Evidence Table 1KQ1: Articles assessing primary process outcomes for all technologies assisting the prescribing phase of medication management

Article InformationHIT Studied
Integrated system
SettingsOutcomes MeasuredResultsOutcome
Abboud (2006) (Abboud et al. 187–198)
Design: Before-after
N = 336 orders
Implementation: 04/2002
Study Start: 10/2003
Study End: 03/2004
CDSS/CDS/CCDS/reminders CPOE/POE system
Integrated CDSS/CDS/CCDS/reminders EHR/EMR system, Formulary, Hospital information system, Imaging systems, Laboratory system, Pharmacy
Pediatric stand alone hospital, 423 Bedsantibiotics courses with no lab order*no significant differences between the baseline and the corollary order periods on courses of antibiotics associated with no laboratory monitoring 31 (19.5%) vs. 31(17.5%), p = NS.
Achtmeyer (2002) (Achtmeyer, Payne, and Anawalt 277–281)
Design: Before-after
N = 1,405 orders for supplemental insulin
Implementation: 12/1998
Study Start: 12/1998
Study End: 07/1999
CDSS/CDS/CCDS/reminders CPOE/POE system
Integrated EHR/EMR system, Imaging systems, Laboratory system
Acute care/tertiary, 290 Beds Academicrate of traditional sliding scale orders for supplemental insulin*rate of traditional sliding scale orders for supplemental insulin in hospitalized patients was reduced when a quick-order CPOE/CDSS system was put in place (97.1% vs. 63.8%, RRR 34%, p <0.001).+
Agostini (2007) (Agostini, Shang, and Inouye 43–48)
Design: Before-after
N = 24,509 patients
Implementation: 04/2002
Study Start: 04/2002
Study End: 03/2003
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system, Formulary
Acute care/tertiary, 944 Beds Academicrate of prescribing of sedative-hypnotics*Prescribing of sedative-hypnotics decreased from 2,208 per 12,356 (18%) patients preintervention to 1,832 per 12,153 (15%) postintervention (OR for the intervention = 0.82, 95% CI = 0.76–0.87), an 18% risk reduction (p <0.001 for pre/post difference).+
Ali (2005) (Ali et al. 110–114)
Design: Before-after
N = 91 patients
Implementation: 02/2000
Study Start: 05/2000
Study End: 05/2002
CPOE/POE systemCritical care units (CCU, ICU, NICU) 25 Beds Academicmean number of orders for vasoactive drips per patient, mean number of orders for sedative infusions per patientCompared to the initial CPOE, the redesign of the CPOE system to incorporate more complex order sets resulted in significantly fewer orders placed per patient (means) for vasoactive drips (4.8 vs. 2.2, p <0.01) and sedative infusions (6.4 vs. 2.9, p <0.01), as a measure of improved workflow efficiency.+
Bailey (2007) (Bailey et al. 586–590).
Design: RCT
N = 853 patients
Implementation: 00/0000
Study Start: 02/2000
Study End: 05/2001
CDSS/CDS/CCDS/reminders
Integrated Hospital information system, Laboratory system
Acute care/tertiary, 1,385 Beds Inpatient hospital based, Academiccompliance rates:
-

patients discharged on a full complement regimen of secondary prevention medications*

-

ACE inhibitor*,

-

statins*,

-

aspirin

-

beta-blockers.

