Table 1 Characteristics and summary of findings of the included studies
Study (country) | Population | Prescribing-associated HIT | Study design | Hospital setting | Sample | Error detection method | Baseline error rate (%) | Technology-generated errors (TGE) detected |
|---|---|---|---|---|---|---|---|---|
Abbass 2011 [37] (USA) | Adult | CPOE (commercial) | Control group | All areas | 1110 (orders) | Chart review | 20.8 | Lack of CDS led to allergy/DDI errors. |
Ali 2010 [39] (UK) | Adult | CPOE (commercial) | Time series analysis | ICU | 14721 (prescriptions) | Routine pharmacist review; Chart review | 16.7 | Allergy alert did not fire if the allergy field was not already completed by the prescriber. |
Al-Sarawi 2019 [38] (Australia) | Adult | ePrescribing (EP) (commercial) | Pre-post | All areas | 4689 (orders) | Chart review | 67.7 | Duplicate orders increased post-CPOE. |
Armada 2014 [40] (Spain) | Adult | CPOE (commercial) | Time series analysis | ICU | 5729 (orders) | Routine pharmacist review | 44.8 | Selection errors made while searching for drugs on drop-down menus. |
Bates 1998 [42] (USA) | Adult | CPOE (homegrown) | Pre-post | 2 medical wards; 2 surgical wards; 2 ICUs | 24453 (pt. days) | Routine pharmacist review; chart review | 5.0 | Increase in therapeutic duplication of sedating drugs, which the CPOE did not prevent. |
Bates 1999 [41] (USA) | Adult | CPOE (homegrown) | Time series analysis | 3 medical units | 7985 (pt. days) | Routine pharmacist review; chart review; medication order review | 1.7 | Missed dose errors (not main outcome of interest) increased with CPOE. |
Bizovi 2002 [67] (USA) | Adult/paediatric | EP (commercial) | Pre-post | ED | 3920 (prescriptions) | Routine pharmacist review; medication order review | 2.3 | Free-text electronic prescriptions had a higher rate of error than the pick-list prescriptions. |
Boling 2005 [57] (USA) | Paediatric | CPOE (commercial) | Pre-post | All areas | 21253 (orders) | Trigger tool methodology; chart review; voluntary error reports | 0.1 (opioids) | None found. |
Colpaert 2006 [43] (Belgium) | Adult | CPOE (commercial) | Prospective controlled trial | 3 units in an ICU | 2510 (prescriptions) | Chart review; medication order review | 27.0 | CPOE errors were mostly duplicate prescriptions. |
Cordero 2004 [58] (USA) | Neonatal | CPOE (commercial) | Pre-post | NICU | 211 (patients) | Chart review | 12.6 (gentamicin) | None found. |
Delgado Silveira 2007 [44] (Spain) | Adult | EP (commercial) | Pre-post | 2 medical units | 4814 (prescriptions) | Routine pharmacist review | 94.2 | Drug interaction errors increased with CPOE, this was not significant. |
Donyai 2008 [45] (UK) | Adult | EP (commercial) | Pre-post | Surgical ward | 4803 (orders) | Routine pharmacist review; chart review; medication order review | 3.8 | Selection errors were found post-EP. 1 wrong-patient error post-EP, authors uncertain if TGE. |
Hernandez 2015 [46] (France) | Adult | CPOE (commercial) | Pre-post | Orthopaedic unit | 2981 (orders) | Chart review; Direct observation | 30.1 | Duplicate orders increased with CPOE. |
Hitti 2017 [47] (Lebanon) | Adult | EP (homegrown) | Pre-post | ED | 2883 (prescriptions) | Chart review | 67.7 | Duplicate errors increased with CPOE. |
Hodgkinson 2017 [68] (Australia) | Adult/paediatric | EP (commercial) | Pre-post | ED and OPD | 1289 (orders) | Routine pharmacist review; medication order review | 95.0 | 50 systems-related errors post-CPOE, such as selection errors or not filling in necessary fields. |
Howlett 2020 [59] (Ireland) | Paediatric | EP (commercial) | Time series analysis | PCCU | 3356 (orders) | Routine pharmacist review | 10.2 | Incorrect formulation and dose errors increased with EP. |
Jani 2008 [60] (UK) | Paediatric | EP (commercial) | Pre-post | Nephrology OPD | 2222 (orders) | Routine pharmacist review; Chart review | 7.1 | Duplicate orders increased with CPOE. Wrong route, frequency, and overdose also found due to selection errors. |
