) where the data has been incorrectly analyzed (eg, when there has been no adjustment for autocorrelation-a common complication with time series data A re‐analysis may be required in the circumstance where effect estimates have been incompletely reported (ie, when an effect estimate is reported without a measure of precision For these reasons, visual displays of ITS data in both primary studies, and systematic reviews of ITS studies, are a valuable part of reporting.Ī further benefit of visually displaying data from an ITS study is that it allows systematic reviewers to extract the data (eg, using digitizing software) and undertake a re‐analysis. These features can be visually displayed, and in well‐designed graphs, the impacts of the interruption on the outcome will likely be evident. Statistical comparisons between the counterfactual and observed data at different points post interruption can be used to estimate the short‐ and long‐term effects of the interruption. By modeling data from the pre‐interruption period, the underlying secular trend can be established and extrapolated to the post‐interruption period, creating a counterfactual for what would have occurred in the absence of the interruption. In an ITS study, data on a group of individuals (eg, hospital, country) are collected at multiple time points both before and after the interruption. The ITS design inherently lends itself to a visual display. However, effective and accurate presentation of the data from ITS studies is needed to enable their inclusion in systematic reviews (including meta‐analysis) and to aid interpretation of the results from the review.
This makes ITS studies a valuable design for inclusion in systematic reviews intended to inform policy decisions.
The interruption could be planned, such as the roll out of a new health policy, or unplanned, such as an unintended environmental exposure.
Interrupted time series (ITS) studies are a common design used in areas such as public health, health policy and health services research to examine the effects of an interruption on an outcome. Our proposed recommendations aim to achieve greater standardization and improvement in the display of ITS data, and facilitate re-use of the data in systematic reviews and meta-analyses. Accurate data extraction (requiring distinct points that align with axis tick marks and labels that allow the points to be interpreted) was possible in only 72 (33%) graphs.ConclusionWe found that many ITS graphs did not meet our recommendations and could be improved with simple changes. We assessed whether 217 graphs from recently published (2013-2017) ITS studies met our recommendations and found that 130 graphs (60%) had clearly distinct data points, 100 (46%) had trend lines, and 161 (74%) had a clearly defined interruption. The adapted recommendations cover plotting of data points, trend lines, interruptions, additional lines and general graph components. Further, well-constructed graphs allow data to be extracted using digitizing software, which can facilitate their inclusion in systematic reviews and meta-analyses.AimWe provide recommendations for graphing ITS data, examine the properties of plots presented in ITS studies, and provide examples employing our recommendations.Methods and resultsGraphing recommendations from seminal data visualization resources were adapted for use with ITS studies. The data from such studies are particularly amenable to visual display and, when clearly depicted, can readily show the short- and long-term impact of an interruption. IntroductionInterrupted Time Series (ITS) studies may be used to assess the impact of an interruption, such as an intervention or exposure.