When individual drug class exclusions were considered, compliance rates increased for patients discharged on a full-complement regimen of secondary prevention medications (70.3% vs. 83.6%, RRR - 19%, p <0.001).
Compliance rates for ACE inhibitor (83.6 vs. 89.9, RRR - 8%, p = 0.01) and statin use (89.3 vs. 94.2%, RRR - 5%, p = 0.02) were significantly higher, while rates for aspirin (96.5% vs. 96.4%, RRR 0%, p = 0.95) and beta- blockers (91.8% vs. 95.9%, RRR - 5%, p = 0.08) remained the same.
+
Bates (1999)
(Bates et al. 313–321)
Design: Time series
N = 1,817 admissions
Implementation: 05/1993
Study Start: 10/1992
Study End: 04/1997
CDSS/CDS/CCDS/reminders
Integrated Billing/administration system Hospital information system
Acute care/tertiary 700 Beds AcademicRate of non-missed dose errors per 1,000 patient-days over 4 time periods*, Rate of non-missed dose errors per admission*The rate of errors (other than missed dose) per 1000 patient-days fell from baseline across all time points for medication errors: non-missed-dose medication errors (142, 51.2, 74, 2666; p = 0.0001).
The results were similar for non-missed-dose error rate per admission (0.64, 0.27, 0.28, 0.11, p = 0.0001). Non-intercepted serious medication errors declined significantly over time (7.6, 7.3, 1.7, 1.1, p = 0.0003).
+
Bates (1994) (Bates, Boyle, and Teich 996)
Design: Before- after
N = 62 Physicians (Medical Interns and 1st and 2nd year surgical residents)
Implementation: 00/0000
Study Start: 00/0000
Study End: 00/0000
CPOE/POE system
Integrated Hospital information system
Unspecified HospitalTime spent ordering by medical interns*, Time spent ordering by surgical residents*, Time spent on daily and one-time orders*, Time spent on sets of orders*When time spent ordering was compared between pre- order entry and post-order entry periods, the percent for medical interns increased from 5.3% to 10.5% (p <0.001) representing 44 additional minutes per day, while for surgical house officers the corresponding figures were an increase from 6.4% to 15.5% (p <0.001), 73 minutes per day. Daily and one-time orders accounted for the majority of this change, increasing almost threefold in percent total time (2.2% before, vs. 7.2% after order entry). However, sets of orders took less total time after order entry (1.7% vs. 3.1%).
Bates (1998) (Bates et al. 1311–1316)
Design: RCT
N = 4,220 patients
Implementation: 00/0000
Study Start: 02/1993
Study End: 07/1995
CPOE/POE system
Integrated Billing/administration system, Formulary, Hospital information system
Acute care/tertiary, 726 Beds Academicthe rate of nonintercepted serious medication errors/1,000 patient days -phase 1 to 2*, the rate of nonintercepted serious medication errors/1,000 patient days -CPOE vs. CPOE+team, Transcription errorsIn paired analyses comparing phase 1 and phase 2 (Table 2), the rate of nonintercepted serious medication errors fell 55%, from 10.7 events per 1,000 patient- days to 4.86 events (p = 0.01). For the RCT in the post-CPOE phase, comparing CPOE alone with CPOE plus team showed no significant difference in error rates (4.81 vs. 6.01, p = 0.49). Transcription errors (CPOE to paper in pharmacy) fell 84%, p <0.001.+
Bell (2010) (Bell et al. e770–e777)
Design: RCT
N = 19,450 patients
Implementation: 00/0000
Study Start: 04/2007
Study End: 04/2008
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory care, Academicproportion of children with asthma having at least 1 prescription for controller medication*, proportion of children with asthama having an up-to-date asthma care plan*, proportion of children with asthma having spirometry performed*Increases in the number of prescriptions for controller medications, over time, was 6% greater (p = 0.006) and 3% greater for spirometry (p = 0.04) in the intervention urban practices. Filing an up-to-date asthma care plan improved 14% (p = 0.03) and spirometry improved 6% (p = 0.003) in the suburban practices with the intervention.+
Berner (2006) (Berner et al. 171–179)
Design: RCT
N = 59 internal medicine residents
Implemgentation: 00/0000
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Handheld
Ambulatory care, Academicproportion of unsafe NSAID prescribing per physician at followupThe proportion of cases per physician with unsafe NSAID prescriptions were similar at baseline for control (0.29) and intervention residents (0.27). At followup, the rates were statistically different, with lower proportions for intervention residents after adjustment for baseline rates (0.45 control vs. 0.23 intervention, p <0.05). Note that the control group prescribing degraded over time while the intervention group was stable.
Bernstein (2005) (Bernstein et al. 225–231)
Design: Before-after
N = 1,158 prescriptions
Implementation: 00/0000
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
Emergency department, Academicpercentage of proprietary antibiotics prescribed*The percentage of proprietary antibiotics prescribed before and after insertion of the electronic prompt was 26.6% vs. 20.7%, RRR 22%, p = 0.03.+
Bertoni (2009) (Bertoni et al. 678–686)
Design: RCT
N = 8,878 patients
Implementation: 00/0000
Study Start: 06/2001
Study End: 04/2006
CDSS/CDS/CCDS/reminders
Handheld
Ambulatory careadherence to guideline-screening*, adherence to guideline-appropriate lipid management*There was no difference in screening rates between the CDSS-PDA group and the control. The control group had a 10.8% drop in appropriate management from baseline, while the PDA group had a 1.1% drop, p <0.01. Stable adherence was observed in the PDA intervention group, whereas a decline in guideline adherence was observed in the control group.
Bloomfield (2005) (Bloomfield et al. 258–263)
Design: RCT
N = 9,105 patients
Implementation: 04/2002
Study Start: 10/2001
Study End: 10/2003
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory carerate of prescription lipid therapy (before-after)rate of lipid therapy prescriptions increased significantly after implementation of the prompts in the intervention clinics (8.3% vs. 39.1%, RRR -371, p <0.0001) no statistically significant difference in prescription rates (40.7% for progress notes, 36.9% for patient letters, and 39.4% for reminders, p = 0.60) alternative logistic regression analysis, significant interaction between group and site, indicating that the efficacy of the prompts differed by site.+
Bogucki (2004) (Bogucki et al. 278–280)
Design: Before- after
N = 2,124 orders for parenteral corticosteroids
Implementation: 04/2002
Study Start: 04/2003
Study End: 07/2003
CDSS/CDS/CCDS/reminders
Integrated CDSS/CDS/CCDS/reminders
CPOE/POE system
EHR/EMR system, e-MAR
Pediatric stand alone hospital, 324 Bedsrate of methylprednisone ordering*There was a significant reduction in methylprednisone prescribing following the implementation of the alert in relation to the total number of parenteral corticosteroids ordered (21.5% vs. 9.7%, RRR 55%, p <0.0001).+
Bouaud (2001) (Bouaud et al. 1–4)
Design: Before-after
N = 127 decisions/orders
Implementation: 00/0000
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory care Otherrate of compliance with CPGBefore using OncoDoc, phyicians compliance with CPG was 61.42%. Using the system significantly increased actual compliance to 85.03% (p <0.0001).+
Buising (2008) (Buising et al. 35)
Design: Time series
N = 740 patients
Implementation: 01/2005
Study Start: 04/2003
Study End: 09/2006
CDSS/CDS/CCDS/reminders
Integrated Hospital information system
Acute care/tertiary, 350 Beds Academicproportion of patients receiving appropriate antibiotic therapy*proportion of patients receiving appropriate antibiotic therapy increased significantly between each time period (61.9% baseline vs. 68.7 academic detailing vs. 89.7 CDSS, pairwise comparisons p <0.01) associated ORs for having received the recommended empiric antibiotic therapy were 2.58 between baseline and CDSS periods and 2.03 between academic detailing and CDSS.+
Butler (2006) (Butler et al. 643–653)
Design: Before-after
N = 1,827 patients (1,251 with CHF and 576 with AMI)
Implementation: 07/2002
Study Start: 07/2001
Study End: 0.9/2003
CDSS/CDS/CCDS/reminders
CPOE/POE system
Acute care/tertiary, Academiccompliance rate: ACEI for LVSD*, compliance rate: ACEI for AMI*, compliance rate: aspirin for AMI*, compliance rate: beta-blocker for AMI*Aspirin (95% vs. 95%, RRR 0%, NS), betablocker (88% vs. 95%, RRR -8%, NS), and ACEI (77% vs. 81%, RRR - 5%, NS) use for AMI patients at the time of discharge in the pre-CPOE era was high and remained so in the CPOE period. Similarly for ACEI for CHF patients (74% vs. 87%, RRR - 18%, NS). When examining indicators in the post-CPOE phase, rates were higher in patients for which the tool was used, vs. not used for all 4 medication related indicators (p <0.001).
Chertow (2001) (Chertow et al. 2839–2844)
Design: Time series
N = 19,982 admissions
Implementation: 00/0000
Study Start: 09/1997
Study End: 04/1998
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated Hospital information system
Imaging systems
Acute care/tertiary, 720 Beds Academicrate of appropriate prescribing*, rate of appropriate prescribing involving dosage alterations*, rate of appropriate prescribing involving frequency alterations*The rate of appropriate prescribing was increased with CPOE/CDSS for all orders (51% intervention vs. 30% control, RRR 70%, p <0.001) by dose (67% vs. 54%, RRR 43%, p <0.001), or by frequency (59% vs. 35%, RRR 69%, p <0.001).+
Chisholm (2003) (Chisholm et al. 199–206)
Design: Before- after
N = 790 children admitted to hospital with asthma exacerbations
Implementation: 10/2002
Study Start: 11/2001
Study End: 12/2003
CPOE/POE system
Integrated Billing/administration system, EHR/EMR system, Laboratory system
Pediatric stand alone hospital, 323 Bedssystemic corticosteroids use*, metered-dose inhaler use*More use was made of systemic corticosteroids (OR 5.61, 95% CI 3.46 to 9.11) and of metered- dose inhalers (OR 1.42, CI 1.04 to 1.94) after implementation of standard order sets in the CPOE for asthma patients.+
Choi (2004) (Choi et al. 1–6)
Design: Before- after
N = 307 patients
Implementation: 02/2003
Study Start: 12/2002
Study End: 04/2003
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated EHR/EMR system
Ambulatory careError rates per patient*Error rates per patient significantly declined in the intervention site following implementation of the nurse CPOE with CDSS (17.4% vs. 3.1%, RRR 82%, p = 0.0075). In the control group, error rates remained unchanged (8.6% vs. 6.9%, NS). At baseline, the control group rate was statistically lower than the intervention group (8.6% vs. 17.4%, p = 0.04).+
Christakis (2001) (Christakis et al. e15)
Design: RCT
N = 38 providers
Implementation: 00/0000
Study Start: 03/0000
Study End: 05/0000
CDSS/CDS/CCDS/reminders
Integrated online prescription writer
Ambulatory care, Academicchange in the frequency of antibiotic prescription*For the primary outcome, providers in the intervention arm had a 44% change in the frequency with which they prescribed antibiotics for <10 days, whereas providers in the control arm had a 10% change, this change in behavior was significantly related to the intervention, although both groups improved (p <0.01).+
Clancy (1992) (Clancy, Gelfman, and Poses 14–18)
Design: Before-after
N = 1,013 patients
Implementation: 02/1985
Study Start: 11/1984
Study End: 05/1985
CDSS/CDS/CCDS/reminders
Integrated Hospital information system
Acute care/tertiary, Academicpneumococcal vaccination rate per admission*Preimplementation of the reminder pneumococcal vaccination rate was 3.4% compared 45% post (p <0.0001).+
Cobos (2005) (Cobos et al. 421–432)
Design: RCT
N = 2,221 patients
Implementation: 04/2000
Study Start: 04/2000
Study End: 05/2002
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory careproportion of patients prescribed lipid lowering drugs (secondary)The proportion of patients prescribed lipid lowering drugs was significantly lower in the CDSS guideline intervention group (59.1% vs. 40.8%, RRR 31%, p <0.0001).+
Colpaert (2006) (Colpaert et al. R21)
Design: RCT
N = 2,510 prescriptions
Implementation: 00/0000
Study Start: 03/2004
Study End: 04/2004
CPOE/POE system
Integrated Billing/administration system, CPOE/POE system, Hospital information system, Laboratory system
Acute care/tertiary, Critical care units (CCU, ICU, NICU) 22 bed unit Beds Academicrate of medication prescribing errors*, minor MPEs*, Intercepted MPEs *, Serious MPEs *The incidence of MPEs was significantly lower in the computerized unit (C-U) compared with the paper based unit (PBU) [44/1,286 (3.4%) vs. 331/1,224 (27.0%); p <0.001]. There were significantly fewer minor MPEs in the C-U than in the PB-U [9 (0.7%) vs. 225 (18.4%); p <0.001)]. Intercepted MPEs were also lower in the C-U [12 (0.9%) vs. 46 (3.8%); p <0.001]. Serious MPEs were also lower in C-U than PBU [23 (1.8%) vs. 60 (4.9%), p <0.001].+
Cordero (2004) (Cordero et al. 88–93)
Design: Before-after
N = 211 infants
Implementation: 02/2000
Study Start: 10/2001
Study End: 09/2002
CPOE/POE system
Integrated Imaging systems, Pharmacy
Acute care/tertiary, Critical care units (CCU, ICU, NICU) Academicmedication turn-around times-caffeine*, medication error rate-gentamicinThe turn- around times for the pre- and post-CPOE loading dose of caffeine were 10.5 ± 9.8 and 2.8 ± 3.3 hours p <0.01, respectively. In the pre-CPOE period, there were 14 (13%) gentamicin prescription dosage errors, in the post- CPOE period there were 0+
Cote (2008) (Cote et al. 1097–1103)
Design: Before-after
N = 601 adult patients
Implementation: 00/0000
Study Start: 12/2005
Study End: 06/2006
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Unspecified Hospitalrate of gastroprotection at discharge*, control vs. physician education vs. alert vs. alert plus educationThe study sought the change in rate of gastroprotection at discharge for all patients; changes only occurred for the group that had both education and alerts compared to control (43% vs. 61%, RRR - 42%, p <0.001). Education alone (42%) or alerts alone (39%) did not change rates of gastroprotection.
Cunningham (2008) (Cunningham, Geller, and Clarke 546–554)
Design: Before- after
N = 1,040 order sets
Implementation: 00/0000
Study Start: 00/0000
Study End: 00/0000
CPOE/POE system
Integrated CDSS/CDS/CCDS/reminders
Acute care/tertiary, General Hospital 667 Bedscompliance to medication order sets*, minutes to first dose of antibioticsMedication orders placed using CPOE were significantly more compliant with hospital protocols (80%) than paper based medication orders at both the CPOE hospital (63%) and the control hospital (64%), and first doses of antibiotics were delivered significantly faster when ordered with CPOE (180 min) than when placed using the standard paper-based system (326 min, p <.01).+
Davis (2007) (Davis et al. e25)
Design: RCT
N = 44 health care providers
Implementation: 11/1999
Study Start: 11/1999
Study End: 12/2003
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
Ambulatory care, Academicchanged physician behavior in accordance with the intervention message screens*Prescribing behavior in accordance with the evidence improved only marginally, by 1% in control group and 4% in the intervention group (absolute difference 3%, 95% CI 1%, 15%).+
de Jong (2009) (de Jong et al. 9–20)
Design: Cross- sectional
N = 749,811 contacts
Implementation: 00/1998
Study Start: 01/2001
Study End: 12/2001
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory careproportion of prescriptions in accordance with DSS*, Herfindahl-Hirschman IndexGPs who use the DSS daily prescribe more according to the advice given in the DSS (89%) than GPs who do not use the DSS (75%, RRR 19%, p = 0.04). There was no significant difference between the Herfindahl- Hirschman Index for both groups (40.3 for daily users and 41.4 for non users, p = 0.3) the variation in prescriptions for a given diagnoses was comparable between groups.+
Devine (2010) (Devine et al. 928)
Design: Before- after
N = 10,169 prescriptions
Implementation: 03/2003
Study Start: 03/2002
Study End: 04/2006
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated EHR/EMR system
Ambulatory careerrorsFrequency of errors declined from 18.2% (Pre-CPOE) to 8.2% (post- CPOE), a reduction in adjusted odds of 70% (OR: 0.30; 95% CI 0.23 to 0.40), p <0.001.+
Dexter (2001) (Dexter et al. 965–970)
Design: RCT
N = 3,416 patients
Implementation: 00/0000
Study Start: 05/1997
Study End: 10/1998
CDSS/CDS/CCDS/reminders
Integrated Pharmacy
General Hospital Academicproportion compliance:
-

