Standard precautions versus clinically triggered contact precautions for prevention of extended-spectrum β-lactamase-producing enterobacterales in acute-care geriatric units: A prospective non-inferiority interventional study protocol (GERSP-study)

Abstract

Introduction Extended-spectrum β-lactamase–producing Enterobacterales (ESBL-E) pose a major global public health challenge, as they are associated with frequent empirical antibiotic treatment failure and increased patient morbidity and mortality. In this context, contact precautions (CP) have traditionally been recommended for hospitalized patients colonized or infected with ESBL-E to prevent healthcare-associated transmission. However, growing evidence has questioned the effectiveness of this strategy in certain care settings. The primary objective of this study is to assess the non-inferiority of standard precautions (SP) compared with clinically triggered CP in reducing the incidence density of ESBL-E acquisition in acute-care geriatric wards. By doing so, this study aims to inform the evaluation of pragmatic infection prevention policies implemented at the institutional level in vulnerable hospital populations. Methods and analysis We will conduct a prospective, two-site interventional study in two geriatric departments of the Metz–Thionville Regional Hospital Center (Metz, France) using an alternating-period (crossover-like) design at the site level, with random allocation of the starting condition (CP first vs SP first). Patients will be screened for ESBL-E carriage at admission and at discharge. During CP periods, contact precautions are implemented only for patients with clinically detected ESBL-E infections identified through routine clinical cultures; asymptomatic ESBL-E carriage identified through surveillance screening does not trigger contact precautions. During SP periods, patients are managed using SP alone. The primary outcome is the incidence density of ESBL-E acquisition per 1,000 patient-days. Secondary objectives include species-specific analyses comparing ESBL-producing Escherichia coli and non-E. coli Enterobacterales. The total inclusion period is four years. Incidence rates will be compared using count regression models adapted to clustered period-based data.