Chapter 11.6: Clinical Decision Support Systems

CLINICAL SCENARIO

On the ambulatory care clinic rotation rounds in which you are the senior attending physician, a senior resident comments that nearly half of the patients seen today have diabetes and many of these patients have complications of diabetes, including previous myocardial infarction, stroke, neuropathy, and nephropathy. Some of the patients are only in their third or fourth decade of life. This leads to a discussion of chronic disease and its effect on disability, quality of life, and mortality, as well as evidence for treatments that improve these outcomes. One of the medical students, who previously worked as a software developer, is convinced that computerized decision support for physicians and patients has to be the way to improve diabetes care. A junior resident from a psychology background points out that a complex chronic disease such as diabetes requires not only awareness of interventions that can decrease mortality and morbidity but also considerable patient investment of learning, time, and expense to address all of the components. You remind the group that diabetes management, with its multiple guideline recommendations, could benefit from clinical decision support systems but that one should ask whether the systems improve patient outcomes. The senior resident, intrigued by the discussion, commits to presenting a well-done randomized clinical trial (RCT) that addresses the effect of computerized decision support systems on diabetes management at the upcoming morning report for all house staff and attending physicians. You tell her that there have been so many studies in this area that it might be more efficient to look for a well-done systematic review.

After the final afternoon rounds, you take a few minutes to search the literature. You believe that, given the progress in technology over the years, it is likely that only recent studies will be applicable to current clinical care. Computers on the ward allow you access to PubMed, so you quickly enter “computerized decision support AND diabetes” in the Clinical Queries search field (http://www.ncbi.nlm.nih.gov/pubmed/clinical; see also Chapter 5, Finding Current Best Evidence). The search retrieves 16 publications, including 4 systematic reviews that address your question. After screening the abstracts, the second, a systematic review and meta-analysis, stands out as directly relevant¹ (see Chapter 22, The Process of a Systematic Review and Meta-analysis).

In this systematic review, the authors searched for all randomized trials based in ambulatory care that involve patients with diabetes and compare a computerized clinical decision support system with usual care with or without additional educational materials. Outcomes of the trials included measures of process of care or patient-important outcomes.