Evaluating quick return restrictions on sickness absence in healthcare employees: A difference-in-differences study.

Abstract

Background: Short intervals between shifts, known as quick returns, have been linked to adverse health effects, and increased risk of occupational accidents, particularly among healthcare employees. To safeguard employee health, the 2020 reform of Working Time Act in Finland limited rest periods under 11 h in irregular shift work.

Objective: To evaluate the changes in quick returns following the 2020 reform of the Working Time Act in Finland and their association with sickness absence among public healthcare employees.

Design and methods: This observational longitudinal study, analysed as a quasi-experiment used a difference-in-differences regression analysis with unit and time fixed effects and robust standard errors. We assessed changes in sickness absence from 2019 to 2021 across hospital work units. The study compared units mandated to limit quick returns (Treatment group; 416 units, over 20,500 employees, 72 % in nursing) with units that had low levels of quick returns prior the reform and did not need to limit quick returns (Control group; 37 units, over 1700 employees, 70 % in nursing). The analysis considered local agreements permitting quick returns, using both intention-to-treat and per-protocol approaches. Sensitivity analysis included regression models with unit level covariates and inverse probability weighting to adjust for initial differences.

Results: The per-protocol approach and simple regression analysis with fixed effects for unit and time over 2019-2020 showed a less pronounced increase in sickness absence by -0.7 percentage points (95 % confidence interval [CI]: -1.3 to -0.1) for the treatment group compared to the control group, indicating a 13 % lower rate of sickness absence. For the period 2019-2021, the estimate was -0.5 percentage points (95 % CI: -1.0 to 0.0). When incorporating covariates and inverse probability weighting, the estimates were more substantial with narrower confidence intervals: -0.9 percentage points (95 % CI: -1.4 to -0.3) for 2019-2020, and - 0.6 percentage points (95 % CI: -1.2 to -0.1) for 2019-2021. The estimates from the intention-to-treat approach were consistent with the per-protocol results.

Conclusion: The Working Time Act reform reduced quick returns, and after the reform, the reduction was associated with a smaller increase in sickness absence among healthcare employees. Policymakers and nursing managers should evaluate and adjust the frequency of quick returns to achieve the potential effects on employee well-being, health, and operational efficiency, which in this study was indicated by the reduction in sickness absence.