Assessment of Pharmacy Interventions Articles with Interrupted Time Series Design in Healthcare Journals: A Systematic Review

Abstract

Background and Objective: Research evidence might be invalid because of biases in methodology. We conducted a systematic review and aimed to assess methodological quality and appropriate analysis of pharmacy intervention studies using Interrupted Time Series design (ITS).

Methods: We included pharmacy intervention with ITS studies published in healthcare journals during January 2007 – December 2011. They were searched from the databases  of  PubMed, SCOPUS, CENTRAL and Web of Science. Hand searching was done in Thai healthcare journals. Two reviewers independently selected studies and assessed risk of bias using the pre-established form. The methodology quality was classified in to good, not good and unclear according to information of essential components of ITS appeared in the studies.

Result: This review included 36 eligible studies. From these, 50.0 % (95% CI: 32.9 to 67.0) had good quality of methodology, 47.2% (95% CI: 30.4 to 64.5) were unclear. There were 58.3% (95% CI: 54.8 to 85.8) used appropriate analysis. Of the 36 included studies,15 (41.7%, 95% CI: 25.5 to 59.2) showed high quality in both methodology and analysis.

Conclusion: The results suggest that published pharmacy interventions studies with ITS designs should be improved in the reporting of methodology and analysis.

Keywords: Interrupted Time Series, Pharmacy Interventions, Systematic Review