Abstract
Factors driving vancomycin surgical prophylaxis are poorly understood. In a national Veterans Affairs cohort with manually validated data, surgical specialty (cardiac, orthopedics) and perception of high facility methicillin-resistant Staphylococcus aureus (MRSA) prevalence—not MRSA colonization—were the primary drivers of prescribing. A β-lactam allergy was the second most common reason. These data may inform perioperative stewardship.
Keywords: antimicrobial prophylaxis, vancomycin, stewardship
Although prevalence of methicillin-resistant Staphylococcus aureus (MRSA) colonization has decreased over the past decade, and rates of MRSA surgical site infections (SSIs) have decreased within the Department of Veterans Affairs (VA) [1, 2], use of vancomycin as an agent for surgical prophylaxis remains common [3]. Furthermore, recent studies demonstrate that choice of vancomycin for surgical prophylaxis may be associated with higher rates of adverse events [4, 5], including SSI, among patients who received the drug due to an allergy or intolerance to β-lactam antibiotics [6].
Despite these secular trends and potential for increased patient harm [4], factors driving choice of vancomycin for surgical prophylaxis are unknown. Understanding the institutional, surgical, and patient-level factors that drive vancomycin prescribing may improve perioperative antimicrobial stewardship and thereby improve outcomes following major surgical procedures. Thus, we sought to characterize drivers of vancomycin prophylaxis among a large, national, multicenter cohort of Veterans undergoing major surgical procedures.
METHODS
Veterans who underwent major cardiac, orthopedic total joint replacement, vascular, or colorectal procedures and were entered into the External Peer Review Program (EPRP) database during the period from 1 October 2008 through 30 September 2013 were included. Importantly, during the study period, the EPRP program included manual review of a selected number of surgeries for type and duration of prophylaxis; if vancomycin was selected, then the reason for vancomycin use was abstracted manually by a trained nurse chart reviewer and entered into the database. The manual review program was discontinued after the relevant Surgical Care Improvement Project SCIP infection (INF) measures were sunsetted in 2015 [7].
Patients in the EPRP dataset were then matched to the VA Surgical Quality Improvement Program database, which included postoperative variables, such as 30-day SSI outcomes and 7-day acute kidney injury (AKI) outcomes; because 2 databases were used, outcomes were not available for all patients included in the cohort. Both datasets are described in more detail in previously published work [4, 5]. Preoperative MRSA colonization status was extracted from the VA Corporate Data Warehouse, and was defined as any positive MRSA nasal swab and/or culture during the 30 days prior to surgery, inclusive of the date of the surgical procedure [5].
Reasons for vancomycin prescribing were classified into 5 categories: β-lactam allergy and/or intolerance; patient perceived to be at high risk of MRSA colonization (eg, prolonged hospitalization prior to surgery, nursing home residence); patient history of MRSA and/or known colonization of MRSA; facility perceived to have high risk of MRSA and/or surgery perceived to carry high risk of MRSA infection; and other/not recorded. The EPRP program allowed patients to be entered into multiple categories.
After classification, reasons for vancomycin use were stratified by surgical specialty, and factors driving vancomycin use were compared. Potential confounding factors determined a priori (eg, age, sex, race, diabetes, American Society of Anesthesiologists score, MRSA preoperative screen, and smoking status) were also evaluated.
Recognizing limited power given a low SSI event rate and thus a limited number of outcomes, we conducted an exploratory univariate analysis evaluating the rate of SSI among cardiac and orthopedic surgery patients who received vancomycin alone for allergy or intolerance only, compared with patients who received vancomycin alone for any other reason; for the purposes of this exploratory analysis, rates of SSI among patients who received a β-lactam without vancomycin surgical prophylaxis are also presented. We conducted a similar analysis exploring the relationship between AKI among patients who received vancomycin alone due to an allergy or intolerance, vancomycin alone for another reason, and a β-lactam without vancomycin.
The VA Boston Institutional Review Board approved this research prior to data collection and analysis.
RESULTS
Among 79 092 cardiac, orthopedic, vascular, and colorectal procedures performed at 109 different VA facilities nationwide, 20 388 (25.8%) received vancomycin either alone (n = 6885 [33.8%]) or in combination with another agent for surgical prophylaxis. Among these patients, 14 200 had a reason for vancomycin documented in the EPRP dataset and 6188 had no indication recorded (Table 1). In total, 2906 (3.68%) patients were colonized with MRSA preoperatively; 1116 of these patients (38.4%) received vancomycin. Use of vancomycin as a prophylactic agent was highest among cardiac surgeries (48.9%) followed by orthopedic procedures (20.8%); utilization in vascular and colorectal procedures was substantially lower (22.6% and 3.3%, respectively) (Supplementary Figure 1).
