Surgical site infections are the second most common reason for unplanned hospital readmissions after hysterectomy1,2 and result in increased morbidity and health care costs. The estimated rate of surgical site infection after hysterectomy varies between 1% and 4%.3–8 Surgical site infection rates in hysterectomy have been publicly reported since 2013 and, since January 2014, reimbursements may be withheld by the Centers for Medicare & Medicaid Services if the rates exceed expected values.9 In 2012, the National Healthcare Safety Network identified an elevated posthysterectomy surgical site infection rate at our institution. The standardized infection ratio, which is risk-adjusted for facility and patient-level factors, was 2.19, a level “above expected” because a ratio greater than 1.0 suggests a higher than expected level.10 The problem was addressed by hospital epidemiology and infection control with input from other departments, and a multidisciplinary surgical site infection prevention steering committee (henceforth, the committee) was formalized in 2014 and tasked with decreasing the standardized infection ratio to less than 1.0 by 2015. The surgical site infection rate at the start of data collection was 5.36%; the objective of this initiative was to drive the rate below 1.0%. A prevention bundle was created that included chlorhexidine gluconate-impregnated wipes, maintenance of normothermia, standardized preoperative skin preparation and antibiotic dosing, dressing maintenance, and direct feedback when the protocol was breached. Our analysis evaluated the effects of this bundle and whether it can be a sustainable method of reducing surgical site infection rates. We use the Standards for Quality Improvement Reporting Excellence 2.0 framework for publication of quality improvement studies.11
MATERIALS AND METHODS
Yale New Haven Hospital is the primary teaching hospital for Yale School of Medicine and medical students, residents, and fellows are involved in the majority of gynecologic surgeries. Hysterectomies are performed by a large group of academic and community-based gynecologists and by subspecialists. During the study period (2013–2015) the annual number of hysterectomies ranged from 706 to 832.
The bundle was designed and implemented by the committee led by representatives from gynecology, anesthesia, hospital epidemiology and infection control, and peri- and postoperative nursing leadership. The group developed a gynecology-specific bundle based on existing evidence and best practice guidelines,12,13 some of which were extrapolated from colorectal and other surgery literature7,14–17 with the understanding that the elements should be evidence- or consensus-based and not cause harm to the patient. The committee also sought to make the interventions sustainable by implementing plans for ongoing presence of the committee and continued oversite. Before formation of the committee, individual groups had piloted single interventions for surgical site infection prevention. Bandage removal after 24–48 hours, use of chlorhexidine skin wipes, preoperative warming, and an instructional video summarizing best practices for abdominal–perineal preparation had all been piloted. Nevertheless, these interventions had yielded disappointing results because of inconsistency of education, adherence, and monitoring processes. Although data suggest these measures correlate with reduced incidence of surgical site infection, there is evidence that single interventions are unlikely to reduce the overall incidence of surgical site infections.14,15,18 As each bundle component was formally implemented, it was introduced to the relevant health care providers with departmental grand rounds and didactic presentations at nursing staff meetings. As of 2016, there is an annual required online teaching module for all surgical team members. The first four components were instituted at the outset of data collection, and the final three elements were introduced in a stepwise fashion (Table 1).
The Surgical Site Infection Prevention Bundle elements are:
- Chlorhexidine gluconate (implemented at the start of the study period): Chlorhexidine gluconate-impregnated wipes were dispensed at the preoperative visit or in the preoperative unit before surgery and patients were counseled regarding their use.15,16,19
- Patient-controlled preoperative warming (implemented at the start of the study period): While awaiting surgery, patients were provided with forced-air warming devices to promote normothermia.12
- Aseptic gynecologic skin preparation (implemented at the start of the study period): A single abdominal preparation (2% chlorhexidine–70% isopropyl alcohol solution) and a single vaginal preparation (4% chlorhexidine–4% isopropyl alcohol solution) were made standard across all obstetric and gynecologic services.19–22 Chlorhexidine-based preparations were selected as a result of activity in the presence of blood (especially relevant to vaginal preparation).23 Mandatory education for perioperative staff, surgeons, and housestaff included viewing a gynecology department-produced video on optimal technique for aseptic cleansing of perineal and vaginal areas. This video addressed the absence of a standard vaginal preparation protocol at our hospital and remains in use today.
