BackgroundThe optimal regimen for perioperative antimicrobial prophylaxis is controversial. Use of combination prophylaxis with a beta-lactam plus vancomycin is increasing; however, the relative risks and benefits associated with this strategy are unknown. Thus, we sought to compare postoperative outcomes following administration of 2 antimicrobials versus a single agent for the prevention of surgical site infections (SSIs). Potential harms associated with combination regimens, including acute kidney injury (AKI) and Clostridium difficile infection (CDI), were also considered.Methods and findingsUsing a multicenter, national Veterans Affairs (VA) cohort, all patients who underwent cardiac, orthopedic joint replacement, vascular, colorectal, and hysterectomy procedures during the period from 1 October 2008 to 30 September 2013 and who received planned manual review of perioperative antimicrobial prophylaxis regimen and manual review for the 30-day incidence of SSI were included. Using a propensity-adjusted log-binomial regression model stratified by type of surgical procedure, the association between receipt of 2 antimicrobials (vancomycin plus a beta-lactam) versus either single agent alone (vancomycin or a beta-lactam) and SSI was evaluated. Measures of association were adjusted for age, diabetes, smoking, American Society of Anesthesiologists score, preoperative methicillin-resistant Staphylococcus aureus (MRSA) status, and receipt of mupirocin. The 7-day incidence of postoperative AKI and 90-day incidence of CDI were also measured. In all, 70,101 procedures (52,504 beta-lactam only, 5,089 vancomycin only, and 12,508 combination) with 2,466 (3.5%) SSIs from 109 medical centers were included. Among cardiac surgery patients, combination prophylaxis was associated with a lower incidence of SSI (66/6,953, 0.95%) than single-agent prophylaxis (190/12,834, 1.48%; crude risk ratio [RR] 0.64, 95% CI 0.49, 0.85; adjusted RR 0.61, 95% CI 0.46, 0.83). After adjusting for SSI risk, no association between receipt of combination prophylaxis and SSI was found for the other types of surgeries evaluated, including orthopedic joint replacement procedures. In MRSA-colonized patients undergoing cardiac surgery, SSI occurred in 8/346 (2.3%) patients who received combination prophylaxis versus 4/100 (4.0%) patients who received vancomycin alone (crude RR 0.58, 95% CI 0.18, 1.88). Among MRSA-negative and -unknown cardiac surgery patients, SSIs occurred in 58/6,607 (0.9%) patients receiving combination prophylaxis versus 146/10,215 (1.4%) patients who received a beta-lactam alone (crude RR 0.61, 95% CI 0.45, 0.83). Based on these associations, the number needed to treat to prevent 1 SSI in MRSA-colonized patients is estimated to be 53, compared to 176 in non-MRSA patients. CDI incidence was similar in both exposure groups. Across all types of surgical procedures, risk of AKI was increased in the combination antimicrobial prophylaxis group (2,971/12,508 [23.8%] receiving combination versus 1,058/5,089 [20.8%] receiving vancomycin alone...

Recent work has demonstrated that propensity score matching may lead to increased covariate imbalance, even with the corresponding decrease in propensity score distance between matched units. The extent to which this paradoxical phenomenon might harm causal inference in real epidemiologic studies has not been explored. We evaluated the effect of this phenomenon using insurance claims data from the Pharmaceutical Assistance Contract for the Elderly (1999-2002) and Medicaid Analytic eXtract (2000-2007) databases in the United States. For each data set, we created several 1:1 propensity-score-matched data sets by manipulating the size of the covariate set used to generate propensity scores, the index exposure prevalence in the prematched data set, and the matching algorithm. We matched all index units, then progressively pruned matched sets in order of decreasing propensity score distance, calculating covariate imbalance after each pruning. Although covariate imbalance sometimes increased after progressive pruning of matched sets, the application of commonly used propensity score calipers for defining an acceptable match stopped pruning near the lowest region of the imbalance trend and resulted in an improvement over the imbalance in the prematched data set. Thus, propensity score matching does not appear to induce increased covariate imbalance when standard propensity score calipers are applied in these types of pharmacoepidemiologic studies.

Coarsened exact matching (CEM) is a matching method proposed as an alternative to other techniques commonly used to control confounding. We compared CEM with 3 techniques that have been used in pharmacoepidemiology: propensity score matching, Mahalanobis distance matching, and fine stratification by propensity score (FS). We evaluated confounding control and effect-estimate precision using insurance claims data from the Pharmaceutical Assistance Contract for the Elderly (1999–2002) and Medicaid Analytic eXtract (2000–2007) databases (United States) and from simulated claims-based cohorts. CEM generally achieved the best covariate balance. However, it often led to high bias and low precision of the risk ratio due to extreme losses in study size and numbers of outcomes (i.e., sparse data bias)—especially with larger covariate sets. FS usually was optimal with respect to bias and precision and always created good covariate balance. Propensity score matching usually performed almost as well as FS, especially with higher index exposure prevalence. The performance of Mahalanobis distance matching was relatively poor. These findings suggest that CEM, although it achieves good covariate balance, might not be optimal for large claims-database studies with rich covariate information; it might be ideal if only a few (<10) strong confounders must be controlled.