RESUMEN
OBJECTIVES: Bartonella spp., renowned for cat-scratch disease, has limited reports of dissemination. Tissue and blood cultures have limitations in detecting this fastidious pathogen. Molecular testing (polymerase chain reaction, PCR) and cell-free DNA have provided an avenue for diagnoses. This retrospective observational multicenter study describes the incidence of disseminated Bartonella spp. and treatment-related outcomes. METHODS: Inclusion criteria were diagnosis of bartonellosis via diagnosis code, serology testing of blood, polymerase chain reaction (PCR) of blood, 16/18S tests of blood or tissue, cultures of blood or tissue, or cell-free DNA of blood or tissue from January 1, 2014, through September 1, 2021. Exclusions were patients who did not receive treatment, insufficient data on treatment course, absence of dissemination, or retinitis as dissemination. RESULTS: Patients were primarily male (n = 25, 61.0%), white (n = 28, 68.3%), with mean age of 50 years (SD 14.4), and mean Charlson comorbidity index of 3.5 (SD 2.1). Diagnosis was primarily by serology (n = 34, 82.9%), with Bartonella henselae (n = 40, 97.6%) as the causative pathogen. Treatment was principally doxycycline with rifampin (n = 17, 41.5%). Treatment failure occurred in 16 (39.0%) patients, due to escalation of therapy during treatment (n = 5, 31.3%) or discontinuation of therapy due to an adverse event or tolerability (n = 5, 31.3%). CONCLUSIONS: In conclusion, this is the largest United States-based cohort of disseminated Bartonella spp. infections to date with a reported 39% treatment failure. This adds to literature supporting obtaining multiple diagnostic tests when Bartonella is suspected and describes treatment options.
RESUMEN
Objective: Determine the impact of limited implementation of a rapid blood culture identification (BCID) panel. Design: Retrospective cohort study. Methods: From February to April 2022, positive blood cultures identified via e-Plex BCID (Roche, Carlsbad, CA) were compared to those identified using standard microbial identification techniques. The primary outcomes assessed were time to optimal therapy, time to de-escalation of anti-MRSA (methicillin-resistant Staphylococcus aureus) agents, and time to de-escalation of anti-pseudomonal agents. Additional analysis investigated the impact of the availability of antimicrobial stewardship program support. This study was conducted at Grady Health System, a large metropolitan safety-net hospital in the southeastern United States. Results: A total of 253 blood cultures were included in this study (153 BCID and 100 standard). Blood culture identification use was associated with a reduction in median time to optimal antimicrobial therapy (43.4 vs 72.1 h, P < .001) and median time to de-escalation of anti-MRSA agents (27.7 vs 46.7 h, P = .006), and a trend towards reduction of median time to de-escalation of anti-pseudomonal agents (38.8 vs 54.8 h, P = .07). These reductions persisted when controlling for patient age, sex, intensive care unit status, Charlson Comorbidity Index, and antimicrobial stewardship program availability. Conclusions: Despite restricted use and lack of 24/7 antimicrobial stewardship program availability, BCID panel utilization was associated with earlier initiation of optimal therapy and pathogen identification with subsequent de-escalation of broad-spectrum antimicrobials, as compared to standard antimicrobial techniques. This suggests the potential for benefit from adopting novel diagnostic technologies outside of idealized fully-resourced settings.