- Sterile dressing (implemented at the start of the study period): Postoperatively, sterile dressings were maintained for a minimum of 24 hours and were removed by 48 hours postoperatively. Use of topical skin adhesives was left up to the individual surgeon and was not considered part of the primary sterile dressing. Patients who were discharged before 24 hours postoperatively were instructed to remove their dressings at home, between 24 and 48 hours after surgery.12
- Maintenance of intraoperative warming (implemented March 1, 2014): Patients' intraoperative temperatures were monitored by anesthesia and forced-air warming devices were used to prevent core temperatures less than 36°C.24
- Antibiotic standardization and redosing (implemented October 1, 2014): The antibiotic protocol consisted of administration of 2 g cefazolin (3 g if weight greater than 120 kg) 1 hour or less before incision. Patients with cephalosporin or penicillin allergies were given 900 mg intravenous (IV) clindamycin and 2 mg/kg of ideal body weight IV gentamicin instead of cefazolin.13,25 The addition of 500 mg IV metronidazole preoperatively was standardized for cases in which bowel involvement was anticipated or for procedures anticipated to be longer or at higher risk for infection.17,25 This group included patients undergoing concurrent lymph node dissection or omentectomy, because these patients are at higher risk for anaerobic infection, according to internal epidemiologic data. The protocol specified redosing at 3 hours of operative time (to ensure complete dosing by 3.6 hours, double the half-life of cefazolin8,13) or for blood loss of 1.5 L or greater. This regimen differed from our prior protocol, which followed Surgical Care Improvement Project measures: 1 g IV cefazolin for patients less than 80 kg, 2 g IV cefazolin for patients 80 kg or greater, with no mention of metronidazole for gynecologic surgery.26
- Timely and constructive direct feedback (implemented December 1, 2014): A structured method of feedback was provided to individual physicians and frontline perioperative staff. The committee, which met monthly to monitor and review progress, kept track of compliance with appropriate intraoperative core temperature, adherence to antibiotic prophylaxis, redosing of antibiotics, and surgical preparation. When deviations from the protocol were observed, a committee member personally contacted the responsible member of the perioperative team. Feedback conversations were designed to be surgeon to surgeon, anesthesiologist to anesthesiologist, nurse to nurse to ascertain reasons for failure of the protocol and to reiterate the requirements of the intervention. These conversations stressed team support and education to promote awareness and understanding of the bundle.1,22,27,28
This quality improvement project was approved by the institutional review board of the Yale Human Research Protection Program. Data for hysterectomies performed from April 2013 to December 2015 at Yale New Haven Hospital were prospectively gathered by trained epidemiology and infection control staff. All abdominal hysterectomies, traditional laparoscopic and robotic-assisted laparoscopic hysterectomies, and laparoscopic-assisted vaginal hysterectomies, including those with concurrent bowel or bladder procedures, were included. Hysterectomies performed for obstetric patients and those performed by a strictly vaginal route were excluded, because surgical site infections related to these procedures are not Centers for Medicare & Medicaid Services-reportable events and were not prospectively monitored using the same National Healthcare Safety Network guidelines.
Data were collected using the Centers for Disease Control and Prevention's National Healthcare Safety Network guidelines.29 The cases included in this study were identified by hospital discharge International Classification of Disease, 9th Revision procedure codes: 68.31, 68.39, 68.41, 68.49, 68.61, and 68.69 as well as by staff examining the hospital surgical schedule weekly to confirm that all relevant inpatient and outpatient hysterectomies were identified. Standard Centers for Disease Control and Prevention surgical site infection definitions (superficial, deep, and organ space)12 were used to identify infections occurring within 30 days of the procedure. A small subset of low-volume surgeons whose offices were not on the hospital-wide electronic medical record was also contacted by letter to ensure collection of all data.