Table 1.
Baseline Characteristics of Patients Who Received Vancomycin for Surgical Prophylaxis
| Characteristic | Reason for Vancomycin Usea | |||||
|---|---|---|---|---|---|---|
| Allergy or Intolerance | Facility and/or Surgery With High MRSA Risk | High Patient Risk of MRSA | Patient With Known MRSA or History of MRSA | Other Reason Documented | No Reason Documented | |
| No. of patients | 4860 | 7372 | 2188 | 503 | 803 | 6188 |
| Vancomycin only | 4049 (83.3) | 1723 (23.4) | 764 (34.9) | 190 (37.8) | 407 (50.7) | 601 (9.7) |
| Age, y, mean (SD) | 65.66 (9.24) | 65.15 (8.75) | 66.14 (8.91) | 65.69 (9.23) | 65.14 (8.61) | 64.82 (9.21) |
| Female sex | 306 (6.3) | 207 (2.8) | 34 (1.6) | 16 (3.2) | 27 (3.4) | 194 (3.1) |
| Race/ethnicity | ||||||
| American Indian | 24 (0.5) | 46 (0.6) | 9 (0.4) | 4 (0.8) | 8 (1.0) | 40 (0.6) |
| Asian | 10 (0.2) | 39 (0.5) | 19 (0.9) | 0 (0.0) | 4 (0.5) | 14 (0.2) |
| Black | 549 (11.3) | 1000 (13.6) | 253 (11.6) | 65 (12.9) | 118 (14.7) | 1118 (18.1) |
| Hawaiian/Pacific Islander | 20 (0.4) | 46 (0.6) | 16 (0.7) | 2 (0.4) | 8 (1.0) | 48 (0.8) |
| Other/not known | 321 (6.6) | 436 (5.9) | 164 (7.5) | 26 (5.2) | 57 (7.1) | 405 (6.5) |
| White | 3936 (81.0) | 5805 (78.7) | 1727 (78.9) | 406 (80.7) | 608 (75.7) | 4563 (73.7) |
| Diabetes | 1561 (32.1) | 2343 (31.8) | 835 (38.2) | 159 (31.6) | 244 (30.4) | 1869 (30.2) |
| ASA score > 2 | 4362 (89.8) | 6570 (89.1) | 2154 (98.4) | 447 (88.9) | 729 (90.8) | 5264 (85.1) |
| Smoking | 1410 (29.0) | 1890 (25.6) | 673 (30.8) | 144 (28.6) | 230 (28.6) | 1727 (27.9) |
| Preoperative MRSA statusb | 220 (4.5) | 269 (3.6) | 157 (7.2) | 190 (37.8) | 35 (4.4) | 333 (5.4) |
| Surgical type | ||||||
| Cardiac | 2075 (42.7) | 4173 (56.6) | 1683 (76.9) | 113 (22.5) | 390 (48.6) | 2903 (46.9) |
| Colorectal | 78 (1.6) | 41 (0.6) | 41 (1.9) | 28 (5.6) | 40 (5.0) | 151 (2.4) |
| Orthopedic | 2192 (45.1) | 2840 (38.5) | 228 (10.4) | 317 (63.0) | 290 (36.1) | 2719 (43.9) |
| Vascular | 515 (10.6) | 318 (4.3) | 236 (10.8) | 45 (8.9) | 83 (10.3) | 415 (6.7) |
Data are presented as no. (%) unless otherwise indicated.
Abbreviations: ASA, American Society of Anesthesiologists; MRSA, methicillin-resistant Staphylococcus aureus; SD, standard deviation.
aPatients could be entered into > 1 category.
bPreoperative MRSA status includes any positive nasal swab up to and including the day of surgery.
Only 3.2% of the vancomycin use for surgical prophylaxis was administered for known MRSA colonization or history of MRSA infection (503/15 726). Perceptions of high facility rates of MRSA and/or high-risk procedures for MRSA were the most common reasons documented for vancomycin administration overall (n = 7372 [46.9%]), and among cardiac and orthopedic surgeries (49.5% and 48.4%, respectively; Supplementary Figure 2). Allergy and/or intolerance to a β-lactam antibiotic was the second most common reason overall (n = 4860 [34%]) and the most common reason among vascular (43.0%) and colorectal (48.4%) procedures. A perceived high personal risk of MRSA, either due to prolonged healthcare exposure, prolonged hospital admission, nursing home residence, or MRSA colonization was the third most common reason (n = 2188 [15%]).
