Author; years searched | Number and design of studies | Population | Intervention | Control group | Outcomes | Summary of findings |
Eslami et al; 1966–200641 | 8 RCT 2 non-RCT 1 interrupted time series 35 pre/post intervention 19 observational 1 usability 1 mixed method | Newborn Paediatric Adult | CPOE CPOE with unknown CDSS | Unspecified | Medication errors Adverse drug events (ADEs) Length of stay Mortality Adherence to guidelines Costs of implementation and organisational efficiency Ordering and turnaround time Alerts, appropriateness of alerts, accepted and ignored alerts Compliance with alerts/order sets Usability satisfaction | CPOE systems improved adherence to guidelines, improved appropriateness of alerts, decreased costs, improved organisational efficiency and usability and provider satisfaction. Some studies suggested that errors, ADEs and even mortality increased after CPOE implementation. |
Wolfstadt et al; 1966-March 200740 | 1 controlled cross sectional 1 interrupted time series 8 pre/post intervention | Paediatric Adult | CPOE with basic CDSS CPOE with advanced CDSS | Usual care | ADEs | CPOE with CDS resulted in statistically significant decrease in ADEs (5 studies); non-statistically significant decrease (4 studies) and no change (1 study). |
Ammenwerth et al; 1966–200623 | 2 RCT 1 cross-sectional 1 retrospective review 5 prospective study 7 pre/post 5 interrupted time series 4 cohort 2 observational studies | Paediatric Adult | CPOE CPOE with basic CDSS CPOE with advanced CDSS | Handwritten prescription CPOE CPOE with basic CDSS | Medication errors ADEs | 23 studies showed a relative risk reduction ranging from 13% to 99%; 6 studies showed a relative risk reduction in potential ADEs between 35% and 98%. Studies differed substantially in their setting, design, quality and results. |
van Rosse et al; Start–200724 | 1 controlled cross sectional 8 retrospective 3 prospective | Newborn Paediatric Adult | CPOE CPOE with basic CDSS CPOE with advanced CDSS CPOE with unknown CDSS | Handwritten prescription | Medication errors ADEs Mortality Timeliness of service/implementation process outcomes | CPOE systems reduced the relative risk of medication errors (24%–99%) but had no effect on mortality or ADEs. |
Reckmann et al; 1950–200739 | 1 cross-sectional study 2 interrupted time series 7 pre/post intervention 1 cross over 1 comparative cohort | Newborn Paediatric Adult | CPOE with basic CDSS CPOE with advanced CDSS | Unspecified | Medication errors Severity of post-CPOE errors (ie, harm) New types of post-CPOE errors | 9 studies showed a significant decrease in the medication error rates. |
Nuckols et al; January 2007–September 201338 | 13 pre/post intervention 1 difference in differences design 2 comparison of similar units | Adult | CPOE CPOE with basic CDSS CPOE with advanced CPSS CPOE with unknown CDSS | Handwritten prescription | Medication errors Preventable ADEs | Implementing CPOE was associated with a greater than 50% decline in preventable ADE and medication error rates. |
Prgomet et al; January 2000–January 201618 | 19 pre/post intervention 1 prospective study | Paediatric Adult | CPOE CPOE with basic CDSS CPOE with advanced CDSS | Handwritten prescription CPOE | Medication errors ICU length of stay ICU mortality Hospital mortality | Paper to commercial CPOEs in ICUs resulted in 85% reduction in medication errors, and 12% reduction in ICU mortality rates. There were no differences in length of stay or hospital mortality rates. |
CDS, clinical decision support; CDSS, clinical decision support systems; ICU, intensive care unit; RCT, randomised controlled trial; SRs, systematic reviews.