Background: Successful return to work is regarded as one of the most important outcome factors for working-age post stroke patients. The present study will estimate the effect of various predictors on the odds of returning to work after stroke. Nearly twenty thousand 20-57 year-old stroke patients in Denmark who were gainfully occupied prior to the stroke will be included in the study.
Methods/design: Stroke patients will be followed prospectively through national registers. Multi-level logistic regression will be used to model the odds of being gainfully occupied ca. two years after the stroke as a function of the following predictors: Age (20-49 years, 50-57 years) gender, occupational class, self-employment (yes; no), onset calendar year (1996, 1997, ..., 2006), diagnosis (subarachnoid haemorrhage; intracerebral haemorrhage; cerebral infarction; stroke, not specified as haemorrhage or infarction) and 'type of municipality' (the variable is set to 1 if the person lived in a municipality which had a brain injury rehabilitation centre at the time of the stroke. Otherwise it is set to 0). Municipalities will be treated as the subjects while individual observations within municipalities are treated as correlated repeated measurements.
Discussion: Since our follow-up is done through registers and all people in the target population are included, the study is free from sampling bias, recall bias and non-response bias. The study is also strengthened by its size. The major weakness of the study is that it does not contain any stroke severity measures. Thus, it cannot accurately predict whether a particular stroke patient will in fact return to work. The study is, however, quite useful from a public health perspective. It can be used to estimate the proportion of patients in a certain group that is expected to return to work, and thereby provide a comparison material, which e.g. municipalities can use to evaluate their success in returning their stroke patients to work.