Surgical site infections are a major cause of morbidity in patients undergoing operative procedures.1–4 Surgical site infections not only result in substantial pain and suffering but also are costly to treat. One study estimated that the development of a surgical site infection resulted, on average, in more than $10,000 of excess hospital costs and prolonged the length of stay by more than 4 days.1 For women undergoing hysterectomy, wound complications have been reported to occur in more than 20% of patients in some reports.5
Over the course of the past four decades, numerous studies have suggested that the administration of perioperative antibiotics reduces infectious morbidity for high-risk surgical procedures.6–10 For hysterectomy, a clean contaminated procedure in which the vagina is entered, one meta-analysis of 17 studies that included 2,752 patients noted that antibiotic prophylaxis reduced the infection rate by 65%.8 Based on these data, the American College of Obstetricians and Gynecologists (the College) recommends antibiotic prophylaxis for hysterectomy, urogynecologic procedures, hysterosalpingogram, and induced abortion.9 For procedures with a low risk of infection, these guidelines do not recommend antibiotic prophylaxis for lower-risk clean procedures, including operative and diagnostic laparoscopy, tubal sterilization, hysteroscopy, and laparotomy.9 Numerous other professional societies have developed similar guidelines for prophylaxis for high-risk surgeries.10–13
In addition to published guidelines, national efforts have been developed to promote proper antibiotic use.4 Despite these efforts, little is known about the actual adherence to recommendations for the allocation of perioperative antibiotics in women undergoing gynecologic surgery. We performed a population-based analysis to determine the patterns and predictors of guideline-based use of antibiotics in women who underwent gynecologic surgery.
MATERIALS AND METHODS
The Perspective database, a voluntary database that captures data from more than 500 acute-care hospitals from throughout the United States, was used.14 Participating hospitals submit electronic updates on a quarterly basis. The data are audited regularly to ensure quality and integrity. The database captures clinical and demographic data, diagnoses, procedures, and all billed services rendered during a hospital stay,15 and therefore contains information regarding all drugs, devices, radiologic tests, laboratory tests, and therapeutic services rendered during a patient's hospitalization. In 2006, nearly 5.5 million hospital discharges representing approximately 15% of all hospitalizations were captured in the database.16 The database has been validated and used in a large number of outcomes studies. The Columbia University Institutional Review Board deemed this study exempt.
We analyzed women aged 18 years or older who underwent inpatient or outpatient gynecologic surgery between 2003 and the first quarter of 2010. Surgical procedures were identified through International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) coding. Based on published practice guidelines by the College, procedures were classified as either antibiotic-appropriate procedures, those operations for which antibiotics were recommended, or antibiotic-inappropriate procedures, surgeries for which antibiotics were not routinely recommended.9 Antibiotic-appropriate procedures included abdominal hysterectomy (ICD-9-CM codes 68.3, 68.39, 68.4, 68.49, 68.90), vaginal hysterectomy (ICD-9-CM codes 68.5, 68.59), and laparoscopically assisted vaginal hysterectomy (ICD-9-CM codes 68.31, 68.41, 68.51). Antibiotic-inappropriate procedures included myomectomy (ICD-9-CM code 68.29), open and laparoscopic oophorectomy with or without salpingectomy (ICD-9-CM codes 65.3x, 65.4x, 65.6x, 65.6x), open and laparoscopic ovarian cystectomy (ICD-9-CM code 65.2x), dilation and curettage with or without hysteroscopy (ICD-9-CM codes 69.0, 69.09, 68.12), and laparoscopic tubal ligation (ICD-9-CM codes 66.2, 66.21, 66.22, 66.29).9
Patients were classified through a hierarchy, based on the order of procedures listed, so that they were only included once even if they underwent multiple procedures. For example, if a patient underwent hysteroscopy along with a tubal ligation, she was classified in the hysteroscopy group. Pregnant patients, including those with ectopic gestations and patients who underwent abortion, were excluded from the analysis (ICD-9-CM codes 630.x–679.x). To capture elective or planned surgery, we only included women who underwent surgery on the first day of hospitalization.17
