The effectiveness of computerized drug-lab alerts: A systematic review and meta-analysis

Abstract Background Inadequate lab monitoring of drugs is a potential cause of ADEs (adverse drug events) which is remediable. Objectives To determine the effectiveness of computerized drug-lab alertsto improve medication-related outcomes. Data sources Citations from the Computerized Clinical Decision Support SystemSystematic Review (CCDSSR) and MMIT (Medications Management through Health Information Technology) databases, which had searched MEDLINE, EMBASE, CINAHL, Cochrane Database of Systematic Reviews, International Pharmaceutical Abstracts from 1974 to March 27, 2013. Study selection Randomized controlled trials (RCTs) of clinician-targeted computerized drug lab alertsconducted in any healthcare setting. Two reviewers performed full text review to determine study eligibility. Data abstraction A single reviewer abstracted data and evaluated validity of included studies using Cochrane handbook domains. Data synthesis Thirty-six studies met the inclusion criteria (25 single drug studies with 22,504 participants, 14 targeting anticoagulation; 11 multi-drug studies with 56,769 participants). ADEs were reported as an outcome in only four trials, all targeting anticoagulants. Computerized drug-lab alertsdid not reduce ADEs (OR 0.89, 95% CI 0.79–1.00, p =0.05), length of hospital stay (SMD 0.00, 95%CI −0.93 to 0.93, p =0.055, 1 study), likelihood of hypoglycemia (OR 1.29, 95% CI 0.31–5.37) or likelihood of bleeding, but were associated with increased likelihood of prescribing changes (OR 1.73, 95% CI 1.21–2.47) or lab monitoring (OR 1.47, 95% confidence interval 1.12–1.94) in accordance with the alert. Conclusions There is no evidence that computerized drug-lab alertsare associated with important clinical benefits, but there is evidence of improvement in selected clinical surrogate outcomes (time in therapeutic range for vitamin K antagonists), and changes in process outcomes (lab monitoring and prescribing decisions).