How common is change of primary diagnosis during an episode of sickness benefit? A register study of medical sickness certificates issued 2010-2012 in Sweden.

AIM: The aims of this study were to investigate how common it is to change primary diagnosis between different diagnostic chapters during a sick-leave spell, and to explore patterns of diagnostic changes. METHODS: The unit for analysis was episode of sickness benefit, that is, sick leave >14 days, which commenced between 2010 and 2012 in Sweden. For each case, the primary diagnosis was retrieved from the first and last/latest medical sickness certificate, respectively. The number of days of sickness benefit was linked to the cases. Any change of primary diagnosis was analysed by diagnostic chapter according to the ICD-10, and this was done separately for women and men. RESULTS: In total, 803,041 cases of sickness benefit (63% women) were included in the study. During a sick-leave spell, 7.1% of female cases and 6.6% of male cases changed their primary diagnosis to a diagnosis from another diagnostic chapter. The change of primary diagnosis increased with the number of days with sickness benefit. For female cases, this increase was from 2.0% for cases that lasted 15-30 days to 20.2% for cases that lasted >365 days. For male cases, the corresponding increase was from 1.8% to 21.2%. A change of primary diagnosis was least common among those initially sick-listed for mental disorders and musculoskeletal disorders. The patterns of diagnostic changes were rather similar for women and men. CONCLUSIONS: A change of diagnosis during a sick-leave spell needs to be taken into consideration by the sickness insurance system and in the actions taken by its administration. Registry-based studies of sickness insurance need to consider diagnostic changes in both the study design and the interpretation of results.