101 Residual work capacity and (in)ability to work fulltime among applicants with cancer with the findings by Stein describing a higher risk on physical impairments for bone cancer (a cancer of the locomotor system) [34]. Our findings showed that being diagnosed with an additional disease other than cancer lowered the odds for no residual work capacity. This seems counter intuitive, because one should expect that when diagnosed with more than one disease would have larger impact on work ability. In order to find an explanation, we discussed these results with insurance physicians. They thought a possible explanation might be that when they assess applicants with such a severe disease, of which the impact on work capacity is so obvious they feel that further exploration of the medical situation is unnecessary. In these cases, they usually do not register any additional diagnosis. The dataset used for our study only included data registered by the insurance physicians at the time of the assessment, data supporting this possible explanation of the insurance physicians cannot be verified by our dataset. However, we did not see this association of multimorbidity for the specific cancer groups with inability to work fulltime; in none of the specific cancer groups, an association of multimorbidity was found with inability to work fulltime. This might indicate that the cancer diagnosis itself already has such a major impact on work capacity, that an additional diagnosis does not increase the risk for being assessed with an inability to work fulltime. Strengths and limitations In this study, we used register data of a year cohort of applicants assessed for a long-term work disability benefit after (in most cases) 2 years of sick leave. Using register data is a strength of our study, as it covers the entire Dutch population. Another strength of our study is the large sample size of work disability benefit assessments by skilled insurance physicians adhering to professional guidelines and assessment methods. Furthermore, our sample seems representative, as the prevalence of the specific cancer groups in our sample is in line with prevalence nation- and worldwide [36]. The prevalence of cancers of the male genital organs in our sample was, however, lower than in the society. This might be due to the fact that this type of cancer is specifically higher prevalent among older males, who are not part of the working population anymore. A study limitation is that register data was not collected for research purposes and did not contain data on other possible determinants such as severity of diseases, time from cancer diagnosis, stage of cancer, the treatment received, and physical and psychosocial work demands. Although the UWV uses a biopsychosocial approach in the work disability assessment, important factors described in this model, as mentioned above, are lacking in the register data. The absence of these determinants can also affect the generalizability of the findings. Furthermore, for the analysis on inability to work fulltime, we had to exclude 254 cases due to missing data on 5
RkJQdWJsaXNoZXIy MTk4NDMw