Demographic data analyzed included age (younger than 40, 40–49, 50–59, 60–69, and 70 years or older), year of surgery, race (white, black, Hispanic, and other), marital status, and insurance status (Medicare, Medicaid, commercial, self-pay, and unknown). Hospitals in which patients were treated were characterized based on location (metropolitan, nonmetropolitan), region of the country (Northeast, Midwest, West, South), size (fewer than 400 beds, 400–600 beds, and more than 600 beds), and teaching status (teaching, nonteaching). Risk adjustment for medical comorbidities was performed using the Elixhauser comorbidity index. Patients were categorized based on the number of medical comorbidities as 0, 1, 2, and 3 or more, as previously reported.18
A composite surgical volume for all procedures was determined for hospitals and surgeons. Annualized procedural volume was estimated for each surgeon or hospital by dividing the total number of procedures performed by the number of years during which the individual hospital or surgeon contributed at least one operation.19,20 The volume distributions were examined visually and volume was classified into approximately equal patient-based quartiles for surgeons (lowest, 18.5 cases per year or less; second, 18.51–33.99 cases per year; third, 34.00–56.60 cases per year; and highest, more than 56.6 cases per year) and hospitals (lowest, fewer than 320.15 cases per year; second, 320.15–524.15 cases per year; third, 524.16–888.25 cases per year; and highest, more than 888.25 cases per year).
The database captures all drugs administered to patients and records the hospital day on which the drug is administered. To capture perioperative antibiotic use, we classified antibiotics administered on hospital day 1 as perioperative antibiotics. Antibiotics were categorized based on whether the antibiotic has been recommended in published guidelines for gynecologic surgery.9 For this classification schema, we used a permissive definition of guideline-adherent antibiotics and included any drug or class of drugs previously described in the College guidelines, including cefazolin, cefotetan, cefoxitin, cefuroxime, ampicillin-sulbactam, clindamycin, gentamicin, aztreonam, metronidazole, ciprofloxacin, enoxacin, gatifloxacin, levofloxacin, lomefloxacin, moxifloxicin, nalidixic acid, norfloxacin, ofloxacin, trovafloxacin, grepafloxacin, sparfloxacin, and temafloxacin. Other antibiotics were categorized as nonguideline-based antibiotics.9
Frequency distributions between categorical variables were compared using χ2 tests. Separate analyses were performed for antibiotic-appropriate and antibiotic-inappropriate procedures. Hierarchical logistic regression analysis was used to determine factors associated with use of perioperative antibiotics. These models included all patient, physician, and hospital characteristics, as well as physician-specific and hospital-specific random effects to account for clustering. For antibiotic-appropriate procedures, we modeled failure to use antibiotics, whereas for antibiotic-inappropriate procedures we modeled use of perioperative antibiotics. All outcomes are reported as risk ratios (RRs) with 95% confidence intervals (CIs). All analyses were performed with SAS 9.3.
A total of 1,036,403 women, including 545,332 (52.6%) who underwent procedures for which antibiotics were recommended and 491,071 (47.4%) who underwent operations for which antibiotics were not recommended, were identified. Within the cohort of women who had procedures for which antibiotics were recommended, 475,180 (87.1%) received appropriate antibiotic prophylaxis, whereas 12,451 (2.3%) received nonguideline-recommended antibiotics and 57,901 (10.6%) received no prophylaxis. Within this cohort, use of antibiotics increased from 88.0% in 2003 to 90.7% in 2010 (P<.001). Likewise, for each procedure, antibiotic use increased over time. For abdominal hysterectomy, the rate of antibiotic use was 87.4% in 2003 and increased to 89.0% in 2010; for laparoscopically assisted vaginal hysterectomy, the rate of 89.2% increased to 91.5%; and for vaginal hysterectomy, the rate of 88.9% increased to 92.4% (P<.001 for all) (Fig. 1A).
