473. Impact of Adjunctive Clindamycin in Invasive β-Hemolytic Streptococcal Infections: A Propensity Score-matched Analysis of 1956 β-lactam Treated Patients from 118 US Hospitals

Abstract

BackgroundClindamycin suppresses invasive β-hemolytic streptococcal (iBHS) toxin production and its cidality, unlike β-lactams, is independent of bacterial inoculum. Evidence favoring the use of adjunctive clindamycin in iBHS is predominantly from in vitro and animal studies. Clinical studies, limited to single-center series or active surveillance, have yielded mixed results. Furthermore, its role in infections caused by non–group A or B β-hemolytic streptococci (NABS) remains poorly defined. Previously we were unable to demonstrate a clindamycin survival benefit in patients with group A streptococci (GAS) bacteremia. Here we examined the impact of adjunctive clindamycin on survival among patients with either invasive GAS or NABS infections from all sites.MethodsClinical characteristics, in vitro susceptibility and antibiotic therapy were examined for unique inpatient encounters with iBHS (GAS + NABS) infections in the Cerner HealthfactsTM Database. β-Lactam treated cases receiving clindamycin were matched 1:1 to non-clindamycin cases by propensity of receiving clindamycin using nearest neighbor propensity score matching with downstream adjustment using logistic regression. In-hospital mortality, length of stay (LOS) and Sequential Organ Failure (SOFA) score trajectory was compared based on the matched sample.ResultsOf 1,956 inpatients with iBHS infection treated with β-lactams at 118 hospitals between 2000 and 2015, 459 (23.5%) received adjunctive clindamycin. Propensity score matching generated 389 case-pairs with good covariate balance (Table 1, Figure 1). In-hospital mortality did not differ between matched clindamycin and non-clindamycin cases (7.2% vs. 8.0%, P = 0.66, aOR 0.88, [95% CI .49–1.57]) (Figure 2). Receipt of clindamycin early (within 24h vs. no clindamycin) also did not display a survival advantage (aOR 1.04 [0.51–2.14]). On day 4 of therapy Median SOFA score (P = 0.586) and SOFA delta (day 0–day 4) were similar between the two groups (P = 0.13; Figure 3). Amongst survivors, median [IQR] LOS was greater in the clindamycin group (8[5,12] vs. 6[4,9]; P = 0.001)ConclusionAdjunctive clindamycin was not associated with decreased mortality or degree of organ dysfunction among patients with iBHS infection already treated with β-lactams. Figure 2: Odds Ratio of In-Hospital Mortality Abbreviations: GAS: Group A Streptococcus , ICU: intensive care unit, iβHS: invasive β-Hemolytic streptococcus, NABS: Non group A β-Hemolytic streptococci , OR: odds ratioFigure 2 Legend: The figure reports the odds ratios (ORs) of in-hospital mortality and 95% confidence intervals in the unmatched and unadjusted analysis, matched and unadjusted analysis, and primary analysis of all propensity-matched pairs with downstream adjustment with logistic regression for proven ißHS, vasopressor use and ICU (within 24h of culture sampling), as well as sensitivity analyses on propensity matched pairs of 2:1 match and subgroup analysis on propensity matched pairs of (1) proven ißHS alone (2) probable ißHS alone (3) ICU admission (within 24h of culture sampling) (4) patients receiving vasopressor therapy (within 24h of culture sampling) (5) Group A ißHS alone (6) Non Group A ißHS alone. There was no statistically significant difference in the ORs for in-hospital mortality between clindamycin and propensity-matched non-clindamycin cases in the primary analysis (*) as well as all sensitivity and subgroup (†) analyses.Figure 3: SOFA Score Trajectory by Survival Status Abbreviations: B: Non-Clindamycin cases, BC: Clindamycin CasesFigure 3 Legend: SOFA score by day of therapy of clindamycin and non-clindamycin matched cases based on survival status from day zero (day prior to antibiotic therapy) to day four of therapy. The linear mixed models were used to assess the time trends and the clindamycin effect on the longitudinal SOFA scores. SOFA trajectory was examined for the post matching sample. Square root transformation was applied to the SOFA score to meet the normality assumption Mean Baseline SOFA scores prior to therapy were similar amongst clindamycin and non-clindamycin subjects (mean [standard deviation(SD)] SOFA score: 1.88[2.48] vs. 1.96[2.60]; p= 0.634). On day 4 of therapy SOFA scores were similar between remaining 310 clindamycin and 286 non-clindamycin hospitalized patients (mean[SD] SOFA score: 1.79[2.88] vs. 1.67 [2.49]; p=0.586). The SOFA delta (day 0 SOFA score - day 4 SOFA score) was similar between the two groups (p=0.1331). When examined amongst survivors only, SOFA scores on day 4 of therapy were similar between and 272 non-clindamycin and 310 clindamycin hospitalized patients (mean[SD] SOFA score: 1.45 [2.20] vs. 1.52 [2.44])This work was funded in part by the Intramural Research Program of the National Institutes of Health Clinical Center and the National Institute of Allergy and Infectious Diseases.The opinions expressed in this abstract are those of the authors and do not represent any position or policy of the National Institutes of Health, the United States Department of Health and Human Services, or the United States government.Disclosures All authors: No reported disclosures.