gms | German Medical Science

Background and status of (inter)national research: Ensuring hospitals’ intensive care capacity for the management of substantial numbers of COVID-19 patients is a central goal during the global SARS-CoV-2 pandemic. In order to support decision-making (e.g. patient transferals and capacity management for elective surgical interventions) in the context of hospital burden various risk assessments and models were developed. The knowledge about the predictive power of risk assessments for the development of actual hospital burden can provide an important foundation for the selection of appropriate local hospital measures. However, evaluations on the ability of such assessments to predict short-term trends in local hospital bed occupancy are scarce.

Research question and objective: The aim of this analysis is to evaluate the extent to which the risk assessments made within a state-wide COVID-19 reporting corresponded to parameters reported in the intensive care register of the German interdisciplinary association for intensive care and emergency medicine (DIVI).

Method or hypothesis: In Bavaria, a continuous data-based COVID-19 reporting was established and distributed among stakeholders within locals crisis management structures (integrated command centers (ILS)) during the first three waves. The COVID-19 reporting provided a risk assessment as well as recommendations for local decision-making based on selected indicators related to the number of inpatients with COVID-19 and officially reported SARS-CoV-2 cases. The aggregation of these indicators resulted in a “traffic-light” risk assessment on the ILS-level, intended to provide directive information on short-term hospital occupancy trends. To evaluate the informational value of these indicators for the short-term development of actual hospital burden, we investigated the predictive capacity of the “traffic-light” risk assessments for diverse criteria indicative of hospital burden as reported in the DIVI intensive care register and compared the results with other models to predict the hospital burden.

Discussion: Our experiences during the course of the pandemic show that an early implementation of a data-based COVID-19 reporting provides helpful support for the capacity management within local crisis structures. This analysis will provide further insight whether the risk assessment based on selected indicators was able to reflect the situation in the hospitals as reported in the DIVI intensive care register.

Practical implications: The analysis can give insights whether the aggregation of the selected indicators is a valid tool for assessing and predicting short-term hospital burden and can thus serve as a basis for (local) decision-making.

Appeal for practice (science and/or care) in one sentence: Within local crisis management structures, decision-making on patient transferals and capacity management for elective and non-elective patients can benefit from data-based risk assessments.