Table 1 displays the characteristics of the patients for whom antibiotics were recommended. In an adjusted multivariable model, patients who underwent surgery more recently, those residing in nonmetropolitan areas, those with three or more medical comorbidities, women living in areas other than the eastern United States, and patients treated by higher-volume surgeons were more likely to receive antibiotics. High-volume surgeons were 41% less likely to omit antibiotics (RR 0.59; 95% CI 0.52–0.66) than low-volume surgeons, whereas nonteaching hospitals were 25% less likely than teaching hospitals to omit antibiotics (RR 0.75; 95% CI 0.70–0.81).
Among women who underwent a procedure for which antibiotics were not recommended, antibiotics were administered to 197,226 (40.2%) women. Use of antibiotics increased over time from 33.4% in 2003 to 43.7% in 2010 (P<.001). Increased antibiotic use over time from 2003 to 2010 was noted for all of the following individual procedures: myomectomy (63.1–66.5%); oophorectomy (79.0–82.6%); laparoscopic oophorectomy (47.2–63.3%); cystectomy (69.1–77.1%); laparoscopic cystectomy (37.2–50.9%); dilation and curettage with or without hysteroscopy (23.1–33.2%); and laparoscopic tubal ligation (14.9–29.6%) (P<.001 for all) (Fig. 1B).
Use of antibiotics in patients who underwent procedures for which antibiotics were not required is shown in Table 2. In a multivariable model, patients treated more recently were more likely to receive antibiotics (RR 1.32; 95% CI 1.29–1.36 for 2010 compared with 2003). Women with Medicaid (RR 0.93; 95% CI 0.92–0.95) and uninsured (RR 0.97; 95% CI 0.94–0.99) women, as well as those treated at nonmetropolitan centers (RR 0.94; 95% CI 0.90–0.98), were less likely to receive antibiotics, whereas patients in areas other than the eastern United States were more likely to receive antibiotics. High-volume hospitals and high-volume surgeons were less likely to administer perioperative antibiotics to these women; compared with low-volume surgeons, high-volume surgeons were 13% less likely to prescribe antibiotics (RR 0.87; 95% CI 0.83–0.92). Women who underwent laparotomy were more likely to receive antibiotics than those who underwent laparoscopic surgery.
Our findings suggest that the use of perioperative antibiotics in women undergoing gynecologic surgery is poorly aligned with published guidelines. Although use of antibiotics is high (87%) for women who should receive antibiotics, antibiotics are being increasingly administered to women who are likely to receive little benefit from the drugs.
To promote the use of appropriate antibiotic prophylaxis, a national collaborative developed the Surgical Care Improvement Project in 2003.4,21–26 The project’s initiatives include perioperative quality measures, including several focused on the proper allocation of antibiotics. Importantly, hysterectomy is one of the procedures targeted in the Surgical Care Improvement Project collaborative. Although hospital participation in the Surgical Care Improvement Project is voluntary, reimbursement by the Centers for Medicare and Medicaid Services is reduced by 2% if hospitals do not report their outcomes.26 Our data suggest that use of antibiotic prophylaxis for hysterectomy is high.
Although use of antibiotics for appropriate procedures was relatively high, we also noted that antibiotics were frequently (40%) administered for procedures for which the drugs are of little benefit. Other studies also have noted that the introduction of quality measures often leads to unintended overuse of medications.27–29 After the Centers for Medicare and Medicaid Services adopted a quality measure that required antibiotic administration within 4 hours of hospital arrival for patients with community-acquired pneumonia, there was a dramatic increase in the use of antibiotics for patients who did not have pneumonia.28,29 Similarly, guidelines for venous thromboembolism prophylaxis for nonsurgical patients may promote pharmacologic prophylaxis in patients at low risk and in those with contraindications to anticoagulation.27 Inappropriate use of antibiotics in patients at low risk not only exposes patients to toxicity with little potential benefit but also leads to substantial resource use and may alter patterns of antimicrobial resistance. Concern about noncompliance with a quality metric may, in part, explain the overuse of perioperative antibiotics for gynecologic procedures that we noted.
