P-876. Decision Analysis of Short versus Ultra-Long Course Antibiotics for Complicated Intraabdominal Infection

TL;DR

This study compares short and ultra-long antibiotic courses for intraabdominal infections using a decision model to determine which strategy offers better outcomes.

Contribution

The novel contribution is the use of a decision analysis model to compare short versus ultra-long antibiotic courses for complicated intraabdominal infections.

Findings

Short course therapy had a higher overall utility (0.97) compared to ultra-long course therapy (0.92) in base-case analysis.

Probabilistic sensitivity analysis showed short course therapy was optimal in over 99% of model iterations.

The model suggests short course therapy may be more effective in terms of utility despite potential recurrence risks.

Abstract

Short course antibiotic therapy (4 days) following source control for intraabdominal infections has become standard of care. However, recent data suggests that ultra-long course durations (28 days) after source control may drastically reduce recurrence rates. A randomized trial of ultra-long course therapy for complicated intraabdominal infections is currently enrolling. The purpose of this study is to evaluate these two treatment strategies using a decision analysis model.

A decision tree was created from the perspective of a healthcare administrator. The model evaluates the rates of recurrence between these two treatment strategies as well as potential negative impacts of overtreatment and per-day risk of antibiotic-associated adverse events. Model values were taken from the published literature or estimated by the authors. Base-case and probabilistic sensitivity analyses were performed. The model was created using TreeAge Pro Healthcare (TreeAge Software, Inc., Williamstown, MA).

On base-case analysis, short course therapy maximized utility with an overall utility of 0.97 versus 0.92 for ultra-long course therapy. Probabilistic sensitivity analysis, after 100,000 iterations identified short course therapy as the optimal strategy for >99% of model iterations.

In this exploratory model, short course therapy resulted in higher overall utility values compared to ultra-long course treatment. Further work is needed to confirm these findings, including completion of the currently enrolling EXTEND trial.

All Authors: No reported disclosures