pneumococcal vaccination*,

-

influenza vaccination*

-

subcutaneous heparin

-

aspirin at discharge

The use of the reminders led to a higher ordering rate all 4 preventive therapies for eligible patients; pneumococcal vaccination (0.8% vs. 35.8%, RRR - 4375%, p <0.001), influenza vaccination (1.0% vs. 51.4%, RRR - 5040%, p <0.001), subcutaneous heparin (18.9% vs. 32.2%, RRR -70%, p <0.001) and aspirin at discharge (27.6% vs. 36.4%, RRR - 32%, p <0.001).+
Dexter (2004) (Dexter et al. 2366–2371)
Design: RCT
N = 1,677 patients
Implementation: 11/1998
Study Start: 11/1998
Study End: 12/1999
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
General Hospital, Academicrate of receipt of vaccination -influenza*, rate of receipt of vaccination - pneumococcal*Patients in the standing order group received both vaccinations more often than patients in the pop-up reminder group; for the influenza vaccine 30% reminder vs. 42% standing order, p <0.001; for the pneumococcal vaccine 51% vs. 31%, p <0.001.+
Durieux (2000) (Durieux et al. 2816–2821)
Design: Time series
N = 1,971 patients
Implementation: 00/0000
Study Start: 12/1997
Study End: 07/1999
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system, Hospital information system
Acute care/tertiary, 1,000 Beds Academicrate of appropriate anticoagulant prescribing*Physicians complied with guidelines in 82.8% of cases during control periods and in 94.9% of cases during intervention periods (RRR - 15%, p <0.001). During each intervention period, the proportion of appropriate prescriptions ordered increased significantly. Each time the CDSS was removed, physician compliance with guidelines reverted to that observed before initiation of the intervention.+
Eslami (2006) (Eslami et al. 803–809)
Design: Cross- sectional
N = 392 orders
Implementation: 00/2002
Study Start: 05/2002
Study End: 12/2004
CPOE/POE system
Integrated EHR/EMR system, Laboratory system
Acute care/tertiary, Critical care units (CCU, ICU, NICU) 28 in 3 units BedsDosing error*The dose was wrong (i.e. there was >10% deviation from the guideline) in 73% (165/227) of the orders that used the default value (essentially as suggested by the CPOE) and in 77% (127/165) of the orders in which the default value was not administered (p = 0.4).
Evans (1998) (Evans et al. 232–238)
Design: Before- after
N = 1,681 patients
Implementation: 00/0000
Study Start: 07/1992
Study End: 06/1995
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system, Imaging systems, Laboratory system
Acute care/tertiary, 520 Beds Academicmean number of days with excessive antibiotic dosing*, usage rate of antiinfectives*During the intervention period, there were significantly fewer days when doses of antiinfective agents were excessive than during the preintervention period (2.7 days vs. 5.9 days per patient, respectively; p <0.002). There was an increase in the use of antiinfectives following the intervention reminder (67% vs. 73%, RRR 9%, p <0.03).+
Evans (1990) (Evans et al. 351–354)
Design: Before- after
N = 7,656 patients
Implementation: 00/0000
Study Start: 06/1985
Study End: 09/1986
CDSS/CDS/CCDS/reminders
Hospital information system
Integrated Laboratory system, Pharmacy
Unspecified Hospitalmean number of antibiotic doses per patient, proportion of patients receiving preoperative antibiotics, proportion of patients receiving antibiotics for too long,Surgical patients received an average of 19 antibiotic doses before implementation of the implementation of the ‘stop orders’ and 13 after (p <0.001). There were non significant changes in the proportion of patients receiving preoperative antibiotics (64% vs. 66%, NS) or those receiving antibiotics for too long (40% vs. 35%, NS).+
Evans (1994) (Evans et al. 878–884)
Design: RCT
N = 482 cultures
Implementation: 00/000
Study Start: 07/1990
Study End: 01/1991
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system, Laboratory system
Acute care/tertiary 520 Bedsrate of prescribing antibiotics to which all of the isolated pathogens were susceptibleThe computer group had a higher rate of prescribing antibiotics to which all of the isolated pathogens were susceptible (77% vs. 94%, RRR 22%, p <0.001).+
Feldstein (2006) (Feldstein et al. 1009–1015)
Design: RCT
N = 9,910 patients with 239 care providers in 15 primary care clinics
Implementation: 12/2002
Study Start: 01/2000
Study End: 08/2004
CDSS/CDS/CCDS/reminders
Integrated CDSS/CDS/CCDS/reminders
CPOE/POE system, EHR/EMR system
Ambulatory careinteracting prescription rate (/10,000 warfarin users/month), slope for interacting prescription rateWhen baseline trends were controlled for, the overall interacting prescription rate decreased immediately after the alerts were implemented, with an estimated reduction of 329.7 interacting prescriptions per 10, 000 warfarin users in the first month (p = 0.002). The alerts also significantly changed the trend in the interacting prescription rate, with a preintervention increasing rate of 1.1 and a postintervention decreasing rate of 21.3 (slope change −22.4; p = 0.01). Academic detaining did not have an effect on interacting prescription rates.+
Feldstein (2006) (Feldstein et al. 450–457)
Design: RCT
N = 311 women
Implementation: 00/0000
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system, Laboratory system
Ambulatory carerate of completion of BMD or medication for osteoporosis, The same pattern was evidence for medication onlycontrol group had fewer women who had BMD completer or medication for osteoporosis compare with the reminder and reminder plus education groups (5.9% control, 51.5% reminders, and 33% reminders and education, p <0.01 for both comparisons with control RRR for reminders alone 690% and RRR for reminders and education 460%). The same pattern was evidence for medication only (5.0% control, 27.7% reminders and 20.2% reminders plus education; p <0.01 for comparisons with control.+
Field (2009) (Field et al. 480–485)
Design: RCT
N = 833 patients (10 physicians and 213,967 patient days)
Implementation: 00/000
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated EHR/EMR system
Long term care (nursing homes)proportion of appropriate orders*, proportion of inappropriate drugs avoidedThe proportion of appropriate antidepressant order rates for patients with renal insufficiency was higher in the CDSS group (52% vs. 63%, OR 1.2, 95% CI 1.0 to 1.4). More inappropriate drugs were avoided (15% vs. 46%, OR 2.6, CI 1.4 to 5.0). Improvements were seen in frequency and missing information but not for doses in the CDSS group.+
Fiks (2009) (Fiks et al. 159–169)
Design: RCT
N = 22,586 patients
Implementation: 00/0000
Study Start: 10/2006
Study End: 05/2007
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system, EHR/EMR system
Ambulatory care, Academicrates of up-to-date influenza vaccination*, rates of captured opportunities for vaccination*Rates of up-to- date influenza vaccination increased from 44.2% to 48.2% (control) and from 45.0% to 53.0% (intervention), a 4.0% (95% CI: − 1.3% to 9.1%) greater but NS. Overall rates of captured opportunities for vaccination increased 3.8% (12.3% to 16.1%) control practices and 4.8% (14.4% to 19.2%) intervention sites, difference 1% (95% CI: − 2.4% to 4.9%).
Filippi (2003) (Filippi et al. 1497–1500)
Design: RCT
N = 15,343 patients
Implementation: 00/0000
Study Start: 05/2001
Study End: 11/2001
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system, EHR/EMR system
Ambulatory careAntipletlet drug treatmentnumber of treated patients significantly increased in the intervention group (OR 1.99, 95% CI 1.79 to 2.22).+
Fischer (2003) (Fischer et al. 2585–2589)
Design: Before-after
N = 1,045 orders
Implementation: 00/00
Study Start: 00/00
Study End: 00/00
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
Acute care/tertiary, Inpatient hospital based Academicdefined daily dose -IV, defined daily dose –oral