We found that surgical volume was an important predictor of guideline adherence. Compared with low-volume surgeons, high-volume physicians were 41% less likely to omit antibiotics in patients for whom the drugs were indicated, and were 13% less likely to prescribe antibiotics when they were not indicated. Although the relationship between surgical volume and morbidity and mortality for gynecologic procedures is modest, there is a more robust association between higher surgical volume and decreased resource use for pelvic surgery.30–33 These data suggest that adherence to best practice perioperative guidelines may, at least in part, explain the lower costs incurred by higher-volume gynecologic surgeons. Our group has previously demonstrated that higher-volume surgeons are also more likely to use appropriate perioperative prophylaxis against venous thromboembolism.34,35
Nonclinical factors appeared to play an important role in the allocation of perioperative antibiotics. Patients treated in the eastern United States were more likely to receive antibiotics regardless of procedure type. Previous studies have shown substantial regional variation in surgical practice patterns.17,31 Patients with a greater number of comorbid medical conditions more frequently received unindicated and prolonged antibiotics; whereas older patients were less likely to receive antibiotics for antibiotic-inappropriate procedures, older women more frequently received prolonged antibiotics. Finally, hospital characteristics were associated with antibiotic use; patients treated at nonteaching hospitals were 25% less likely to receive antibiotics when indicated.
Although our analysis benefits from the inclusion of a large number of women, we acknowledge a number of important limitations. We cannot exclude the possibility that prophylaxis was misclassified in a small number of women. However, the database has been validated and used in a number of studies including analyses examining antibiotic use.26,36 A priori we chose an inclusive definition of guideline-based antibiotics and included any drugs or class of drugs that have been described in published guidelines. Finally, we recognize the fact that just because guidelines do not recommend antibiotics, it does not necessarily mean that antibiotics are of no benefit in some scenarios. Administrative data are unlikely to capture some clinical scenarios in which antibiotics may be warranted. Clearly, antibiotic prophylaxis has been poorly studied for many gynecologic procedures, including after complications, and would benefit from more rigorous prospective trials.
The widespread misuse of perioperative antibiotics for gynecologic surgery suggests that strategies to better-align practice patterns with evidence-based recommendations are urgently needed. Although most quality initiatives have focused on promoting an intervention, reducing unnecessary treatments could substantially decrease medical waste and lower costs.37 Previous studies have demonstrated moderate success for pay-for-performance initiatives and public reporting of data.38 Similarly, educational interventions, formalized protocol implementation, and clinical decision support tools linked to electronic order entry have all been proposed to reduce unnecessary testing and treatments.39–41
1. Boltz MM, Hollenbeak CS, Julian KG, Ortenzi G, Dillon PW. Hospital costs associated with surgical site infections in general and vascular surgery patients. Surgery 2011;150:934–42.
2. Ortega G, Rhee DS, Papandria DJ, Yang J, Ibrahim AM, Shore AD, et al.. An evaluation of surgical site infections by wound classification system using the ACS-NSQIP. J Surg Res 2012;174:33–8.
3. Campbell DA Jr, Henderson WG, Englesbe MJ, Hall BL, O'Reilly M, Bratzler D, et al.. Surgical site infection prevention: the importance of operative duration and blood transfusion–results of the first American College of Surgeons-National Surgical Quality Improvement Program Best Practices Initiative. J Am Coll Surg 2008;207:810–20.
4. Bratzler DW, Hunt DR. The surgical infection prevention and surgical care improvement projects: national initiatives to improve outcomes for patients having surgery. Clin Infect Dis 2006;43:322–30.
