Interrupted times series (ITS) evaluation designs involve monitoring particular populations, administrative units, or groups of health professionals for a period of time prior to implementation of an intervention (e.g., patient safety programme implementation; national guidelines dissemination, prescribing policy reorganisation in hospitals), and subsequently for a period of time following the intervention (see figure 1 below). The general objective in ITS studies is to examine whether the data series observed post-intervention differs in important aspects from that in the pre-intervention period. Robust evaluation methodology is vital, but analysis of such designs varies from study to study and there is little published guidance on the design strengths and weaknesses, or contexts in which the design should be considered exist. These uncertainties are hindering the quality and reporting of results as well as the spread of many improvement interventions where randomisation is not feasible.
Broad research questions
- What design and analysis approaches to ITS data are being used in healthcare and how are they reported?
- What methods should be used when analysing an ITS and what aspects should be reported?
Supervision: Professor Craig Ramsay and Dr Shona Fielding
- Jemma Hudson; firstname.lastname@example.org
StatusOngoing - Data collection
M. Taljaard, J. E. McKenzie, C. R. Ramsay, and J. M. Grimshaw. The use of segmented regression in analysing interrupted time series studies: an example in pre-hospital ambulance care. Implementation Science, 9(1):77, 2014.
Database of interrupted time series in the healthcare setting
A database of characteristics of ITS studies from healthcare settings, including frequency of data collection, autocorrelation, seasonality and effect sizes: Database of ITS studies in healthcare.xlsx