DDD
After implementation the use of IV medication (DDD) decreased by 11.1%, p = 0.002 and the oral drug use (DDD) increased by 3.7%, p = 0.002.+
Fischer (2008) (Fischer et al. 2433–2439)
Design: Before-after
N = 12,625,276 prescriptions
Implementation: 10/2003
Study Start: 10/2003
Study End: 5/2005
e-Rx
Integrated Formulary, Insurance
Not specifiedrates of prescribing, tier 1*, rates of prescribing, tier 2*, rates of prescribing, tier 3*20% of prescriptions written by intervention physicians completed using e-Rx intervention group prescribed 1.4% more (95% CI, 0.6% to 2.0%) tier 1 medications, 0.3% fewer (95% CI, −0.8% to 0.2%) tier 2 medications, and 1.0% fewer (95% CI, −1.4% to −0.7%) tier 3 medications than the control group.+
Flottorp (2002) (Flottorp et al. 367)
Design: RCT
N = 26,826 Consultation
Implementation: 00/0000
Study Start: 01/2000
Study End: 01/2001
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory careUse of antibiotics for sore throat, use of antibiotics for UTIsore throat group 3% less likely to receive antibiotics after the intervention (49.5% vs. 43.8%, p = 0.032) UTI (43.4% vs. 46.3%, p = 0.639) Women with symptoms of UTI in the intervention group were 5.1% less likely to have a laboratory test ordered (55% vs. 49.8%, p = 0.046) For sore throat, the numbers were 39.7% vs. 42.0%, p = 0.638 proportion of telephone consultations sore throat: 1.2% greater in the control group than in the intervention group (14.1% vs. 12.9%, p = 0.128) proportion decreased for UTI (18.9% vs. 19.8%, p = 0.874)
Fontan (2003) (Fontan et al. 112–117)
Design: Cross-sectional
N = 4,532 prescriptions
Implementation: 00/1988
Study Start: 02/1999
Study End: 03/1999
Computerized unit dose drug dispensing system (UDDS)
Integrated Hospital information system
Other specialty hospital (rehab, oncology) Pediatric stand alone hospital 510 Bedsprescribing error rate administering error rateErrors were decreased with the use of the eRX and computerized dispensing system compared with the hand- written prescriptions and ward distribution system. Prescribing errors were reduced from 87.9% to 10.6%, RRR 88%, p <0.00001. Administrative errors with time errors were reduced from 29.3% to 22.5%, RRR 23%, p <0.001.+
Fortuna (2009) (Fortuna et al. 897–903)
Design: RCT
N = 257 clinicians
Implementation: 00/1997
Study Start: 03/2006
Study End: 03/2008
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system, e-Rx
Ambulatory care, Academicrelative risk of prescribing heavily marketed medications*The relative risk of prescribing heavily marketed medications in the alert-group during the intervention period was less than in the usual-care group (RRR 0.74; 95% CI 0.57 to 0.96; p = 0.02). The RR of prescribing heavily marketed hypnotics in the alert-plus- education group was less than in the usual-care group (RRR 0.74; 95% CI 0.58 to 0.97, p = 0.03). The prescribing of heavily marketed medications was similar in the alert-only group and the alert-plus- education group (RRR 1.02; 95% CI 0.80 to 1.29; p = 0.90).+
Frances (2001) (Frances et al. 165–166)
Design: RCT
N = 63 physicians and 730 patients
Implementation: 00/0000
Study Start: 03/1997
Study End: 06/1997
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system, Pharmacy
Ambulatory careReceiving aspirin*, History of MI and receiving beta-blocker,*, Receiving cholesterol- lowering agent*proportion of patients had an active prescription for aspirin 37.9% vs. 35.1%, RRR 7%, p = 0.440, NS; proportion of patients with MI who had an active beta- blocker prescription 22.2% vs. 33.3%, RRR - 50%, p = 0.465, NS; proportion of patients receiving a cholesterol- lowering agent (73.2 % vs. 71.0%, RRR - 15%, p = 0.512)
Frank (2004) (Frank, Litt, and Beilby 87–90)
Design: RCT
N = 10,507 patients
Implementation: 00/0000
Study Start: 03/1998
Study End: 03/1999
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory careproportion of opportunities taken for preventive activity*Reminders did not improve adherence to MMR and flu vaccinations, but there was a significant increase in tetanus immunization (1.5% vs. 2.8%, relative change 1.89, 95% CI 1.59, 2.25). and pneumococcal immunization rates (1.6% vs. 2.8%, relative change 1.70, 95% CI 1.10, 2.62). Two of 8 non-medication related preventive care recommendations were significantly improved as well.+
Franklin (2007) (Franklin et al. 279–284) Donyai (2008) (Conyai et al. 230–237) Barber (2007) (Barber, Cornford, and Klecun 271–278) Franklin (2008) (Franklin, Jakclin, and Barber 375–379) Franklin (2007) (Franklin et al. 133–139)
Design: Before-after
N = 4,803 medication orders
Implementation: 06/2003
Study Start: 00/0000
Study End: 00/0000
Automated Dispensing Machine, e- Medication administration system (e-MAR, e- TAR) e-Rx
Integrated Pharmacy
Acute care/tertiary, 28 bed surgery ward of a teaching hospital Inpatient hospital based Academicerror rate for new prescriptions*, error rate for drug administrations*, %administered <1hr (Franklin, Jacklin, and Barber 375–379), rate of pharmacist interventions (Donyai et al. 230–237), total pharmacy time taken on study wardThe prescription error rate for new orders dropped significantly after implementation of the system (3.8% vs. 2.0%, RRR 47%, p = 0.0004). Medication administration error rate also significantly declined (8.6% vs. 4.4%, RRR 49%, p = 0.0003). (Franklin, Jacklin, and Barber 375–379) Post- intervention medication timeliness was improved (%administered <1hr, 79% vs. 89%, p <0.001). (Donyai et al. 230–237) The rate of pharmacist interventions declined significantly after implementation (3.0% vs. 1.9%, AR 1.1 (95% CI 0.2,2.0). (Franklin et al. 133–139) Total pharmacy time taken on study ward increased after implementation (1h 8 min vs. 1h 38 min, p = 0.001). Pharmacists were required to endorse fewer orders (50% vs. 21%, RRR 58%, p <0.0001) and endorsed fewer orders (55% vs. 30%, RRR 45%, p <0.0001).+
Fretheim (2006) (Fretheim, Aaserud, and Oxman e216)
Fretheim (2006) (Fretheim et al. e134)
Design: RCT
N = 139 practices and 501 physicians
Implementation: 00/0000
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory carethiazides prescription rates*, rates of cardiovascular risk assessment, proportion of patients achieving treatment goal at 3 monthsPrescribing of thiazides increased in the reminders + group (11% vs. 15%, RRR 54%, p <0.001, RR 1.94 95% CI 1.49 to 2.49). The groups did not differ for cardiovascular risk assessment (RR 1.04, CI 0.60 to 1.71) or proportion that achieved treatment goal at 3 months (RR 0.98, CI 0.93 to 1.02).
Galanter (2004) (Galanter, Polikaitis, and Didomenico 270–277)
Design: Before-after
N = 620 patients
Implementation: 00/0000
Study Start: 02/2001
Study End: 03/2002
CDSS/CDS/CCDS/reminders
Integrated CDSS/CDS/CCDS/reminders
CPOE/POE system
Laboratory system
Acute care/tertiary, Academiccompliance with digoxin monitoring guidelines - synchronous alerts*, compliance with hypokalemia and hypomagnesemia treatment guidelines - synchronous alerts*, compliance with hypokalemia and hypomagnesemia treatment guidelines - asynchronous alerts*Post implementation, synchronous alerts significantly increased test ordering for digoxin levels, K levels and Mg levels at 1 hr and 24 hrs (p <0.01 for all). Supplementation of Mg at 1 hour was significantly improved, but not at 24 hrs. Supplementation of K was not improved at 1 or 24 hrs. Synchronous alerts resulted in improved compliance at 1 hr and 24 hrs for both K and Mg supplementation (p <0.01).+
Galanter (2005) (Galanter, Didomenico, and Polikaitis 269–274)
Design: Before- after
N = 410 patients
Implementation: 00/0000
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
Acute care/tertiary, Academiclikelihood of a patient receiving contraindicated medication, compliance rates- housestaff compared to other cliniciansThe likelihood of a patient receiving at least one dose of the contraindicated medication decreased from 89% to 47% after alert implementation (p <0.0001), RRR 47%. For the 226 alerts received by housestaff, the alert compliance rate was 42%; for the remaining clinicians the compliance rate was 38% (p = 0.54).+
Gerard (2008) (Gerard et al. 776–779)
Design: Time series
N = 907 orders for flu vaccination
Implementation: 00/2001
Study Start: 00/2003
Study End: 00/2007
CDSS/CDS/CCDS/reminders
Integrated Hospital information system
General Hospital 464 Bedsacceptance rate of pre- selected orders, year 1 vs. year 2, acceptance rate of pre-selected orders, year 2 vs. year 3, vaccination rate, year 1 vs. year 2, vaccination rate, year 2 vs. year 3During the intervention, physicians were significantly more likely to accept pre- selected vaccination orders, Year 1 (47%), Year 2 (77%), Year 3 (83%); however vaccine administration by nurses was suboptimal. EMR functionality improved, patient receipt of vaccine increased significantly, Year 1 [0/36; 0%], Year 2 [8/66; 12%], Year 3 [286/805; 36%].+
