Delays in treating bacteremias with antibiotics to which the causative organism is susceptible are expected to adversely affect patient outcomes. Quantifying the impact of such delays to concordant treatment is important for decision-making about interventions to reduce the delays and for quantifying the burden of disease due to antimicrobial resistance. There are, however, potentially important biases to be addressed, including immortal time bias. We aimed to estimate the impact of delays in appropriate antibiotic treatment of patients with Acinetobacter species hospital-acquired bacteremia in Thailand on 30-day mortality by emulating a target trial using retrospective cohort data from Sunpasitthiprasong Hospital in 2003-2015. For each day, we defined treatment as concordant if the isolated organism was susceptible to at least 1 antibiotic given. Among 1,203 patients with Acinetobacter species hospital-acquired bacteremia, 682 had 1 or more days of delays to concordant treatment. Surprisingly, crude 30-day mortality was lower in patients with delays of ≥3 days compared with those who had 1-2 days of delays. Accounting for confounders and immortal time bias resolved this paradox. Emulating a target trial, we found that these delays were associated with an absolute increase in expected 30-day mortality of 6.6% (95% confidence interval: 0.2, 13.0), from 33.8% to 40.4%.
Keywords: Acinetobacter species; bacteremia; causal inference; empirical antibiotic treatment; patient mortality.
© The Author(s) 2020. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of Public Health.