Reviewing the impact of computerized provider order entry on clinical outcomes: the quality of systematic reviews

Source:
Medicines Management Collection
Publisher:
International Journal of Medical Informatics
Publication date:
27 March 2012

Abstract

Purpose: Computerised provider order entry (CPOE) is central to current efforts at improving clinical care.  Understanding the quality of the evidence for CPOE is important to the practical decision of implementation, patient safety and future design efforts.  This paper presents the results of a systematic analysis of the quality of systematic reviews of empirical CPOE research. Methods: The systematic search process included PubMed, CINAHL, Scopus, Cochrane, INSPEC, and PsychInfo databases from the years 1987 to mid 2010 in English only.  All reviews with a focus on CPOE, electronic ordering, Electronic Health Record, or Health Information Technology were included.  Studies were excluded if they did not mention a systematic review in the title or text, report a formal search process, report results of the search, or specifically include a separate section on CPOE in the results.  Quality was assessed using systematic criteria developed by Oxman and Guyatt, QUOROM, and PRISMA.  All three authors conducted the reviews independently.  Disagreements were resolved through discussion.  Descriptive data were extracted. Results: The search process yielded 185 initial unique references with 13 final reviews meeting the inclusion criteria.  The rating of overall quality in the Oxman and Guyatt scale averaged 4.9 out of a possible 7 and the average mean of the sum of the other questions was 5.69.  The overall QUOROM/PRISMA ratings averaged 63% completion and ranging from 45% to 81%. Conclusions: The quality of these reviews were moderate.  Only one study conducted a full quantitative synthesis, and overall heterogeneity was reported as very high in the 3 studies that measured it.  Recommendations emphasise clarifying the phenomenon of CPOE by avoiding reporting conclusions across sub-group analyses,  increasing emphasis on the development of theoretical models, including more quantitative assessments, and increasing breadth of outcomes.