Gill (2009) (Gill et al. 221–226)
Design: RCT
N = 64,150 patients
Implementation: 00/0000
Study Start: 00/0000
Study End: 10/2006
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory careUp-to-date lipid test*, Lipid medication if not at goal (high risk patients only)*Outcomes improved for most measures from before to 1 year after the intervention (univariate analysis). However, after controlling for confounding variables and for clustering in multilevel modeling, only up-to-date lipid testing for high- risk patients was statistically better in the intervention group as compared to the control group (adjusted OR 15.0, p <0.05). Intervention status was NS
Gilutz (2009) (Gilutz et al. 23–29)
Design: RCT
N = 7,448 patients from 56 control and 56 intervention clinics
Implementation: 00/0000
Study Start: 01/2000
Study End: 12/2003
CDSS/CDS/CCDS/reminders
Integrated Hospital information system, Laboratory system, Pharmacy
Ambulatory care Academicrate of adequate monitoring Positive treatment trend, overall up-titration rate in patients with LDL = 110 mg/dlhigher rate of adequate monitoring documented in intervention arm (54.8% vs. 48.7%, p <0.001). Medication initiation or up- titration recommended for patients with LDL levels above 110 mg/dl results showed overall positive trends were minimally more prominent in the intervention arm (59.1% vs. 53.7%, p <0.003). This difference constitutes a higher rate of drug initiation (2.5%), up- titration (1.8%) and avoiding drug cessation (1.1%). However, overall up- titration in patients with LDL = 110 mg/dl was poor, both in the intervention arm and in the control arm (8.6% vs. 7.4%, NS).+
Ginzburg (2009) (Ginzburg et al. 2037–2041)
Design: Before-after
N = 540 Patients
Implementation: 00/0000
Study Start: 01/2005
Study End: 12/2005
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory careMedication error*Significantly more medication errors were found in the preintervention group than in the postintervention group [(32.6% (n = 103) vs. 20.5% (n = 46), p = 0.002]. Significantly fewer strength overdosing errors occurred in the postintervention group (8.9% vs. 4.0%, p = 0.028).+
Goethe (1997) (Goethe, Schwartz, and Szarek 553–558)
Design: Time series
N = 1,604 alerts
Implementation: 00/0000
Study Start: 01/1994
Study End: 12/1996
CDSS/CDS/CCDS/remindersOther specialty hospital (rehab, oncology) 130 Bedsalert rate, physician response rate to alerts*, compliance with alerts*The rate of alerts went down in the second year (29% vs. 15%, RRR 48%, p <0.001), as did the rate of physician responses to the alerts (67% vs. 55%, RRR 18%, p <0.001) change in practice to comply with guidelines occurred 28% (year 1) compared to 21% (year 2)
Griffey (2009) (Griffey 265)
Design: Time series
N = 2,419 orders
Implementation: 00/0000
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Stand-Alone, CPOE/POE system
Emergency department, AcademicDuring use of the CDSS system, agreement with recommended doses was increased.During use of the CDSS system, agreement with recommended doses was increased (23.0% for off and 31.4% for on, RRR 37%, p = 0.03) reduction similar for benzodiazepine s (p = 0.03), opiates (p = 0.04), and NSAIDS (p = 0.0009).+
Halkin (2001) (Halkin et al. 260–265)
Design: Time series
N = 775,186 prescriptions
Implementation: 11/1997 to 00/1998
Study Start: 01/1998
Study End: 06/1999
CDSS/CDS/CCDS/reminders
Integrated Pharmacy
Pharmacy, HMO pharmacyrate of drug interaction prescriptions 90% pharmacies and 50% physicians compared with baseline, rate of drug interaction prescriptions 95% pharmacies and 90% physicians compared with baselineDispensing of drug interaction prescriptions was reduced by 21.1% and by 67.5% in periods II and III compared with period I (OR, 0.79; 95% CI, 0.75 to 0.83 and OR, 0.28; 95% CI, 0.26 to 0.30, respectively).+
Hicks (2007) (Hicks et al. 429–441)
Design: RCT
N = 1,422 patients
Implementation: 00/0000
Study Start: 07/2003
Study End: 02/2005
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Other, Academicblood pressure controlled, receiving a recommended drug class medication within 1 week of the clinic visit adjustedThis study had 4 groups: usual care, CDS, NPs, and NPs+CDS. No difference was seen across all 4 groups for blood pressure readings: Usual care vs. CDS: 45% vs. 48% controlled, OR 0.96 (CI 0.78 to 1.19). Patients in the CDS group were more likely to have received a recommended drug class medication within 1 week of the clinic visit than patients in the usual care group: adjusted OR 1.32 (CI 1.09 to 1.61).+
Hollingworth (2007) (Hollingworth et al. 722–730)
Devine (2010) (Devine et al. 152–171)
Design: Cross-sectional
N = 146 health care providers (69 in phase 1 and 77 in phase 2)
Implementation: 00/2003
Study Start: 02/2005
Study End: 01/2006
e-Rx
Integrated EHR/EMR system
Ambulatory caretime spent on writing tasks (min/hr), paper vs. desktop vs. laptop, time spent on computer tasks (min/hr) paper vs. desktop vs. laptop, time spent on computer and writing tasks (min/hr), paper vs. desktop vs. laptop, More time in phase 2 compared with handwritten prescriptions for all prescriptions and new prescriptions but not for renewed prescriptions. (Devine et al. 152–171)Prescribers at e-RX sites, both desktop and wireless laptops, spent significantly less time (minutes/hour) on writing tasks that their paper- based colleagues (8.7 paper vs. 5.5 desktop vs. 5.9 laptops, p <0.05), but more time on computer based tasks (3.8 vs. 7.4. vs. 8.1, p <0.05). Overall time on writing tasks and computer tasks together were not different among the three formats (12.4 vs. 12.9. vs. 14.0, NS) and should not greatly disrupt workflow. [In the second phases (Deveine et al. 152–171) (point of care prescribing) the clinicians spend more time than handwritten prescribing for all prescriptions:
25 seconds more (99.5% CI 12 to 38), and more time for new prescriptions:
29 seconds more (CI 14 to 44) but not more time for renewed prescriptions:
13 seconds more (CI −13 to 39).
+
Hulgan (2004) (Hulgan et al. 349–357)
Design: Time series
N = 15,194 quinolone orders
Implementation: 02/2002
Study Start: 02/2001
Study End: 01/2003
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
EHR/EMR system
Acute care/tertiary, Academicchange in weekly proportion of oral quinolone orders*55.5% orders were for oral quinolones before the intervention orders compared with 62.4% after (RRR -12%, p = NR). In the time-series analysis, the intervention increased the proportion of oral quinolone orders per week by 5.6% (95% CI 2.8 to 8.4%; p <0.001).+
Hwang (2002) (Hwang, Park, and Bakken 213–223)
Design: Time series
N = 171 patients
Implementation: 10/1999
Study Start: 06/1999
Study End: 05/2000
CPOE/POE system
Integrated Hospital information system
Imaging systems
Acute care/tertiary, 1,000 plus Beds Academicnumber of daily orders per patient, number of daily medication orders, number of changed orders, number of cancelled orders, number of daily PRN ordersdaily orders per patient significantly increased following POE system introduction compared to both 3- and 6- months post (10.9 vs. 17.4 vs. 19.9, p </0.0001) similar pattern observed for number of daily medication orders (4.2 vs. 6.6 vs. 6.1, p <0.0001) and PRN orders (2.9 vs. 7.9 vs. 8.3, p <0.0001) difference between 3 and 6 months after POE was NS for either measure. The number of changed orders (2.2 vs. 0.2 vs. 0.03, NS) and cancelled orders (3.3 vs. 2.3 vs. 2.2, NS)+
Igboechi (2003) (Igboechi et al. 227–231)
Design: Before-after
N = 10,134 medication errors
Implementation: 06/2001
Study Start: 06/1999
Study End: 05/2002
CPOE/POE system
Integrated CDSS/CDS/CCDS/reminders
EHR/EMR system, Pharmacy
Acute care/tertiary, 350 Beds Inpatient hospital basedtotal potential errors*, Illegible orders, incomplete orders, incorrect orders, drug therapy problemsThe number of documented medication errors decreased postimplementation for total potential errors (p <0.001), illegible orders (p <0.001), incomplete orders, (p <0.001) and incorrect orders (p <0.001) but not for drug therapy problems (p = 0.289). Annual numbers were compared for each of the 2 years before implementation of CPOE and the year after CPOE.+
Jacques (2005) (St Jacques et al. 215–221)
Design: Before-after
N = 287 procedures
Implementation: 00/0000
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated Hospital information system
Acute care/tertiary, Academicantibiotic redosing rate*On-time antibiotic redosing increased significantly after the implementation of the computer reminder system (20% vs. 57%, RRR - 185%, p <0.001).+
Jani (2008) (Jani et al. 214–218)
Design: Before-after
N = 2,222 prescribed drugs
Implementation: 03/2006
Study Start: 07/2005
