An Overview of Interrupted Time Series as a Research Method

Interrupted time series (ITS) study design is increasingly used to evaluate the impacts of guidelines, policy changes, and quality improvement efforts on clinical outcomes. Repeated observations of an outcome of interest over time are “interrupted” by an intervention, dividing the study into pre- and postintervention time periods. Regression analysis then compares the level change and slope of the 2 time periods to determine intervention effects. The benefits of this study design include (1) use of lower-cost, preexisting large data sets, (2) reduced susceptibility to confounding by individual-level characteristics with the ability to test for potential biases, and (3) presentation of results in straightforward, easy-to-interpret figures. However, ITS methods require high-quality, consistent data over an extended time frame, and researchers must have a firm grasp of potential confounding variables and selection of appropriate statistical analysis methods. When properly employed, ITS provides a rigorous study design to evaluate outcomes in pediatric populations that may otherwise be difficult to investigate. This primer familiarizes pediatric hospitalists with the basics of employing this study methodology, introduces examples of its utility in the field, and directs readers to existing literature for the more advanced statistical underpinnings of conducting such a study.