotic use is associated with decreased SSI rates after hysterectomy. Despite the known increased risk of infection in morbidity obese patients, the current guidelines for antibiotic prophylaxis do not address this issue adequately. In this cohort study, we compared the rate of SSIs and other infection-related morbidities in obese women who had combined hysterectomy and panniculectomy who had extended prophylactic antibiotic therapy compared with those who received the standard single-dose prophylactic antibiotic after adjustment for other known risk factors for infection.
Materials and Methods
This retrospective cohort study was approved by the Mayo Foundation Institutional Review Board. The study is reported according to the strengthening the reporting of observational studies in epidemiology initiative.
Identification of the study cohorts
The medical records linkage system of Mayo Clinic was used to identify women who underwent combined abdominal hysterectomy and panniculectomy from January 1, 2005, through December 31, 2008. Only patients who gave their consent for the use of their medical records were included. To minimize variability in surgical technique, we included only patients who had hysterectomy by 1 specific surgical service (C.R.S.). The surgical technique was published previously. Standard antibiotic prophylaxis was administered to all patients and included intravenous cefazolin (2 g) within 60 minutes of the incision. In cases of known allergy to cefazolin, clindamycin or levofloxacin was used. Additional doses were provided if operative time was >4 hours or if high blood loss was documented, per guidelines. In all patients, 2 subcutaneous drains were inserted and used for at least 2 weeks after surgery. An abdominal binder was used for 6 weeks. The extended antibiotic prophylaxis cohort (exposed cohort) included women who continued to receive prophylactic antibiotic therapy with ciprofloxacin (500 mg) every 12 hours until the time of removal of the subcutaneous drains. In patients with allergy to ciprofloxacin, levofloxacin, moxifloxacin or amoxicillin clavulanate were used.
The reference cohort consisted of patients who received only standard antibiotic prophylaxis per guidelines. The standard antibiotic regimen was enforced on January 1, 2008, by the Mayo Clinic Surgical Quality and Safety Committee in accordance with the Surgical Care Improvement Project. This initiative required strict adherence to the current antibiotic prophylaxis guidelines for patients who underwent abdominal hysterectomy. All patients who were included in the study before January, 2008, received the prolonged antibiotic regimen before the implementation of the quality initiative, and all patients received the standard antibiotic regimen after implementation of the quality initiative.
Baseline and procedural data (independent variables)
Baseline data that were obtained for each patient included age, height and weight, BMI, indication for hysterectomy (benign vs malignant), medical comorbidities (eg, diabetes mellitus, corticosteroid use), and the American Society of Anesthesia classification. Procedure-related data included each patient’s minimum intraoperative body temperature, documented amount of blood loss, need for intraoperative blood transfusion, operative time, lymphadenectomy, and the need for a vertical extension of the incision, which is a vertical extension in both the skin and the fascia. Pathologic reports were also reviewed for the weight of panniculectomy and diagnosis. Admission to intensive care, duration of the hospital stay, and readmission were recorded.
Follow-up evaluation and measurement of outcomes (dependent variables)
SSI was the primary outcome for the evaluation of infectious morbidity after combined hysterectomy and panniculectomy in obese women. SSI was defined as infections that were related to the operative procedure that occurred at or near the surgical incision (incisional or organ/space) within 30 days of an operative procedure or within 1 year if an implant was left in place. The clinical criteria that were used to define an SSI included any of the following data: a purulent exudate draining from a surgical site, a positive fluid culture that had been obtained from a surgical site that was closed primarily, a surgeon’s diagnosis of infection, or any surgical site that required reopening. Other secondary outcomes included readmission for infection, incision, and drainage and a need for a secondary closure. To minimize measurement bias of the primary outcome, we used 2 independent identification methods. The first method included identification of cases through cross matching the study cohorts with the institutional data base for SSI. The second method included chart review of all included patients by 1 of the investigators (S.A.E.). Final verification of identified cases was performed by a second reviewer who was masked to case cohort allocation (R.L.T.). Disagreements were resolved by consensus.
Statistical analysis
The mean and SD were used to show continuous variables with normal distribution; median and range were used in skewed data. Categoric variables were presented as numbers and percentages. Baseline characteristics were compared with the use of t tests or Wilcoxon sum rank (Mann Whitney) test for continuous variables according to the distribution; χ 2 test or Fisher’s exact test was used for comparison of categoric variables, accordingly. Regression models were used to adjust for potential bias that was caused by known confounders. Logistic regression modeling was used to explore the association between infection and various confounders that included age; BMI; diabetes mellitus; steroid use; cancer; lymphadenectomy; combined procedures that included appendectomy, diverticulectomy, cholecystectomy, and Marshall-Marchetti-Krantz procedure; vertical midline incision extension; pannus weight; intraoperative temperature; operative time; blood loss; intraoperative transfusion, and the American society of anesthesia classification.
