Studies that evaluated the effects of CPOE with CDS on the development of an adverse event were eligible for inclusion.
The type of CPOE/CDS system was classified as home grown or commercially sold. The systems used in the included studies were mostly home grown; a small number of commercially sold systems were evaluated. CPOE systems were mostly used for all drug orders or for specific drugs (antibiotics or psychotropic drugs orders). Studies used basic or advanced clinical decision support. Basic support included drug-allergy checking, dose guidance, formulary decision support, duplicate therapy checking and drug-drug interaction checking; advanced support included adjusting for dose for renal insufficiency, guidance for medication-related laboratory testing, drug-disease contraindication checking and drug-pregnancy checking.
Studies were grouped into three categories: hospital care, ambulatory care and long-term care. Most studies were undertaken in a hospital setting and focused on older patients, paediatric patients or intensive care unit patients. There was a single study in an ambulatory setting (outpatient clinic).
The primary outcome of interest in the review was prevention of adverse drug events related to the prescribed medications. Adverse events were considered preventable if they were caused by an error or were classified as preventable using a standard approach. Data sources used to identify events in the studies included medical charts and records, and incident reports from staff. Where reported, a range of different methods were used across studies to identify events (such as computer identified, verified by nurse/pharmacist, independent review of data sources by two people and pharmacist monitoring and analysis of drug orders).
The authors did not state how many reviewers selected the studies.