After initial data collection for institutional and National Healthcare Safety Network reporting purposes, the infection prevention team data were released to the study authors. A physician (S.E.A.) and a medical student (E.M.L.) verified all identified cases by entering each patient's medical record number into the electronic medical record and reviewing the appropriate operative documentation to ensure that included participants met eligibility criteria and to collect information on patient demographics, comorbidities, intraoperative characteristics, and length of hospital stay. Specifically, we collected measures of patient age, body mass index (BMI, calculated as weight (kg)/[height (m)]2), presence or absence of diabetes mellitus and cancer diagnoses, involvement of bowel procedure, surgical route, operative time, and estimated blood loss.
To assess readmissions occurring within 30 days after surgery, we queried our electronic medical record to capture all readmissions for our cohort, regardless of reason, across our health care system during the study period. A single physician (S.E.A.) reviewed information in the medical records of all readmissions and the corresponding primary International Classification of Disease, 9th Revision and 10th Revision diagnosis code. In our sample, patients were readmitted for complications of diabetes mellitus, pneumonia, hyponatremia, postoperative pain, pulmonary embolism, cardiac events, bowel obstruction, constipation, and urinary tract infections as well as surgical site infection diagnoses. The surgical site infection–related diagnosis codes observed in our sample were: 780.2 (postprocedural fever), 998.12 (hematoma complicating procedure), 998.59 (other postoperative infection, which included diagnoses of pelvic abscess, vaginal cuff abscess, infected seroma, and infected pelvic hematoma), and N76.2 (acute vaginitis). Full chart review was performed to ensure that patients met criteria for a surgical site infection.
Our primary outcome measure for analysis was surgical site infection meeting Centers for Disease Control and Prevention-defined criteria.12 Our secondary outcomes of interest were 1) postoperative length of hospital stay and 2) surgical site infection–related postoperative readmissions occurring within the 30-day period after the hysterectomy.
Descriptive statistics were calculated to summarize patient demographics and clinical characteristics. Patient characteristics were compared using χ2 test or Fisher exact test for categorical variables and Student t test and Wilcoxon rank-sum test for continuous variables. We conducted multivariable regression analysis to examine the associations between full implementation of the prevention bundle and patient outcomes while adjusting for patient demographic and clinical characteristics. Logistic regression models were used for determining associations of the bundle with surgical site infection and 30-day readmissions related to surgical site infection, and a Poisson model was used for length of hospital stay. The indicator for post–full bundle implementation was forced into the models; other variables measuring patient characteristics were selected for inclusion in the final model based on a backward stepwise selection process (cutoff P value=.05).
Because our bundle consisted of multiple components (Table 1) that were implemented sequentially, we developed an additional logistic regression model to assess the incremental effect of individual bundle components. Specifically, rather than dichotomizing the study period into pre– compared with post–full bundle implementation, we further categorized the study period into four subperiods reflecting when the first four components of the bundle (chlorhexidine gluconate-impregnated wipes, preoperative warming devices, standardization of aseptic surgical preparation, and surgical dressing maintenance) were implemented, when the fifth component (maintenance of intraoperative normothermia) was added, when the sixth component (antibiotic standardization) was added, and when the seventh component (direct feedback) was added. By alternating each of the first three subperiods as the reference group in analysis and comparing it with the next adjacent subperiod, we assessed the potential incremental effect of the additional bundle component on surgical site infection.
All statistical analysis was conducted using GraphPad Prism 6.0 and IBM SPSS Statistics 24. Crude and adjusted odds ratios (ORs) and 95% CIs were reported for results from regression models.
During the study period, a total of 2,099 hysterectomies meeting study criteria were performed. There were 1,352 performed during bundle implementation (April 1, 2013–November 30, 2014) and 747 performed once the bundle implementation was complete (December 1, 2014–December 31, 2015).
The mean age and BMI of patients in the pre– and post–full bundle implementation periods were comparable (P=.39 and .37) (Table 2). There was no statistically significant difference in the proportion of patients who had a diagnosis of diabetes mellitus, a cancer diagnosis associated with the surgery, or bowel involvement during the surgery between the pre– and post–full bundle implementation periods (P=.08, .73, .16). From the pre– to post–full bundle implementation periods, surgical route shifted to fewer open cases (49.9% vs 39.1% open, 48.1% vs 58.8% laparoscopic, 1.99% vs 2.14% laparoscopic-assisted vaginal, P<.001). Operating time was slightly longer among patients in the post–full bundle implementation period (median 138.0 minutes vs 157.0 minutes, P<.001). There was no significant difference in estimated blood loss between patients in the two periods (data not shown).
