Table 2.

Type of CDS Alerts and Data Sources and Mechanisms Used to Identify the ADEs

Source	Type of CDS alerts	Data sources used to identify ADEs	Mechanisms used to identify and classify ADEs
Evans et al.28 1994	Alerts pharmacists when a drug was inadvertently ordered to which a patient was allergic	Nurse charts; nurse/pharmacist reports; laboratory results; pharmacy orders	Computer-identified ADE was verified by nurse or clinical pharmacist based on Naranjo method39 (definite, probable, possible/unlikely)
Bates et al.23 1998	Menu of medications; default doses and range of doses; relevant lab results; consequent orders; limited drug-allergy checking, drug–drug interaction, and drug-laboratory checking	Incident reports from nurse/pharmacist; information from unit staff; medical records review	Review of data sources for possible drug-related incidents; independent review by 2 reviewers; structured criteria used to determine occurrence, severity, and ADE preventability15
Bates et al.24 1999	Hospital-approved standard lists; relevant laboratory results; detect drug-allergy interactions for the most commonly allergic drug families; duplicate order warning; life-threatening drug–drug interactions and drug-laboratory checks notification	Pharmacist reporting; nurse reports; pharmacist medication sheets review; medical records review	Independent review by 2 reviewers; structured criteria used to determine occurrence, severity, and ADE preventability15
Peterson et al.30 2005	Highlighted default dose and frequency; suggested substitution for psychotropic medications based on default dose and frequency for elderly patients	Electronic medical records; hospital inpatient reporting system; altered mental status score	Analysis of administrative data
Upperman et al.32 2005	Most drugs cross-referenced in an online formulary; rules warning of unfavorable clinical parameters in patients status; potential drug–drug, drug–allergy, and drug–food interactions and potential medication errors alerts	ADE rate data pre and post introduction of CPOE implementation	Med-Marx32 categorization system (no harm or harm ADE) used to identify errors in prescribing
Evans et al.26 1995	Maximum 24-h white blood cell count and temperature; calculates renal function and estimated creatinine clearance; antibiotic allergies and current antibiotic therapy; uses all patient admission and patient allergies, drug–drug interactions, toxicity, and cost in selection and type of antibiotic therapy; calculates dosage and frequency	Computerized medical records; prospective ADE surveillance	Not specified
Evans et al.27 1998	Decision-support logic suggests an antiinfective regimen; uses patient allergies, drug–drug interactions, toxicity, contraindications, and cost in the selection of antibiotics; measures renal and hepatic function to calculate dose and dosing interval	Prospective ADE surveillance	Not specified
Colpaert et al.25 2006 (RCT)	Commonly used drug therapy with dose and dosage schemes for renal insufficiency and for patients with severe liver dysfunction; allergies, clinically important interactions, and drug-related complications; protocol-based facilitated medication prescription for specific patient groups	Medication orders; medical and nursing files; laboratory data	Clinical pharmacist analysis of every medication order for possible error; independent panel evaluation of the severity of errors; used NCC MERP guidelines for classifying errors
Mullett et al.29 2001	Laboratory results review and summary report; the pediatric antiinfective management program	Computer alerting program; pharmacist-recorded ADE	Pharmacy staff monitoring and recording
Steele et al.31 2005	Rule based drug–laboratory interactions alerts; can order any rule-associated lab test	Automated order entry forms; random sample of chart reviews	Random sample of chart reviews using the Naranjo scoring scale39

CPOE: computerized physician order entry, CDS: clinical decision support, ADE: adverse drug event, NCC MERP: National Coordinating Council for Medication Error Reporting and Prevention