Comparing the effectiveness of universal admission testing and risk-based testing at emergency admission for preventing nosocomial COVID-19: a multicenter retrospective cohort study in Japan

Abstract

Objective: To compare the effectiveness of universal admission testing (UAT) and risk-based testing (RBT) in preventing nosocomial coronavirus disease 2019 (COVID-19) after the implementation of strict infection control measures.

Design: Retrospective multicenter cohort study.

Setting: Five community hospitals in Japan.

Patients: 14,028 adult patients admitted emergently from June 1, 2022, to January 31, 2023.

Methods: We calculated crude incidence density rates of community-acquired COVID-19 (positive test ≤4 days postadmission), hospital-acquired COVID-19 (positive test ≥8 days postadmission), total postadmission COVID-19 (all cases of positive test postadmission), and primary cases (sporadic and index cases). A generalized estimating equation model was used to adjust for local incidence (new COVID-19 patients per 100,000 population), single-bed room proportion, and admission proportion of patients older than 65 years.

Results: The weekly local incidence in the study areas was less than 1,800 per 100,000 population (1.8%). Two hospitals implemented RBT and 3 implemented UAT. The median admission testing rate was higher in the UAT group than in the RBT group (95% vs 55%; difference 45.2%, 95% CI, 40.3%-48.8%). Crude and adjusted analyses revealed no significant associations between incidence density rates (IRR; >1 indicates higher incidence with UAT) and admission strategies for any of the outcomes: community-acquired cases (adjusted IRR = 1.23; 95% CI, 0.46-3.31), hospital-acquired cases (1.46; 0.80-2.66), total postadmission COVID-19 (1.22; 0.79-1.87), and primary cases (0.81; 0.59-1.12).

Conclusions: Compared with risk-based testing, universal admission testing may have limited additional benefits in preventing nosocomial COVID-19 transmission during a period of low-moderate local incidence.

Figure 1.

Flow diagram of patient enrollment in the study.

Figure 2.

Crude weekly mean incidence density rate (solid line), admission testing rate (dashed line), and local incidence rate in Hyogo Prefecture (bar graph). Universal admission testing group (filled circle, blue) and risk-based testing group (filled triangle, orange). The vertical axis shows (left-most) the incidence density rate per 1,000 at-risk patient days, (left-most) Hyogo Prefecture’s incidence rate (%), and (right-most) the testing rate (%).