Patients who underwent surgery after the bundle was fully implemented had a reduced risk for overall surgical site infection (4.5% vs 1.9%, P=.001; Table 2). Superficial surgical site infection was the most common type and its rate decreased from 2.1% before full bundle implementation to 0.8% after full bundle implementation (P=.02; Table 2). After adjusting for patient characteristics, post–full bundle implementation was associated with an adjusted OR of 0.46 for surgical site infection (95% CI 0.25–0.82, P=.01) (Table 3). Additional multivariable regression models to assess individual bundle components showed no statistically significant difference in risk for surgical site infection associated with maintenance of intraoperative normothermia, antibiotic standardization, or direct feedback (Table 4).
When reviewing the data longitudinally by month, the mean overall surgical site infection rate fell from 5.4% at the beginning of data collection to 2.0% or below in the last 8 months of data collection. The rate of deep and organ space infections was initially 3.0% but fell to a mean of 1.2% (and was zero in some of these months) during the last 8 months of data collection (Fig. 1).
Postoperative days of hospital stay decreased slightly between pre– and post–full bundle implementation (2.68 vs 2.34 days, P=.01) (Table 2). As expected, increasing age, BMI higher than 40, bowel involvement, cancer diagnosis, and diabetes mellitus diagnosis were all associated with longer hospital stays. After adjusting for these patient characteristics in a Poisson regression model, there was no longer a statistically significant difference in postoperative length of stay between the two time periods (adjusted mean ratio 0.95 95% CI 0.90–1.01, P=.09) (data not shown).
During the pre–full bundle implementation period, 1.3% of the patients had a surgical site infection–related readmission within 30 days of the hysterectomy. This decreased to 0.5% after full implementation of the prevention bundle (P=.12). This difference was not statistically significant in multivariable regression analysis after adjusting for other patient characteristics (adjusted OR 2.65, 95% CI 0.90–7.81, P=.08).
This study describes the establishment of a gynecology-specific surgical site infection prevention bundle and its effect on surgical site infection rate. Our findings add to existing surgical site infection prevention bundle data and apply current knowledge specifically to hysterectomy.8,19,27,28 We found a statistically significant association between full bundle implementation and the reduced overall surgical site infection rate. Although the stepwise nature of the bundle implementation made analysis less straightforward, we used regression analyses to examine the effects of individual bundle components.
The last bundle component (direct feedback) is different from the preceding six components and is crucial to change in delivery of care. Direct feedback relies on interpersonal interactions and fosters a sense of a shared mission. Formalizing this component is important to the introduction of new processes, because health care providers may be resistant to change, particularly if the data and intentions behind them are not clear. Because surgical site infection incidence continually decreased after formalization of feedback in December 2014 (Fig. 1), we hypothesize that this component has played an important role in changing behavior. Although our analysis did not find a statistically significant effect of this last bundle component, this may be the result of the limited sample size and warrants further investigation.
This study was limited by several challenges in data collection. Our medical record system changed dramatically as the first bundle components were enacted. As such, the data collected before the initial bundle implementation were not available in full and made comparison with the data recorded in the newly implemented electronic medical record difficult. We were also unable to capture outpatient encounters or hospital readmissions for surgical site infections occurring outside our health care system. Although patient characteristics were similar in our prebundle and postbundle groups, some socioeconomic and preoperative health variables and operative variables such as surgeon experience level were unavailable. This makes it difficult to assess overall generalizability of our findings. Furthermore, the bundle, our quality improvement intervention, was introduced sequentially and not as one package. This incremental process does, however, accurately reflect the gradual process by which change takes place on a large surgical service. Although specific data regarding bundle adherence for each surgery were not available, future electronic data collection should allow analysis of specific effects of individual bundle components as well as assessment of any additional improvements over time.