Prevalence of MRSA colonization was higher among patients who received vancomycin than those who did not (1116/20 349 [5.48%] vs 1790/58 079 [3.08%]; relative risk [RR], 1.78; P < .0001). This was particularly true for patients who had a history of MRSA colonization (190/503 [37.8%]). Limiting the analysis to patients who only had perceived high facility risk of MRSA as the reason for vancomycin use and no other cited factors, there was no increase in prevalence of MRSA colonization when compared to the general surgical population (227/6281 [3.6%] vs 2680/72 717 [3.7%]; P = .79). Patient-level risk in the absence of a known MRSA history (eg, nursing home residence, prolonged hospitalization) was predictive of preoperative MRSA colonization (113/1578 [7.2%]; RR, 1.95; P < .001) among patients who received vancomycin prophylaxis.
Outcomes
The incidence of SSI was similar among patients who received a β-lactam without vancomycin (most commonly cefazolin, followed by cefoxitin), vancomycin alone for an allergy, and vancomycin alone for any other reason in both cardiac cases (β-lactam: 166/10 869 [1.5%]; vancomycin for allergy: 25/1660 [1.5%]; vancomycin for other reasons: 25/1478 [1.7%]) and orthopedic cases (β-lactam: 369/29 113 [1.3%]; vancomycin for allergy: 22/1851 [1.2%]; and vancomycin for other reasons: 15/953 [1.6%]; P = .70). Incidence of AKI was lower in patients who received a β-lactam for surgical prophylaxis compared with patients who received vancomycin alone (RR, 0.88 and 0.70 for cardiac and orthopedics, respectively; both P < .001); risk of AKI after receipt of vancomycin did not differ by indication for vancomycin. Patients who received the antibiotic due to allergy or intolerance to a β-lactam had similar incidence to patients who received it for other reasons (P = .12 and P = .13 for cardiac and orthopedics, respectively).
DISCUSSION
This study demonstrates that the most common reason for vancomycin prophylaxis is a perception of high facility rate of MRSA; however, this variable had no predictive value for identifying patients with preoperative colonization; these findings may be leveraged by antimicrobial stewardship programs to improve choice of surgical prophylaxis regimen for major surgeries and to optimize postoperative outcomes. Other major drivers of perioperative vancomycin use included surgical specialty and type of surgery. Patient-level risk factors, including known MRSA colonization and/or healthcare exposure, were substantially stronger predictors of preoperative colonization status but a less commonly cited driver of vancomycin for surgical prophylaxis.
The second most common reason for vancomycin use was β-lactam allergy; this was the most common reason for colorectal and vascular procedures, where SSI pathogens are more varied and perioperative prophylaxis is typically designed to cover not only typical skin colonizers, but also gram-negative and anaerobic organisms. Our findings expand upon the current literature by characterizing the multiple factors driving perioperative antimicrobial use and suggest strategies that can be used to optimize clinical outcomes.
Recognizing a small number of outcomes and therefore a limited power to detect a difference between them, we found that incidence of SSI was similar in patients who received vancomycin due to allergy or intolerance vs those who received it for another reason. We also found that risk of AKI was elevated in patients who received vancomycin alone—regardless of the factor driving the vancomycin use—when compared to patients who received a β-lactam–based regimen; because the analysis excluded patients who received combinations of vancomycin plus a β-lactam, the increase cannot be driven by the combined toxicity of the 2 agents [8]. These data suggest that implementing antimicrobial stewardship interventions designed to limit the use of vancomycin for surgical prophylaxis, such as preoperative MRSA screening programs and preoperative allergy testing for patients with documented β-lactam allergies [9], may lead to improved postoperative outcomes by reducing unnecessary nephrotoxic exposures. Additional benefits of limiting vancomycin prophylaxis may include increased compliance with appropriate timing of preoperative antimicrobial administration, as vancomycin is often administered outside of recommended time windows; noncompliance with appropriate administration is associated with higher rates of adverse outcomes [6, 10].
