# Early discontinuation of antibiotic therapy initiated in the emergency department in older patients: a retrospective study

## TL;DR

This study examines whether stopping antibiotics early in older patients after emergency treatment is safe and finds mixed results.

## Contribution

The study evaluates the safety and outcomes of early antibiotic discontinuation in older patients after emergency treatment.

## Key findings

- Early antibiotic discontinuation occurred in 23.9% of patients and was associated with fewer antibiotic days and lower mortality.
- Non-inferiority for antibiotic resumption due to recurrence was not demonstrated, with a 4.2% risk difference.
- Discontinuation was linked to no increase in ICU admissions and lower 30-day mortality despite less severe baseline conditions.

## Abstract

Infectious diseases are a leading cause of emergency department (ED) visits in elderly patients, and their accurate diagnosis and management remain a challenge. In this context, empirical antibiotic therapy may be discontinued upon patient admission to the acute geriatric unit (AGU) during routine early clinical reassessment. However, the safety of this practice remains unclear. This study evaluated the risk of antibiotic therapy resumption due to recurrence of the initial clinical infectious syndrome after early discontinuation (≤ 48 h) in older patients admitted to the AGU.

This single-center retrospective observational study included patients aged over 75 years admitted to the AGU with suspected bacterial infections who received at least one dose of antibiotic therapy in the ED.

The primary outcome was the rate of antibiotic therapy resumption for recurrent infections. The secondary outcomes included total antibiotic therapy consumption (in days), side effects, 30-day ED revisits and 30-day all-cause mortality. Noninferiority was tested with a 10% margin.

Among 213 patients admitted to the AGU previously receiving empirical antibiotic therapy in the ED, early antibiotic therapy discontinuation occurred in 51/213 (23.9%) patients. The “early discontinuation” group of patients had lower qSOFA scores and lactate levels. Non-inferiority for antibiotic therapy resumption due to recurrence was not demonstrated because the upper bound of the confidence interval exceeded the prespecified 10% non-inferiority margin (Absolute risk difference 4.2% 95% CI [-5.9; 14.4]). Early discontinuation had lower antibiotic therapy days (1 day vs 7 days, p < 0.001), no increase in ICU admissions (2.0% vs 2.5%, p = 1.000), and lower 30-day all-cause mortality rate (9.8% vs 27.2%, p = 0.017), but the discontinuation group was less severe at baseline (qSOFA, lactate, CRP, albumin).

Although noninferiority was not reached in terms of antibiotic therapy resumption, early discontinuation of empirical antibiotic therapy initiated in the ED appears safe in selected older adults without sepsis criteria, with infrequent antibiotic therapy resumption and decrease in antibiotic therapy consumption with no increase in ICU transfers and mortality.

The online version contains supplementary material available at 10.1186/s12877-026-06979-w.