Adjusted odds ratios were used to evaluate the likelihood for SSI and possible predictors. All statistical analyses were 2-sided with probability values of < .05. All statistical analysis was done with JMP software (version 6.0; SAS Institute Inc, Cary, NC).
Results
From January 1, 2005, through December 31, 2008, 172 women underwent panniculectomy to facilitate gynecologic surgeries. Of those, 5 patients (3%) did not provide consent for research and were excluded. In addition, 22 patients were excluded; 10 patients had a BMI of <30 kg/m 2 , and 12 patients had pelvic operations that did not include hysterectomy. After those exclusions, 145 women were included in the analysis. Forty-three women (29.5%) were included in the extended antibiotic prophylaxis cohort, and 102 women (70.5%) received standard antibiotic prophylaxis ( Figure 1 ). Of 102 women in the extended antibiotic prophylaxis group, 95 women (93%) received ciprofloxacin; 5 women (5%) received levofloxacin; 1 woman (1%) received moxifloxacin, and 1 woman (1%) received amoxicillin clavulanate. The median time was 14 days (range, 7–26 days; 25th percentile of 10 days and 75th percentile of 14 days). Eight women received an additional intraoperative dose of cefazolin: 6 for prolonged surgery, and 2 for blood loss >1500 mL. Of 43 women in the standard antibiotic cohort, 4 women (9%) received at least 1 additional dose of intraoperative doses of prophylactic antibiotics, compared with 4 women (4%) in the extended prophylactic antibiotics group ( P = .24).
The mean age of participants was 56.0 ± 12.1 years (range, 31–85 years). The mean BMI was 42.6 ± 8.4 kg/m 2 (range, 30–86.4 kg/m 2 ), with 80 patients (55.2%) with a BMI of >40 kg/m 2 . The study population included 43 women (29%) with diabetes mellitus; 72 women (49.7%) with cancer; 84 women (57.9%) with American Society of Anesthesia stages III and IV. Along with hysterectomy, lymphadenectomy were performed in 53 women (36.6%); appendectomy or Meckel’s diverticulectomy was performed in 24 women (16.6%); cholecystectomy was performed in 4 women (2.8%), and Marshall-Marchetti-Krantz procedure was performed in 12 women (8.3%). Vertical midline incision was necessary in 35 women (24.1%). There was no significant difference in clinical characteristics in women who received extended prophylactic antibiotic compared with women who received the standard regimen ( Table 1 ).
| Characteristic | Antibiotic prophylaxis | P values | |
|---|---|---|---|
| Extended (n = 102) | Standard (n = 43) | ||
| Age, y a | 55.9 ± 12.2 | 56.3 ± 11.9 | .85 |
| Age >70 y, n (%) | 16 (15.7) | 7 (16.3) | .93 |
| Body mass index (kg/m 2 ) a | 42.9 ± 9.1 | 41.9 ± 6.3 | .45 |
| Body mass index >40 kg/m 2 , n (%) | 55 (53.9) | 25 (58.1) | .64 |
| Diabetes mellitus, n (%) | 27 (26.4) | 16 (37.2) | .20 |
| Steroid use, n (%) | 4 (3.9) | 1 (2.3) | 1.00 b |
| Cancer, n (%) | 47 (46.1) | 25 (58.1) | .18 |
| Endometrial c | 35/47 | 17/25 | |
| Ovarian c | 5/47 | 6/25 | |
| Cervical c | 2/47 | 2/25 | |
| Other cancers c | 5/47 | 0/25 | |
| Lymphadenectomy, n (%) | 33 (32.5) | 20 (46.5) | .11 |
| Appendectomy/diverticulectomy, n (%) | 22 (21.6) | 4 (9.3) | .079 |
| Cholecystectomy, n (%) | 4 (3.9) | 0 | .32 b |
| Marshall-Marchetti-Krantz, n (%) | 10 (9.8) | 2 (4.9) | .51 b |
| Vertical midline incision extension, n (%) | 23 (22.6) | 12 (27.9) | .49 |
| Weight of pannus, g a | 3379.5 ± 2681.4 | 3160 ± 1846.9 | .57 |
| Pannus weight >2500 g, n (%) | 53 (51.9) | 25 (58.1) | .50 |
| Lowest intraoperative temperature (°C) a | 35.9 ± 0.75 | 35.8 ± 0.59 | .25 |
| Intraoperative temperature < 35°C, n (%) | 15 (14.7) | 3 (7.0) | .19 |
| Operative time, min a | 134.5 ± 46.1 | 143.6 ± 56.6 | .36 |
| Operative time >180 min, n (%) | 19 (18.6) | 10 (23.2) | .52 |
| Blood loss, mL a | 365.7 ± 207.9 | 457.1 ± 374.4 | .14 d |
| Blood loss >1000 mL, n (%) | 3 (2.9) | 3 (7.1) | .36 b |
| Intraoperative transfusion, n (%) | 11 (10.8) | 8 (18.6) | .20 |
| American Society of Anesthesiologist score 3 & 4, n (%) | 56 (54.9) | 28 (65.1) | .26 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