Limitations of the study include the retrospective, observational nature of the findings, the predominantly male population within the VA healthcare system, and the small number of outcomes, which restricts our ability to make definitive conclusions about whether SSI rates differ between patients who received vancomycin for allergy vs for other reasons. However, despite these limitations, this is a robust study examining factors driving perioperative vancomycin use, and these findings can be used to augment antimicrobial stewardship programs to optimize perioperative care. Data are derived from a unique high-quality dataset encompassing a national cohort with manually compiled reasons for perioperative vancomycin use; findings provide an important snapshot of factors that drive clinical decision making and that ultimately impact surgical outcomes.
Supplementary Data
Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
Notes
Acknowledgments. The authors thank John Ripollone, MPH, for his assistance with database creation and model development. This work would not have been possible without the collaboration and resources of the Department of Veterans Affairs (VA) External Peer Review Program, the VA Office of Analytics and Business Intelligence, the VA Surgical Quality Improvement Program and the Surgical Quality Data Use Group, and the VA Informatics and Computing Infrastructure.
Disclaimer. The opinions expressed are those of the authors and not necessarily those of the VA or the United States government.
Financial support. This work was funded by VA Health Services Research and Development Award IIR 12–103 (to K. G.); W. B.-E. is supported by the National Heart, Lung, and Blood Institute (grant number 1K12HL138049-01).
Potential conflicts of interest. K. G., K. I., and W. O. are co-investigators on a grant from Pfizer to the VA Boston Healthcare System. K. I. has served as a research consultant for Eupraxia. W. B.-E. has served as an expert witness for DLA Piper, LLC. All other authors report no potential conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
Presented in part: IDWeek 2019, Washington, District of Columbia, 2–6 October 2019. Poster #1100.
References
- 1. Diekema DJ, Pfaller MA, Shortridge D, Zervos M, Jones RN. Twenty-year trends in antimicrobial susceptibilities among Staphylococcus aureus from the SENTRY antimicrobial surveillance program. Open Forum Infect Dis 2019; 6:47–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. O’Brien WJ, Gupta K, Itani KM. A longitudinal study of S. aureus infection in a national cohort of surgical patients [manuscript published online ahead of print 13 August 2019]. Open Forum Infect Dis 2019. doi:10.1093/ofid/ofz350. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Strymish J, Branch-Elliman W, Itani K, Gupta K. National trends in vancomycin use for peri-operative prophylaxis over 5 years. Open Forum Infect Dis 2015; 2. doi:10.1093/ofid/ofv133.235 [Google Scholar]
- 4. Branch-Elliman W, O’Brien W, Strymish J, Itani K, Wyatt C, Gupta K. Association of duration and type of surgical prophylaxis with antimicrobial-associated adverse events. JAMA Surg 2019; 154:590–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Branch-Elliman W, Ripollone JE, O’Brien WJ, et al. Risk of surgical site infection, acute kidney injury, and Clostridium difficile infection following antibiotic prophylaxis with vancomycin plus a beta-lactam versus either drug alone: a national propensity-score-adjusted retrospective cohort study. PLoS Med 2017; 14:e1002340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Blumenthal KG, Ryan EE, Li Y, Lee H, Kuhlen JL, Shenoy ES. The impact of a reported penicillin allergy on surgical site infection risk. Clin Infect Dis 2017; 66:329–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. The Joint Commission. Specifications manual for Joint Commission national quality core measures (2010A1): measure information form SCIP-Inf-3. 2010 Available at: https://manual.jointcommission.org/releases/archive/TJC2010B/MIF0112.html. Accessed 3 August 2018.
- 8. Tong SYC, Lye DC, Yahav D, et al. ; Australasian Society for Infectious Diseases Clinical Research Network Effect of vancomycin or daptomycin with vs without an antistaphylococcal β-lactam on mortality, bacteremia, relapse, or treatment failure in patients with MRSA bacteremia: a randomized clinical trial. JAMA 2020; 323:527–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Reilly CA, Backer G, Basta D, Riblet NB, Hofley PM, Gallagher MC. The effect of preoperative penicillin allergy testing on perioperative non-beta-lactam antibiotic use: a systematic review and meta-analysis. Allergy Asthma Proc 2018; 39:420–9. [DOI] [PubMed] [Google Scholar]
- 10. Burger JR, Hansen BJ, Leary EV, Aggarwal A, Keeney JA. Dual-agent antibiotic prophylaxis using a single preoperative vancomycin dose effectively reduces prosthetic joint infection rates with minimal renal toxicity risk. J Arthroplasty 2018; 33:213–8. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.