Use of a Decision Support Tool on Prognosis of Work Ability in Work Disability Assessments: An Experimental Study Among Insurance Physicians

I Louwerse,J R Anema,A J Van Der Beek,C L I Gielen,M A Huysmans,H J Van Rijssen

doi:10.1007/s10926-020-09907-w

Abstract

Purpose Assessment of prognosis of work disability is a challenging task for occupational health professionals. An evidence-based decision support tool, based on a prediction model, could aid professionals in the decision-making process. This study aimed to evaluate the efficacy of such a tool on Dutch insurance physicians’ (IPs) prognosis of work ability and their prognostic confidence, and assess IPs’ attitudes towards use of the tool. Methods We conducted an experimental study including six case vignettes among 29 IPs. For each vignette, IPs first specified their own prognosis of future work ability and prognostic confidence. Next, IPs were informed about the outcome of the prediction model and asked whether this changed their initial prognosis and prognostic confidence. Finally, respondents reported their attitude towards use of the tool in real practice. Results The concordance between IPs’ prognosis and the outcome of the prediction model was low: IPs’ prognosis was more positive in 72 (41%) and more negative in 20 (11%) cases. Using the decision support tool, IPs changed their prognosis in only 13% of the cases. IPs prognostic confidence decreased when prognosis was discordant, and remained unchanged when it was concordant. Concerning attitudes towards use, the wish to know more about the tool was considered as the main barrier. Conclusion The efficacy of the tool on IPs’ prognosis of work ability and their prognostic confidence was low. Although the perceived barriers were overall limited, only a minority of the IPs indicated that they would be willing to use the tool in practice.

Highlights

Individuals who are unable to work due to a disease or disorder can apply for a work disability benefit
This was just above the minimum number of 28 insurance physicians’ (IPs) that was determined in a sample size calculation as number needed to answer the research questions
If we look at each of the case vignettes separately, we can see that for two of the vignettes that showed low initial agreement, there was an improvement in agreement after IPs were informed about the outcome of the prediction model

Summary

Introduction

Individuals who are unable to work due to a disease or disorder can apply for a work disability benefit. In most European countries, this covers both financial support to compensate loss of income, and interventions to support return to work [1]. A potential solution to this problem is to provide occupational health professionals with evidence‐based decision support tools. These tools are comprised of software and designed to aid decision-making [6]. They match characteristics of individual claimants with a computerized knowledge base to generate patient-specific assessments or recommendations [7]. Previous research has shown that such tools could be a means to achieve more accurate estimates of prognosis,

Objectives

Methods

Results

Conclusion