Development and validation of a risk prediction model for work disability: multicohort study.

Area of Science:

• Occupational Health
• Epidemiology
• Biostatistics

Background:

• Work disability significantly impacts individuals and society.
• Existing risk prediction models often lack multifactorial approaches.
• Identifying at-risk individuals is crucial for preventative strategies.

Purpose of the Study:

• To develop and validate a parsimonious multifactorial prediction score for work disability.
• To assess the discriminative ability of the developed score.

Main Methods:

• Individual-level data from a public-sector employee cohort (n=65,775) and a general population sample (n=13,527) were used.
• A comprehensive set of sociodemographic, health, lifestyle, and work-related variables were analyzed.
• A parsimonious 8-predictor model was derived and validated.

Main Results:

• The final prediction score included age, self-rated health, sickness absences, socioeconomic position, chronic illnesses, sleep problems, BMI, and smoking.
• The score demonstrated high discriminative ability with C-indices of 0.84 (development) and 0.83 (validation).
• A work-related predictor score showed slightly lower discrimination (C-indices 0.79 and 0.78).

Conclusions:

• A rapidly administered, 8-item prediction score reliably identifies individuals at high risk of future work disability.
• This tool can aid in targeted interventions to mitigate work disability.
• The findings support the use of multifactorial algorithms in occupational health risk assessment.