5. Clarke-Pearson DL, Geller EJ. Complications of hysterectomy. Obstet Gynecol 2013;121:654–73.
6. Duff P, Park RC. Antibiotic prophylaxis in vaginal hysterectomy: a review. Obstet Gynecol 1980;55:193S–202S.
7. Mittendorf R, Aronson MP, Berry RE, Williams MA, Kupelnick B, Klickstein A, et al.. Avoiding serious infections associated with abdominal hysterectomy: a meta-analysis of antibiotic prophylaxis. Am J Obstet Gynecol 1993;169:1119–24.
8. Tanos V, Rojansky N. Prophylactic antibiotics in abdominal hysterectomy. J Am Coll Surg 1994;179:593–600.
9. Antibiotic prophylaxis for gynecologic procedures. ACOG Practice Bulletin No. 104. American College of Obstetricians and Gynecologists. Obstet Gynecol 2009;113:1180–9.
10. Leaper D, Burman-Roy S, Palanca A, Cullen K, Worster D, Gautam-Aitken E, et al.. Prevention and treatment of surgical site infection: summary of NICE guidance. BMJ 2008;337:a1924.
11. Morrill MY, Schimpf MO, Abed H, Carberry C, Margulies RU, White AB, et al.. Antibiotic prophylaxis for selected gynecologic surgeries. Int J Gynaecol Obstet 2013;120:10–5.
12. Van Eyk N, van Schalkwyk J. Antibiotic prophylaxis in gynaecologic procedures. J Obstet Gynaecol Can 2012;34:382–91.
13. Bratzler DW, Dellinger EP, Olsen KM, Perl TM, Auwaerter PG, Bolon MK, et al.. Clinical practice guidelines for antimicrobial prophylaxis in surgery. Surg Infections 2013;14:73–156.
14. Wright JD, Hershman DL, Shah M, Burke WM, Sun X, Neugut AI, et al.. Quality of perioperative venous thromboembolism prophylaxis in gynecologic surgery. Obstet Gynecol 2011;118:978–86.
15. Rogo-Gupta L, Rodriguez LV, Litwin MS, Herzog TJ, Neugut AI, Lu YS, et al.. Trends in surgical mesh use for pelvic organ prolapse from 2000 to 2010. Obstet Gynecol 2012;120:1105–15.
16. Lindenauer PK, Pekow PS, Lahti MC, Lee Y, Benjamin EM, Rothberg MB. Association of corticosteroid dose and route of administration with risk of treatment failure in acute exacerbation of chronic obstructive pulmonary disease. JAMA 2010;303:2359–67.
17. Schiavone MB, Herzog TJ, Ananth CV, Wilde ET, Lewin SN, Burke WM, et al.. Feasibility and economic impact of same-day discharge for women who undergo laparoscopic hysterectomy. Am J Obstet Gynecol 2012;207:382.e1–9.
18. van Walraven C, Austin PC, Jennings A, Quan H, Forster AJ. A modification of the Elixhauser comorbidity measures into a point system for hospital death using administrative data. Med Care 2009;47:626–33.
19. Wright JD, Lewin SN, Deutsch I, Burke WM, Sun X, Herzog TJ. Effect of surgical volume on morbidity and mortality of abdominal hysterectomy for endometrial cancer. Obstet Gynecol 2011;117:1051–9.
20. Wright JD, Lewin SN, Deutsch I, Burke WM, Sun X, Herzog TJ. The influence of surgical volume on morbidity and mortality of radical hysterectomy for cervical cancer. Am J Obstet Gynecol 2011;205:225.e1–7.
21. Bratzler DW. The surgical infection prevention and surgical care improvement projects: promises and pitfalls. Am Surg 2006;72:1010–6.
23. Awad SS. Adherence to surgical care improvement project measures and post-operative surgical site infections. Surg Infect (Larchmt) 2012;13:234–7.
24. Berenguer CM, Ochsner MG Jr, Lord SA, Senkowski CK. Improving surgical site infections: using National Surgical Quality Improvement Program data to institute Surgical Care Improvement Project protocols in improving surgical outcomes. J Am Coll Surg 2010;210:737–43.
