Related Experiment Video
Updated: Jun 10, 2026
Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
A Tool for Predicting Disability Pension Applications in the Finnish Public Sector: Sickness Absences as Early
1Faculty of Management and Business Studies, Tampere University, Tampere, Finland.
Purpose:
Identifying employees at the greatest risk of work disability is critical to preventing early exit from the labor market. Understanding how different factors influence work disability risk in the years leading to a pension is important in preventing early exits. Our aim in this study is to model the risk factors for an employee applying for a disability pension by using a novel and extensive register dataset and to evaluate the predicted risk for different occupations in the workforce.
Methods:
We use a logistic model to assess the risk of filing a first-time disability pension application using rich register data. We evaluate the performance of the model and finally present the evaluated risk distribution among current employees. Data are also included for short-term sickness absences, usually not available for register studies. A pension application is used as a follow-up to include employees whose application is rejected but who nonetheless carry a physician's note of diminished work ability, also a new approach to modeling work disability risk. The data cover all public sector occupations in Finland and their diverse disability risk profiles.
Results:
The constructed risk model accurately and transparently predicts the disability pension application risk in different employee groups. Age, occupation, cumulative sickness days, and number of sickness spells of all lengths, including short ones, were positively associated with the risk; proximity of retirement age and high earnings were negatively associated. Validation measures allow for using the model in providing employers with risk profiles of their employees.
Conclusion:
The developed model can be used as a tool for predicting the risk status of employee groups and for assessing the economic effects of sickness absences and disability pension risk. The modeling results reported here are already in practical use by public sector employers in Finland, eg, for early intervention.
