Need for Speed: Lower mortality and earlier targeted treatment with BCID2 panel: Preliminary report from an Indian ICU

Background: BIOFIRE Blood Culture Identification 2(BCID2) panel, a rapid molecular diagnostic platform, identifies the organism and its molecular mechanism of resistance within two hours of the blood culture bottle flagging positive. This is the first Indian study to evaluate the impact of this panel on real-world antimicrobial stewardship practices and clinical outcomes in an Indian ICU. Methods: This pre-post quasi-experimental study; performed at a tertiary care centre ICU in Mumbai, India; comprised of two cohorts: The Intervention/Prospective Cohort[BCID2](1st January 2023- 31st October 2023) and the Control/Retrospective cohort[Standard of Care (SOC)](1st January 2020- 31st October 2020). ICU patients with documented bacteremia/fungemia were included. Phenotypic Blood culture and susceptibility testing [using current CLSI breakpoints on VITEK-2(BioMérieux, France)] was performed as SOC in both cohorts. BCID2 panel(Biomerieux, France) was run additionally in the Intervention cohort when the blood culture bottle flagged positive. Primary and secondary outcomes were recorded. (Figure 3) Interim analysis for the first six months (January-June) is presented.Statistical analysis was performed on MedCalc(V22). Continuous and categorical variables were compared using Mann-Whitney U test and chi-square respectively. Univariate analysis was performed to analyse predictors of mortality. Logistic regression analysis was performed for variables demonstrating p<0.05 to assess independent association. Tests were interpreted at 5% significance. Results: The final analysis comprised of 145 patients in the Intention to Treat (ITT) Analysis. Both cohorts had similar baseline characteristics (Figure 2). Time to appropriate therapy was shorter in the BCID2 cohort, but was statistically insignificant [5.75 hours vs 15 hours, p=0.369). However, after excluding those on appropriate antibiotics prior to Gram stain and where contaminants were not treated [Per protocol (PP) analysis, Figures 1 and 3], the BCID2 cohort showed a statistically significant shorter time from Gram stain to appropriate therapy compared to SOC [17 hours vs 35.75 hours, p=0.026]. Time to organism identification, and in-hospital mortality were significantly lower in the BCID2 cohort (ITT analysis) (Figure 3).Logistic regression analysis in the ITT analysis showed that BCID2 use was independently associated with 67% lower incidence of mortality [aOR: 0.33 (95%CI:0.14-0.75), p=0.0078]. Pitts Bacteremia score [aOR: 1.38 (95%CI:1.20-1.6), p=0.0001] and invasive candidiasis [aOR: 20.32 (95% CI:1.83-225.77), p=0.014] were other independent predictors of mortality. Discussion: Despite its limitations (non-randomized study, lack of data on antibiotic consumption and financial outcomes), the preliminary findings of this study demonstrated earlier targeted treatments and mortality benefit in a real-world setting with the BCID2 panel use. This study also gave us an opportunity to reflect on the gaps in our antimicrobial prescribing practices and the need to implement earlier de-escalation strategies. Conclusion: With a significant impact on stewardship measures and clinical outcomes (especially mortality), BCID2 panel holds promise in an Indian ICU.