Effect of delays in concordant antibiotic treatment on mortality in patients with hospital-acquired Acinetobacter spp. bacteremia in Thailand: a 13-year retrospective cohort

Abstract

Background: A quantitative understanding of the impact of delays to concordant antibiotic treatment on patient mortality is important for designing hospital antibiotic policies. Acinetobacter spp are among the most prevalent pathogens causing multidrug-resistant hospital-acquired infections in developing countries. We aimed to determine the causal effect of delays in concordant antibiotic treatment on 30-day survival of patients with hospital-acquired Acinetobacter spp bacteremia in a resource-limited setting. Methods: We included patients with Acinetobacter spp–related hospital-acquired bacteremia (HAB) in a hospital in Thailand over a 13-year period. We classified patients into 4 groups: those with no delays to concordant antibiotic treatment; those with a 1-day delay; those with 2-day delays; and those with >2 days of delay. We adopted an analytical approach that aimed to emulate a randomized controlled trial and compared the expected potential outcomes of patients between the exposure groups using a marginal structural model with inverse-probability weightings to adjust for confounders and immortal time bias. Results: Between January 2003 and December 2015, 1,203 patients had HAB with Acinetobacter spp., of which 682 patients (56.7%) had ≥1 days of delay in concordant antibiotic treatment. These delays were associated with an absolute increase in 30-day mortality of 6.6% (95% CI 0.2%-13.0%), from 33.8% to 40.4%. Among the 1,203 patients, 521 had no delays to concordant antibiotic treatment (i.e. concordant therapy on the day of blood collection), 224 patients had a 1-day delay, 119 had a 2-day delay, and 339 had a delay of ≥3 days. The crude 30-day mortality was substantially lower in patients with ≥3 days of delay in concordant treatment compared to those with 1 to 2-days of delays. After adjusting for measured confounders and immortal time bias, the expected probability of dying in the hospital within 30-days of blood collection if patient had no delays in concordant therapy was 39.7% (95% CI: 32.3-47.2%), for a 1-day delay it was 42.7% (95% CI: 29.8-55.7%), for a 2-day delay it was 51.0% (95% CI: 38.9-63.2%), and for a ≥3 days was 40.9% (36.0-45.7%).

Conclusions: Delays to concordant antibiotic therapy are linked to increased mortality among patients with HAB due to Acinetobacter spp. Accounting for confounders and immortal time bias is necessary when attempting to estimate causal effects of delayed concordant treatment and, in this case, it helped resolve paradoxical results in crude data.

Funding: The Mahidol Oxford Tropical Medicine Research Unit (MORU) is funded by the Wellcome Trust [grant number 106698/Z14/Z]. CL is funded by a Wellcome Trust Research Training Fellowship [grant number 206736/Z/17/Z]. MY is supported by a Singapore National Medical Research Council Research Fellowship [grant number NMRC/Fellowship/0051/2017]. BSC is funded by the UK Medical Research Council and Department for International Development [grant number MR/K006924/1]. DL is funded by a Wellcome Trust Intermediate Training Fellowship [grant number 101103]. The funder has no role in the design and conduct of the study, data collection, or in the analysis and interpretation of the data.

Disclosures: None