Derivation and validation of a prognostic model for workers disabled by depression

Abstract

AimMajor Depressive Disorder is associated with decreased productivity, and is a common cause of disability. We used administrative data from a private Canadian disability insurer, SSQ Life Insurance Company Inc. (SSQ), to develop a prognostic model for disability resolution.MethodWe used parametric survival models (Royston-Parmar models) that allow more flexibility for the shape of the survival distribution. We employed a random split technique with a ratio of 70:30 for creating a derivation and validation set.ResultsOur study cohort was comprised of 3,113 long term disability (LTD) claimants with a primary diagnosis of Major Depressive Dis-order, of which 2,679 cases (87%) were closed when our data was captured and 434 (13%) were censored. The median time on LTD benefits was sig-nificantly shorter among disabled workers residing in Quebec (approximately 200 days) than in Ontario (ap proximately 550 days, p<0.001). Our final fitted model demonstrated that older age, female gender, shorter claim approval time and residing in Ontario vs. Quebec were associated with slower LTD claim closure among workers disabled due to depression. We produced a receiver operating characteristic curves for 1-year follow-up period for the training and validation datasets and area under the curve estimates were 0.70 and 0.69, respectively.ConclusionOur prognostic model has modest but acceptable discrimination and good calibration. We confirmed that work­ers receiving LTD benefits for depression who reside in Quebec, Canada are more likely to resolve their insurance claims quickly than those who reside in Ontario, Canada.

Keywords

Additional Information

Publisher’s Note
Bayrakol MP remains neutral with regard to jurisdictional and institutional claims.

Rights and Permissions

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0). To view a copy of the license, visit https://creativecommons.org/licenses/by-nc/4.0/

View full license details →