Study End: 07/2006
e-Rx
Integrated EHR/EMR system, e-MAR, Pharmacy
Pediatric stand alone hospital, Ambulatory careerror rate*, error free visit rateThe overall prescribing error rate was 77.4% (95% CI = 75.3% to 79.4%) for handwritten items and 4.8% (95% CI = 3.4% to 6.7%) with e-Rx (RRR, 94%, p <0.001). Pre-e-Rx, 1153 items (73.3%; 95% CI = 71.1% to 75.4%) were missing essential information, and 194 items (12.3%; 95% CI = 10.8% to 14%) were judged to be illegible. Post-EP, only 9 items (1.4%; 95% CI = 0.7% to 2.6%) were missing essential information, and illegibility errors were eliminated. The number of patient visits that were error- free increased from 21% to 90% (69% difference; 95% CI = 64% to 73.4%; RRR - 324%, p <0.001) after the implementation of e-Rx.+
Javitt (2008) (Javitt, Rebitzer, and Reisman 585–602)
Design: RCT
N = 39,508 patients
Implementation: 01/2001
Study Start: 01/2001
Study End: 12/2001
CDSS/CDS/CCDS/reminders
Integrated Billing/administration system
Laboratory system, Pharmacy
Ambulatory careresolution rate-add a drug alert*, resolution rate-stop a drug*, resolution rate -do a test*Resolution rate for “add a drug” CCs was 8.6 % higher in the study group than the control group (p <0.05). There was, however, no significant difference in the resolution rates for “stop a drug” CCs (change −6%, NS). Resolution rates for “do a test” CCs were 5.8% higher in the study group, p <0.05.+
Johnson (2010) (Johnson et al. 321–325)
Design: RCT
N = 3,285 patients
Implementation: 00/0000
Study Start: 04/2007
Study End: 08/2007
CDSS/CDS/CCDS/reminders e-Rx
Integrated EHR/EMR system
Ambulatory care, Pharmacy, Not specified, Academicrate of callbacks generated*There was no significant difference in the callback rates between the “SYW off” and the “SYW on “periods (0.4% vs. 0.45%; p = 0.47).
Kadmon (2009) (Kadmon, et al. 935–940)
Design: Time series
N = 5,000 Medication orders
Implementation: 11/2004
Study Start: 09/2004
Study End: 09/2007
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system, Hospital information system
Acute care/tertiary, Critical care units (CCU, ICU, NICU) 12 bed PICU unittotal prescription error rate(combination of the 3 error types)*, pADE*, rule violations*, medication prescription errors*Among the 5,000 prescriptions reviewed, 273 (5.5%) contained prescription errors. Implementation of CPOE associated with a slight, nonsignificant decrease in prescription error rate (between periods 1 and 2; 8.2% vs. 7.8%, p = 0.66). Decreases in rate of prescription errors after CDSS implementation were statistically significant between periods 2 and 3 (7.8% vs. 4.4%, p = 0.0004) and after prescription authorization between period 3 and 4 (4.4% vs. 1.4%, p <0.0001). The rate of potential ADEs decreased slightly between periods 1 and 2 (from 2.5% to 2.4%, p = 0.9) and significantly in periods 3 and 4 (to 0.8% and 0.7%, respectively; p <0.005). Rate of MPEs decreased slightly between periods 1 and 2 (from 5.5% to 5.3%, p = 0.79), but new types of MPEs appeared. A significant decrease in period 3 (to 3.8%; p <0.05) and a dramatically significant decrease in period 4 (to 0.7%; p <0.0005) was noted. 3 RVs were found between period 1 and 2 (0.002% vs. 0.001%, p = 0.3). No RVs were found in period 3 and 4.+
Kaplan (2006) (Kaplan et al. 461–467)
Design: Time series
N = n/a orders
Implementation: 04/2002
Study Start: 12/2001
Study End: 01/2004
CPOE/POE system
Integrated Formulary, Hospital information system, Imaging systems, Pharmacy
Pediatric stand alone hospital, 423 Bedsrate of verbal orders*, rate of unsigned verbal orders*Overall, there was a significant decrease in the rates of verbal orders (from 22% to 10%) and unsigned verbal orders (from 43% to 9%) between the period before CPOE implementation and 21 months after CPOE implementation (p = 0.0001 for both).+
Karson (2007) (Karson et al. 1004)
Design: Time series
N = 74,494 verbal orders
Implementation: 0/0000
Study Start: 01/2005
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
Acute care/tertiary, 900 Beds Academiccompliance with co- signing within 24 hours, compliance with co- signing by month endAt baseline, 49% of verbal orders were co- signed within 24 hours. This increased to 63% after the first intervention and 93% after the second intervention (p <0.001). At month end, the compliance rate was 61% at baseline, 94% after the first intervention and 98% after the second (p <0.001).+
Kazemi (2010) (Kazemi et al. e5)
Design: Observational study
N = 158 patients (neonates)
Implementation: 00/2007
Study Start: 12/2007
Study End: 09/2008
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated Hospital information system
Acute care/tertiary, 400 Beds Academicrate of non-intercepted errors for orders (POE Errors vs. NOE Errors)*, rate of nonintercepted errors for ordered medication*, rate of nonintercepted errors for patient days*, rate of nonintercepted errors medication-days *,The rate of nonintercepted errors for orders decreased from 22.7% to 14.5% (RR 0.64; 95% CI 0.53 to 0.77, p <0.001). For ordered medication it dropped from 12.8% to 7.6% respectively (RR 0.60; 95% CI 0.50 to 0.71, p <0.001). However, the highest rate difference (9.5%) was seen when calculated according to patient days (24.5% vs. 15%, RR 0.61; 95% CI 0.49 to 0.77; p <0.001). The rate difference for medication- days were 5.8% (14.4% vs. 8.6%, RR 0.60; 95% CI 0.49 to 0.74; p <0.001).+
Kim (2006) (Kim et al. 495–498)
Design: Before- after
N = 2,375 chemo orders
Implementation: 02/2003
Study Start: 07/2001
Study End: 02/2004
CPOE/POE system
Integrated EHR/EMR system
Other specialty hospital (rehab, oncology) Academicrate of improper dosing on treatment plans, rate of improper dosing on orders, rate of treatment plans and orders not matching, rate of missing cumulative dose calculations, rate of incorrect dosing calculationsAfter CPOE deployment, daily chemotherapy orders were less likely to have improper dosing on orders (2.3% vs. 0.1%, RRR 97%, p <0.05), incorrect dosing calculations (5.8% vs. 0.5%, RRR 91%, p <0.05), missing cumulative dose calculations (18% vs. 5.7%, RRR 68%, p <0.05), and incomplete nursing checklists (4.8% vs. 2.5%, RRR 48%, p <0.05). There was no difference in the likelihood of improper dosing on treatment plans (4.0% vs. 2.6%, RRR 35%, NS) and a higher likelihood of not matching medication orders to treatment plans (1.1% vs. 6%, RRR -445%, p <0.05).+
Kim (2008) (Kim et al. 416–421)
Design: Time series
N = not given not given
Implementation: 00/2002
Study Start: 02/2004
Study End: 04/2006
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
Acute care/tertiary, 750 Beds Academic3rd generation cephalasporin use (daily doses/1,000 patient days)The use of third generation antibiotic cephalosporin use decreased significantly from 103.2 doses/1,000 patient days to 84.9 postimplementation. It increased once the feedback element was stopped (84.9 vs. 115.1, p <0.05).+
Kirk (2005) (Kirk et al. 817–824)
Design: Observational study
N = 4,274 prescriptions
Implementation: 00/2000
Study Start: 03/2003
Study End: 08/2003
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
Acute care/tertiary, Ambulatory care Academicerror rateThe computer calculated dose error rate was 12.6% compared with the traditional prescription error rate of 28.2% (RRR 55%, p <0.001).+
Kitahata (2003) (Kitahata et al. 803–811)
Design: Time series
N = 1,204 patients with HIV
Implementation: 04/1998
Study Start: 03/1996
Study End: 09/1999
CDSS/CDS/CCDS/reminders
Integrated Hospital information system
Ambulatory care, Academicrate of prophylaxis for mycobacterium avium complex infection, rate of prophylaxis for pneumocystis cairnii pheumoniaAfter implementation of the CDSS patients were more likely to be given prophylaxis for mycobacterium avium complex infection (Hazard Ratio 3.84, CI 1.58 to 9.32, = 0.003) but not for pneumocystis cairnii pheumonia (Hazard Ratio 1.14, CI 0.84 to 1.59, NS).+
Koide (1999) (Koide et al. 11–19)
Design: Before-after
N = 1,024 prescriptions for 111 patients and 68 physicians
Implementation: 09/1994
Study Start: 09/1994
Study End: 09/1996
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system, Hospital information system, Laboratory system
Acute care/tertiary, 1,040 Beds Academicrate of ‘appropriate’ prescribing (normal value of ALT or AST within 3 mon)ths127/491 (25.9%) preintervention prescriptions were classified as ‘appropriate’. 353/533 (66.2%) postintervention prescriptions were classified as ‘appropriate’. This sudden increase in level of 40.3% occurring immediately after the start of the intervention was highly significant (p <0.0001).+
Kooij (2008) (Kooij et al. 893–898)
Design: Time series
N = 1,565 patients
Implementation: 00/0000
Study Start: 11/2005
Study End: 06/2006
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