During the study period, there was a decrease of approximately 10% in open procedures from the prebundle to postbundle period. This abrupt change was in part driven by the departure of a single high-volume surgeon who performed many open hysterectomies. We are aware of no other widespread changes in surgical technique or perioperative care occurring during the study period. The shift in proportions of open and minimally invasive hysterectomies contributed to the reduction in surgical site infection rate in open hysterectomies, yet the rate reduction held when stratified by surgical route. Although the reduction in surgical site infection rate was not statistically significant in our laparoscopic groups, our study may not be powered to assess differences based on specific surgical route. Examination of a larger database may clarify the effect of different surgical routes.
The implementation of our perioperative infection prevention bundle was made possible through a multidisciplinary effort in which physicians, nurses, and hospital administration collaborated to address one major quality deficit. The bundle included gynecology and institution-specific components; vaginal and abdominal preparation was standardized with a teaching video and antibiotic guidelines were set based on hospital epidemiology data. Successful implementation was aided by peer teaching and feedback, and the surgical site infection rate decreased throughout the study period and was stable at below expected rates in the last 6 months of the study period.
Although bundle adherence rates were not a study outcome, internal monthly compliance data from July to December 2015 are encouraging. For example, overall compliance with skin preparation protocols went from 86% in July 2015 to 95% in December 2015. One hundred percent of patients received appropriate antibiotics within 1 hour of incision (with 0–7% of benign gynecology patients and 37–65% of gynecology–oncology patients receiving metronidazole, depending on the month).
Although additional analyses are needed to further elucidate the relationships among adherence rates, specific bundle components, hysterectomy routes, length of hospital stay, and surgical site infection reduction and overall surgical costs, we believe that a multidisciplinary, gynecology-specific approach to implementation and maintenance of the surgical site infection prevention bundle serves patients well and will become a mainstay of gynecologic surgical care.
1. Young H, Knepper B, Vigil C, Miller A, Carey JC, Price CS. Sustained reduction in surgical site infection after abdominal hysterectomy. Surg Infect (Larchmt) 2013;14:460–3.
2. Merkow RP, Ju MH, Chung JW, Hall BL, Cohen ME, Williams MV, et al. Underlying reasons associated with hospital readmission following surgery in the United States. JAMA 2015;313:483–95.
3. Broex EC, van Asselt AD, Bruggeman CA, van Tiel FH. Surgical site infections: how high are the costs? J Hosp Infect 2009;72:193–201.
4. de Lissovoy G, Fraeman K, Hutchins V, Murphy D, Song D, Vaughn BB. Surgical site infection: incidence and impact on hospital utilization and treatment costs. Am J Infect Control 2009;37:387–97.
5. Roy S, Patkar A, Daskiran M, Levine R, Hinoul P, Nigam S. Clinical and economic burden of surgical site infection in hysterectomy. Surg Infect (Larchmt) 2014;15:266–73.
6. Bakkum-Gamez JN, Dowdy SC, Borah BJ, Haas LR, Mariani A, Martin JR, et al. Predictors and costs of surgical site infections in patients with endometrial cancer. Gynecol Oncol 2013;130:100–6.
7. Tanner J, Padley W, Assadian O, Leaper D, Kiernan M, Edmiston C. Do surgical care bundles reduce the risk of surgical site infections in patients undergoing colorectal surgery? A systematic review and cohort meta-analysis of 8,515 patients. Surgery 2015;158:66–77.
8. Pellegrini JE, Toledo P, Soper DE, Bradford WC, Cruz DA, Levy BS, et al. Consensus bundle on prevention of surgical site infections after major gynecologic surgery. AANA J 2017;85:1–12.
9. Centers for Medicare and Medicaid Services (CMS), HHS. Medicare program; hospital inpatient prospective payment systems for acute care hospitals and the long-term care hospital prospective payment system and fiscal year 2013 rates; hospitals' resident caps for graduate medical education payment purposes; quality reporting requirements for specific providers and for ambulatory surgical centers. Final rule. Fed Regist 2012;77:53257–750.
10. Edwards JR, Peterson KD, Mu Y, Banerjee S, Allen-Bridson K, Morrell G, et al. National Healthcare Safety Network (NHSN) report: data summary for 2006 through 2008, issued December 2009. Am J Infect Control 2009;37:783–805.
