longitudinal, time series, interrupted, research, health service
Changes to health services require monitoring and analysis to determine if the desired improvements are
achieved. Longitudinal studies, such as interrupted time series, involve the collection and analysis of data
to explore the outcome of interest over time. Interrupted time-series design provides an estimate of the
associated effect of an intervention using longitudinal data. Health services research increasingly uses these
studies to inform policy and practice development. In many cases, the data which allow this investigation have
been collected over many days, weeks, months, or even years. Analysis and interpretation need to allow for
any underlying trends and the possible presence of seasonality. This analysis aims to identify changes in
longitudinal health service data and the times at which they occur. We outline two case studies, one without
seasonality (Case Study 1) and one with (Case Study 2). We present an overview of the challenges of
longitudinal studies and identify solutions for future researchers.