Predictive validity of general work ability assessments in the context of sickness insurance.

The data came from the first 300 AFU cases during a 2013 to 2014 testing period, so it may not reflect how the assessment is used now. The register data only showed whether people received sickness benefits, not whether they returned to work, so conclusions about return to work are speculative. People with lower educational demands in their jobs or who were self-employed had lower risk of remaining on benefits, but the study did not show they returned to work; they may have taken jobs with poor conditions or been supported by others. Missing scale scores were filled in using the person’s average score on other items in the same scale, which can affect the ratings. Educational demands were taken from job classification, not the person’s actual education, which may have affected the job-related findings. The study did not separate sickness pay from disability pension, because both required reduced work ability. The authors suggest self-rated work ability could partly act as a self-fulfilling prophecy, so prediction does not prove accuracy or causality.

The authors declared no conflicts of interest.

Do not treat a high self-rated work ability score as proof that a person will return to work. The study measured continued sickness benefits, not actual work. Also, 39% of people assessed as not having reduced work ability and having benefits withdrawn later had a new sick leave period.