General Hospital, Academicrate of prophylaxis, control vs. CDSS, rate of prophylaxis, CDSS vs. stopping CDSSPatients who needed PONV prophylaxis were more likely to be prescribed medication if clinicians were provided with electronic DS (73%) than before DS (38%) or after the electronic DS was stopped (37%), p <0.001.+
Kooij (2009) (Kooij et al. 187–191)
Design: Before- after
N = 5,652 patients
Implementation: 00/0000
Study Start: 11/2005
Study End: 06/2006
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Acute care/tertiary, Academicpercentage of patients who received dexamethasone*, percentage of patients who received granisetron*, percentage of patients who received both dexamethasone and granisetron*Dexamethasone was given to 46% of the control period. In the decision support period, rate increased significantly to 95% and after deactivating the automated reminders, it decreased to 47% in the post decision support period (p <0.001) For granisetron, these percentages were 53%, 81%, and 51%, respectively (p <0.001). Percentage of patients receiving both medications was 39% in the control period, increased to 79% in the decision support group and decreased to 41% in the post decision support group (p <0.001).+
Kralj (2003) (Kralj et al. 197–203)
Design: Case control
N = 11,644 patient-physician encounter
Implementation: April 2000
Study Start: 12/1999
Study End: 11/2001
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory carechanges in prescribing rates of erythropoietin between clinics at baseline compared with during the intervention groupThe mean difference in prescribing rates between experimental and control clinic at baseline was 0.36 (p = 0.044). Whereas in the intervention period the difference in the rates between them almost tripled to .093 (p = 0.000). The rate of erythropoietin prescribing increased by 14.2% (p = 0.05) at the experimental clinic. It declined by 15.9% (p = 0.12, NS) in the control clinic.+
Krall (2004) (Krall, Traunweiser, and Towery 1–9)
Design: RCT
N = 1,076 patients
Implementation: 00/1994
Study Start: 01/2000
Study End: 02/2000
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Acute care/tertiaryproportion of patients no longer eligible for alerts at the end of the month*Following implementation of the alert, more patients were ‘no longer eligible for alerts at the end of the month’ (25.8% pre vs. 54.3% post, RRR - 103%, p <0.001).+
Kucher (2005) (Kucher et al. 969–977)
Design: RCT
N = 2,506 patients
Implementation: 00/0000
Study Start: 09/2000
Study End: 01/2004
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system, Hospital information system
Acute care/tertiary Academicreceived pharmacological interventionsMore patients in the CDSS group received pharmacological interventions. (13% vs. 24%, RRR 69%, p <0.001).+
Lapane (2008) (Lapane et al. 442–446)
Design: Mixed methods
N = 276 primary care prescribers and their staff
Implementation: 00/2003
Study Start: 04/2006
Study End: 08/2006
e-RxAmbulatory careself reported drug alert overrides*22/145 prescribers (15%) reported overriding drug- allergy alerts most of the time or ‘always’ with variation in frequency of overriding drug alerts by e-Rx software system ranging from 9% to 50% (p = 0.656 for overall comparison by e-Rx software system). Nearly 1 in 4 respondents reported overriding drug–dose alerts ‘most of the time ‘or ‘always’ (range 13% to 33%; p = 0.006). More than 40% indicated they override drug– drug interactions ‘most of the time’ or ‘always’ (range, 25% to 50%; p = 0.374).
Lecumberri (2008) (Lecumberri et al. 699–704)
Design: Time series
N = 19,338 patients
Implementation: 09/2005
Study Start: 01/2005
Study End: 06/2007
CDSS/CDS/CCDS/reminders
Integrated hospital guidelines
Unspecified Hospital Academicnumber of alerts, proportion of alerted patients receiving thromboprophylaxisan electronic alert was sent to 32.8% and 32.2% of all hospitalized patients, respectively. Appropriate prophylaxis among alerted patients was ordered in 89.7% (2006) and 88.5% (in 2007) of surgical patients, and in 49.2% (in 2006) and 64.4% (in 2007) of medical patients.
Ledwich (2009) (Ledwich et al. 1505–1510)
Design: Before- after
N = 2,477 vaccine possibilities (patients)
Implementation: 00/0000
Study Start: 10/2006
Study End: 12/2007
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory care, Academicinfluenza vaccination rates, pneumococcal vaccination rates*PostBPA influenza vaccination rates significantly increased (47% to 65%; p <0.001), at both sites. PostBPA pneumococcal vaccination rates likewise significantly increased (19% to 41%; p <0.001).+
Lesprit (2009) (Lesprit et al. 1058–1063)
Design: Observational study
N = 932 prescriptions
Implementation: 11/2006
Study Start: 11/2006
Study End: 10/2007
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system, Laboratory system
Acute care/tertiary, 960 Beds Academicactual duration of treatment in days compared to prescribed*Of the 482 prescriptions requiring intervention, the physicians complied with 80.3% of the recommendations. There was a significant reduction in the actual duration of antibiotic treatment compared to the originally prescribed duration (8 to 7 days (p <0.0001).+
Lester (2005) (Lester et al. 22–29)
Design: RCT
N = 235 patients and 14 clinicians
Implementation: 07/2003
Study Start: 07/2003
Study End: 07/2004
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory care Academicproportion of patients receiving statins*, proportion of patients receiving statins at 1 yr*At 1 month, more patients in the email group had received statins than control patients (3% vs. 15%, RRR 400, p <0.001). At 1 year the difference in receipt of statins had disappeared (17% vs. 25%, NS).+
Lin (2008) (Lin et al. 620–626)
Design: Time series
N = 1,123 high severity order checks
Implementation: 00/1997
Study Start: 01/2001
Study End: 01/2006
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated CPOE/POE system
EHR/EMR system
Acute care/tertiary, General Hospital, 444 Beds Ambulatory care, Long term care (nursing homes)override rates-severe drug-drug alerts*, override rates-severe drug-allergy alerts*There were 215 high severity order checks in 2001 (0.5% of orders) and 908 in 2006 (2.5% of orders). Rate of overrides for drug-drug checks remained the same between 2001 and 2006 (88% vs. 87%, NS). Rate of overrides for drug-allergy order checks increased significantly from 2001 to 2006 (69% vs. 81%, RRR - 17%, p <0.005).
Linder (2009) (Linder et al. 231–240)
Design: RCT
N = 111,820 patients
Implementation: 00/0000
Study Start: 11/2005
Study End: 05/2006
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system, Imaging systems, Laboratory system
Ambulatory careRate of antibiotic prescribing to patients with ARI *In the intent-to- intervene analysis, clinicians prescribed antibiotics to 43% of patients with ARI diagnoses in control clinic compared to 39% in the intervention clinic (OR. 0.8; 95% CI 0.6 to 1.2; p = 0.30). The ARI Smart Form did not significantly reduce overall antibiotic prescribing, was used by 33% of intervention clinicians (86/262) at least once. Appropriate antibiotic prescribing rate was 88% (n = 990 visits) in the as-used analysis.
Liu (2008) (Liu et al. 1109–1112)
Design: Time series
N = 858 patients
Implementation: 00/1989
Study Start: 01/2005
Study End: 12/2006
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
EHR/EMR system
Acute care/tertiarypercentage of no prophylactic antibiotic after clean surgery, mean number of days of antibiotic treatmentIn clean procedures, the percentage of no prophylactic antibiotic after surgery increased in the long run (overall 76% vs. 84% vs. 93%, no analysis); the increase was significant for 2 of the 4 surgery types (p <0.005). In clean-contaminated procedures, the duration of prophylactic antibiotic after surgery (mean number of days) was significantly reduced in 2 of the 3 surgery types (p <0.001).+
Madaras-Kelly (2006) (Madaras-Kelly et al. 155–169)
Design: Time series
N = not reported
Implementation: 00/0000
Study Start: 07/2001
Study End: 06/2004
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated Hospital information system
Acute care/tertiary, 87 Bedsuse of antibiotics*Use of aminopenicillin beta-lactam inhibitors, all fluoroquinolones and levofloxacin decreased while use of first-generation cephalosporins and trimethoprim sulfamethoxazole increased (p <0.05 for each).+
Mahoney (2007) (Mahoney et al. 1969–1977)
Design: Before-after
N = 2,843,135 inpatient medication orders
Implementation: 02/2002
Study Start: 02/2002
Study End: 06/2006
CDSS/CDS/CCDS/reminders
CPOE/POE system Pharmacy information system
Integrated EHR/EMR system, Hospital information system
General Hospital, Pediatric stand alone hospital, 966 in 2 hospitals Beds Pharmacy, Inpatient hospital based, Academicrate of:
-

