Measuring Health System Resilience During the COVID‐19 Pandemic Using Dynamic Indicators of Resilience Based on Sick‐Leave Data

TL;DR

This study uses sick-leave data to measure how resilient the Dutch healthcare system was during the COVID-19 pandemic.

Contribution

The study introduces dynamic indicators of resilience (DIORs) estimated from sick-leave data to assess healthcare system resilience during disruptive events.

Findings

Sick-leave absenteeism rates increased significantly post-pandemic.

DIORs showed increased autocorrelation during the pandemic, indicating reduced resilience.

Resilience trends varied across regions but were consistent across healthcare sectors.

Abstract

Healthcare system resilience is generally understood as the capacity of a healthcare system to prepare, withstand, and adapt to disruptive health events while maintaining the continuity and quality of essential health services. So‐called dynamic indicators of resilience (DIORs) allow us to examine resilience by analysing patterns of functioning of the healthcare system in time series data. The aim of this study was to examine whether DIORs can be estimated from time series data of the functioning of the Dutch healthcare system before, during and after the COVID‐19 pandemic, and whether these DIORs are indicative of the resilience of the Dutch healthcare system during the COVID‐19 pandemic.

To select a measure of healthcare functioning, healthcare experts completed a questionnaire in which they selected the five most relevant indicators of healthcare availability (table s14). Based on the questionnaire results and datasets available, time series data of sick‐leave absenteeism rates among Dutch healthcare workers before, during and after the COVID‐19 pandemic were used to quantify the functioning of the Dutch healthcare system. DIORs were estimated using moving window techniques on the time series data of each healthcare sector, each safety region in the Netherlands, and all healthcare sectors and safety regions in the Netherlands combined.

Short‐term sick‐leave increased from 3.2% to 4.5% and long‐term from 3.0% to 4.0% post‐pandemic (p < 0.001). DIORs showed significantly increasing autocorrelation during the pandemic (Kendall's τ = 0.46–0.52), indicated an increased loss of resilience of the Dutch healthcare system as the COVID‐19 pandemic progressed. Trends were consistent across healthcare sectors but varied across regions, with some regions showing stable or improving resilience.

Our results indicate that DIORs, estimated from time series data of sick‐leave absenteeism rates among healthcare workers in the Netherlands during the COVID‐19 pandemic, potentially provide useful insights into healthcare system's resilience during and following disruptive health events, such as the COVID‐19 pandemic.

Figures