11. Goodman D, Ogrinc G, Davies L, Baker GR, Barnsteiner J, Foster TC, et al. Explanation and elaboration of the SQUIRE (Standards for Quality Improvement Reporting Excellence) guidelines, V.2.0: examples of SQUIRE elements in the healthcare improvement literature. BMJ Qual Saf 2016;25:e7.
12. 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.
13. Antibiotic prophylaxis for gynecologic procedures. ACOG Practice Bulletin No. 104. American College of Obstetricians and Gynecologists. Obstet Gynecol 2009;113:1180–9.
14. Waits SA, Fritze D, Banerjee M, Zhang W, Kubus J, Englesbe MJ, et al. Developing an argument for bundled interventions to reduce surgical site infection in colorectal surgery. Surgery 2014;155:602–6.
15. Schweizer ML, Chiang HY, Septimus E, Moody J, Braun B, Hafner J, et al. Association of a bundled intervention with surgical site infections among patients undergoing cardiac, hip, or knee surgery. JAMA 2015;313:2162–71.
16. Keenan JE, Speicher PJ, Thacker JK, Walter M, Kuchibhatla M, Mantyh CR. The preventive surgical site infection bundle in colorectal surgery: an effective approach to surgical site infection reduction and health care cost savings. JAMA Surg 2014;149:1045–52.
17. Hendren S, Fritze D, Banerjee M, Kubus J, Cleary RK, Englesbe MJ, et al. Antibiotic choice is independently associated with risk of surgical site infection after colectomy: a population-based cohort study. Ann Surg 2013;257:469–75.
18. Meeks DW, Lally KP, Carrick MM, Lew DF, Thomas EJ, Doyle PD, et al. Compliance with guidelines to prevent surgical site infections: as simple as 1-2-3? Am J Surg 2011;201:76–83.
19. Savage MW, Pottinger JM, Chiang HY, Yohnke KR, Bowdler NC, Herwaldt LA. Surgical site infections and cellulitis after abdominal hysterectomy. Am J Obstet Gynecol 2013;209:108.e1–10.
20. Darouiche RO, Wall MJ Jr, Itani KM, Otterson MF, Webb AL, Carrick MM, et al. Chlorhexidine-alcohol versus povidone-iodine for surgical-site antisepsis. N Engl J Med 2010;362:18–26.
21. Hibbard JS. Analyses comparing the antimicrobial activity and safety of current antiseptic agents: a review. J Infus Nurs 2005;28:194–207.
22. Anderson DJ, Podgorny K, Berríos-Torres SI, Bratzler DW, Dellinger EP, Greene L, et al. Strategies to prevent surgical site infections in acute care hospitals: 2014 update. Infect Control Hosp Epidemiol 2014;35(suppl 2):S66–88.
23. Solutions for surgical preparation of the vagina. Committee Opinion No. 571. American College of Obstetricians and Gynecologists. Obstet Gynecol 2013;122:718–20.
24. Kurz A, Sessler DI, Lenhardt R. Perioperative normothermia to reduce the incidence of surgical-wound infection and shorten hospitalization. Study of Wound Infection and Temperature Group. N Engl J Med 1996;334:1209–15.
25. Bratzler DW, Dellinger EP, Olsen KM, Perl TM, Auwaerter PG, Bolon MK, et al. Clinical practice guidelines for antimicrobial prophylaxis in surgery. Surg Infect (Larchmt) 2013;14:73–156.
26. Rosenberger LH, Politano AD, Sawyer RG. The surgical care improvement project and prevention of post-operative infection, including surgical site infection. Surg Infect (Larchmt) 2011;12:163–8.
27. Clinical guidelines and standardization of practice to improve outcomes. Committee Opinion No. 629. American College of Obstetricians and Gynecologists. Obstet Gynecol 2015;125:1027–9.
© 2018 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
28. Sutherland T, Beloff J, Lightowler M, Liu X, Nascimben L, Kaye AD, et al. Description of a multidisciplinary initiative to improve SCIP measures related to pre-operative antibiotic prophylaxis compliance: a single-center success story. Patient Saf Surg 2014;8:37.