The use and effectiveness of electronic clinical decision support tools in the ambulatory/primary care setting: a systematic review of the literature

Abstract

The 1999 Institute of Medicine (IOM) report To Err is Human alerted the healthcare industry and the public to the lack of consistency in the delivery of quality care to the US population. Clinical decision support systems (CDSS) have become a leading response to this report, and to the growing demand for the promotion of standards-based care delivery. The objective of this paper is to evaluate the recent literature for both the types and effectiveness of electronic CDSS in the primary care setting. An electronic search of the literature was conducted utilising MEDLINE (1996-2006), CINAHL (1982-2006) and all EBM Reviews--Cochrane DSR, ACP Journal Club, DARE and CCTR. The search included various combinations of the MeSH search terms 'clinical decision support systems', 'primary health care', 'ambulatory care' and 'practice guidelines' and was limited to articles published from 2000 to 2006. Studies were selected for review if they involved either non-randomised observational or randomised controlled trials (RCTs) utilising CDSS as a single intervention, were performed in an ambulatory primary care setting and included quantifiable outcome measures. Seventeen studies were included in the review, including five non-randomised observational studies and 12 RCTs. Thirteen studies (76%) found either positive or variable outcomes related to CDSS intervention with four studies (24%) showing no significant effect. Although there is validation that CDSS has the potential to produce statistically significant improvement in outcomes, there is much variability among the types and methods of CDSS implementation and resulting effectiveness. As CDSS will likely continue to be at the forefront of the march toward effective standards-based care, more work needs to be done to determine effective implementation strategies for the use of CDSS across multiple settings and patient populations.