25. Potenza B, Deligencia M, Estigoy B, Faraday E, Snyder A, Angle N, et al.. Lessons learned from the institution of the Surgical Care Improvement Project at a teaching medical center. Am J Surg 2009;198:881–8.
26. Stulberg JJ, Delaney CP, Neuhauser DV, Aron DC, Fu P, Koroukian SM. Adherence to surgical care improvement project measures and the association with postoperative infections. JAMA 2010;303:2479–85.
27. Baker DW, Qaseem A. Evidence-based performance measures: preventing unintended consequences of quality measurement. Ann Intern Med 2011;155:638–40.
28. Kanwar M, Brar N, Khatib R, Fakih MG. Misdiagnosis of community-acquired pneumonia and inappropriate utilization of antibiotics: side effects of the 4-h antibiotic administration rule. Chest 2007;131:1865–9.
29. Welker JA, Huston M, McCue JD. Antibiotic timing and errors in diagnosing pneumonia. Arch Intern Med 2008;168:351–6.
30. Wright JD, Ananth CV, Lewin SN, Burke WM, Lu YS, Neugut AI, et al.. Robotically assisted vs laparoscopic hysterectomy among women with benign gynecologic disease. JAMA 2013;309:689–98.
31. Wright JD, Burke WM, Wilde ET, Lewin SN, Charles AS, Kim JH, et al.. Comparative effectiveness of robotic versus laparoscopic hysterectomy for endometrial cancer. J Clin Oncol 2012;30:783–91.
32. Rogo-Gupta LJ, Lewin SN, Kim JH, Burke WM, Sun X, Herzog TJ, et al.. The effect of surgeon volume on outcomes and resource use for vaginal hysterectomy. Obstet Gynecol 2010;116:1341–7.
33. Wallenstein MR, Ananth CV, Kim JH, Burke WM, Hershman DL, Lewin SN, et al.. Effect of surgical volume on outcomes for laparoscopic hysterectomy for benign indications. Obstet Gynecol 2012;119:709–16.
34. Ritch JM, Kim JH, Lewin SN, Burke WM, Sun X, Herzog TJ, et al.. Venous thromboembolism and use of prophylaxis among women undergoing laparoscopic hysterectomy. Obstet Gynecol 2011;117:1367–74.
35. Wright JD, Lewin SN, Shah M, Burke WM, Sun X, Neugut AI, et al.. Quality of venous thromboembolism prophylaxis in patients undergoing oncologic surgery. Ann Surg 2011;253:1140–6.
36. Rothberg MB, Pekow PS, Lahti M, Brody O, Skiest DJ, Lindenauer PK. Antibiotic therapy and treatment failure in patients hospitalized for acute exacerbations of chronic obstructive pulmonary disease. JAMA 2010;303:2035–42.
37. Brook RH. The role of physicians in controlling medical care costs and reducing waste. JAMA 2011;306:650–1.
38. Eijkenaar F, Emmert M, Scheppach M, Schoffski O. Effects of pay for performance in health care: a systematic review of systematic reviews. Health Policy 2013;110:115–30.
39. Levick DL, Stern G, Meyerhoefer CD, Levick A, Pucklavage D. Reducing unnecessary testing in a CPOE system through implementation of a targeted CDS intervention. BMC Med Inform Decis Mak 2013;13:43.
40. Vegting IL, van Beneden M, Kramer MH, Thijs A, Kostense PJ, Nanayakkara PW. How to save costs by reducing unnecessary testing: lean thinking in clinical practice. Eur J Intern Med 2012;23:70–5.
41. May TA, Clancy M, Critchfield J, Eberling F, Enriquez A, Gallagher C, et al.. Reducing unnecessary inpatient laboratory testing in a teaching hospital. Am J Clin Pathol 2006;126:200–6.