A Systematic Comparison of Age, Comorbidity and Frailty of Two Defined ICU Populations in the German Helios Hospital Group from 2016–2021

TL;DR

This study compares age, comorbidities, and frailty in ICU patients across German hospitals from 2016–2021, finding subtle shifts in patient demographics and severity.

Contribution

The study introduces two distinct ICU case definitions and provides a longitudinal analysis of age and severity trends in ICU populations.

Findings

The mean age of CodeBased ICU cases slightly decreased, while BedBased ICU cases slightly increased over six years.

CodeBased ICU patients had higher comorbidity and frailty scores compared to BedBased ICU patients in 2021.

The proportion of ICU admissions for patients aged 60–69 and ≥80 years showed a marginal increase over time.

Abstract

Background/Objectives: The demographic change raises concerns about the provision of adequate, long-term healthcare. Our study was driven by the decision to test other studies’ findings about how patients’ age and comorbidities are significantly increasing in German intensive care units (ICUs) over time. The goal of this study was to analyze the age and age-related characteristics, e.g., comorbidities and frailty, in ICU populations from 86 hospitals in the German Helios Group over a period of 6 years. Methods: For this retrospective observational study, we derived two different definitions of ICU cases, with (i) CodeBased ICU cases being defined by typical ICU procedures (e.g., OPS 8-980, 8-98f and/or duration of ventilation > 0 h) derived from the German administrative dataset of claims data according to the German Hospital Remuneration Act and (ii) BedBased ICU cases being based on the actual presence of a patient on a designated ICU bed; this was taken from the Helios hospital bed classification system. For each ICU definition, the size of the respective ICU population, age, Elixhauser Comorbidity Index (ECI) and Hospital Frailty Risk Score (HFR) were analyzed. Further patient characteristics, treatments and outcomes are reported. Trends in cases with and without COVID-19 were analyzed separately. Results: We analyzed a total of 6,204,093 hospital cases, of which 281,537 met the criteria for the CodeBased ICU definition and 457,717 for the BedBased ICU definition. A key finding of our study is that a change in age in absolute and relative terms is observable and statistically significant: the mean age of CodeBased ICU cases, 68.7 (14.4/−0.06), is marginally decreasing, and that of BedBased ICU cases, 69.1 (15.9/0.07) (both with a p-value of <0.001), is marginally increasing. Age analysis excluding COVID-19 cases does not change this key finding. A longitudinal analysis shows a continuously decreasing number of ICU admissions and a marginally positive trend of patients who are 60–69 and ≥80 years old: CodeBased ICU, 1.04/1.02; BedBased ICU, 1.03/1.03, all with a p-value of <0.001. A severity analysis based on the ECI and HFS shows that both are higher in CodeBased ICU cases (2021 ECI:18.0 (12.9); HFS: 10.7 (7.3); both p-values < 0.001) than in BedBased ICU cases (2021 ECI: 12.3 (12.4); HFS: 7.4 (7.1); p-values of 0.3 and 0.12). Further testing results per definition are reported. Conclusions: The observed age-related trends suggest that there has been a further increase in demand for intensive care from a frailer population. We recommend further studies to critically evaluate the increasing frailty within the ICU population and to test the associated presumed need for increased ICU capacities.

Figures