Kadmon 2009 [61] (Israel) | Paediatric | CPOE (commercial) | Time series analysis | PICU | 5000 (orders) | Medication order review | 5.5 | Prescriptions were found to be prescribed by nurses, due to doctors using computers where a nurse was already logged in. |
Kazemi 2011 [62] (Iran) | Neonatal | CPOE (commercial) | Time series analysis | Neonatal unit | 4508 (medication days) | Medication order review | 51.9 | ‘Neighbouring cell’ errors were noted, where a prescriber chose a nearby cell in error or used incorrect data to do dose calculations. |
Kenawy 2019 [69] (Egypt) | Adult/paediatric | EP (commercial) | Pre-post | 4 OPDs (cardiology, nephrology, paediatric, neurology) | 25057 (orders) | Voluntary error reports | 28.3 | Indication and omission prescribing errors increased with EP. |
King 2003 [63] (Canada) | Paediatric | CPOE (commercial) | Pre-post | 3 medical wards; 2 surgical wards | 12460 (patients) | Voluntary error reports | 0.1 | None found. |
Liao 2017 [48] (USA) | Adult | CPOE (commercial) | Time series analysis | ICU | 3988 (pt. days) | Chart review | 86.6 | Reduction in errors only evident 2 years post-implementation. |
Mahoney 2007 [70] (USA) | Adult/paediatric | CPOE (commercial) | Pre-post | All areas | 2843165 (orders) | Routine pharmacist review | 0.33 | Duplicate errors did not significantly decrease with CPOE. |
Mills 2017 [49] (UK) | Adult | EP (commercial) | Pre-post | All areas | 318 (patients) | Chart review; medication order review | 99.4 | 8/37 errors post- EPÂ were selection errors on menus. |
Pontefract 2018 [50] (UK) | Adult | CPOE (commercial) | Pre-post | All areas | 2422 (patients) | Trigger tool methodology; routine pharmacist review; chart review | 5.0 | Increase in insulin prescribing errors with CPOE in 1 site due to lack of CDS. |
Potts 2004 [64] (USA) | Paediatric | CPOE (homegrown) | Pre-post | CCU | 13828 (orders) | Routine pharmacist review; medication order review | 30.1 | Dose errors related to trailing decimal points or missing weights occurred with CPOE. |
Riaz 2014 [51] (Pakistan) | Adult | EP (homegrown) | Control group | 2 OPD and 2 ED | 2040 (prescriptions) | Medication order review | 52.0 | Omission errors higher on EPÂ prescriptions which caused error increase. |
Rouayroux 2019 [52] (France) | Adult | CPOE (commercial) | Time series analysis | Cardiology and diabetes depts. | 3086 (pt. days) | Routine pharmacist review | 12.1 | Unit of use errors and duplicate orders increased with CPOE. |
Shawahna 2011 [71] (Pakistan) | Adult/paediatric | EP (homegrown) | Pre-post | All areas | 32662 (orders) | Chart review; medication order review | 21.7 | None found. |
Shulman 2005 [53] (UK) | Adult | CPOE (commercial) | Pre-post | ICU | 3465 (prescriptions) | Routine pharmacist review | 6.4 | Errors related to overdose increased with CPOE, with the potential to cause serious morbidity or mortality. Orders were frequently unsigned and therefore invalid. |
Spencer 2005 [54] (USA) | Adult | CPOE (commercial) | Pre-post | 2 medical units | 4339 (pt. discharges) | Voluntary error reports | 1.4 | 23 reported errors caused by CPOE, including allergy errors, duplicate orders, input errors, and discrepancies when transcribing to pharmacy computer. |
Van Doormaal 2009 [55] (The Netherlands) | Adult | CPOE (site 1 commercial) (site 2 homegrown) | Pre-post | 4 medical wards | 1195 (patients) | Chart review; medication order review | 78.6 | Overriding of alerts occurred with CPOE due to alert fatigue. |
Venkataraman 2016 [65] (UK) | Paediatric | EP (homegrown) | Pre-post | PCCU | 251 (prescriptions) | Routine pharmacist review | 32.6 | Wrong-patient error due to manual input of date of birth. |
Warrick 2011 [66] (UK) | Paediatric | EP (commercial) | Time series analysis | PICU | 624 (prescriptions) | Chart review | 8.8 | Infusions were prescribed with no diluent or rate with EP. |
Westbrook 2012 [56] (Australia) | Adult | CPOE (commercial) | Difference in differences | 6 medical wards | 15194 (pt. days) | Routine pharmacist review | 48.5 | Selection errors occurred with CPOE. |