drug allergy violations*,

-

excessive doses*,

-

incomplete or unclear orders*,

-

therapeutic duplication*

Medication errors decreased after implementation of the CIT with respect to drug allergy violations (OR 0.14, 95% CI 0.11 to 0.17, p <0.001), excessive doses (OR 0.68, 95% CI 0.62 to 0.74, p <0.001) and incomplete or unclear orders (0.35, 95% CI 0.32 to 0.38, p <0.001), but no decease in therapeutic duplications. Turnaround time between drug ordering and administration decreased from 90 minutes to 11 minutes. The override rate also decreased (7.1 to 2.9%, RRR 59%, p = 0.001).+
Martens (2007) (Martens et al. S403–S416)
Design: RCT
N = 77 physicians (GPs)
Implementation: 04/2004
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated, EHR/EMR system
Ambulatory carequinolone prescriptions, inhaled corticosteroids for newly diagnosed COPD in patients >40 years, first choice drugs for sore throats GPs got either reminders on antibiotics, asthma, and COPD or cholesterol. Reminders were either to stop prescribing drugs or to prescribe a specific first-line drug.No differences were seen for either group to prescribe a drug or in the cholesterol reminder group. GPs in the antibiotics, asthma and COPD group showed changes in 3 of 8 drug categories. Outcome measures were sum scores for drug volume: lower scores were improvements in prescribing. Reminder physicians prescribed fewer quinolones (4.6 (95% CI 2.8 to 8.1) vs. 1.5 (95% CI 0.8 to 2.2); fewer inhaled corticosteroids for newly diagnosed COPD in patients >40 yr (0.5 (95% CI 0.3 to 0.9) vs. 0.0 (95% CI 0 to 0.1), p = 0.00); and better first choice drugs for sore throats (0.8 (95% CI 0.3 to 2.4) vs. 0.2 (95% CI 0.0 to 9.4), p = 0.03.
Mattison (2010) (Mattison et al. 1331–1336)
Design: Before-after
N = not reported medication orders
Implementation: 10/2004
Study Start: 06/2004
Study End: 08/2008
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
Acute care/tertiary, 621 Beds Academicrate of prescribing not-recommended medications*, rate of prescribing medications with recommended dosage reductions*In before-and-after comparisons, the mean (SE) rate of prescribing not-recommended medications dropped from 11.56 (0.36) to 9.94 (0.12) orders per day (difference, 1.62 [0.33]; p <0.001). There were no appreciable changes in the rate of ordering medications for which only dose reduction was recommended or that were not targeted after CPOE implementation.+
Maynard (2010) (Maynard et al. 10–18)
Design: Time series
N = 3,285 patients
Implementation: 04/2006
Study Start: 00/2005
Study End: 00/2007
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
Acute care/tertiary, 350 Beds Academicpercent of patients on adequate prophylaxis*The percent of patients on adequate prophylaxis improved in each of the 3 years from a baseline of 58% in 2005 to 78% in 2006 (unadjusted relative benefit = 1.35; 95% CI 1.28 to 1.43), and 93% in 2007 (unadjusted relative benefit =1.61; 95% CI 1.52 to 1.69).+
McCluggage (2010) (McCluggage et al. 70–75)
Design: Before-after
N = 522 patients
Implementation: 02/2007
Study Start: 08/2006
Study End: 04/2007
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
EHR/EMR system
Acute care/tertiary, Academicoptimal regimen prescribed*The percentage of patients whose initial vancomycin regimen matched the nomogram recommendation was higher in the postimplementation group compared with the preimplementation group (35.8% vs. 23.7%, p = 0.0028).+
McGregor (2006) (McGregor et al. 378–384)
Design: RCT
N = 4,507 patients
Implementation: 00/000
Study Start: 05/2004
Study End: 08/2004
CDSS/CDS/CCDS/reminders
Integrated Laboratory system, Pharmacy
Acute care/tertiary 648 Beds Inpatient hospital based, Academicmean time spent on antimicrobial managementTeam members spent 3.2 hours per day on management of antimicrobials with the decision support system compared with 4 hours per day without. No statistical testing was done.+
McMullin (1999) (McMullin et al. 2077–2082)
Design: Before-after
N = 265 patients
Implementation: 01/1996
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system, Laboratory system, Pharmacy
Acute care/tertiary, Pharmacy, Inpatient hospital basedrate of concomitant orders for contraindicated medications with cisapride*The rate of ordering contraindicated drugs with cisapride was reduced with COPE (9% vs. 3.1%, RRR 65%, p <0.001).+
Miskulin (2009) (Miskulin et al. 1081–1088)
Design: Cohort study
N = 8,941 patients
Implementation: 00/2005
Study Start: 11/2005
Study End: 04/2006
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory careEPO use*, time spent on anemia management (hours per month)After adjustment for center and baseline differences, the log weekly EPO dose in patients treated using CDS was 4% less than those dosed manually (RR 0.96; 95% CI, 0.77 to 1.18, NS). CDS was associated with a nearly 50% decrease (p <0.001) in the time spent by dialysis unit staff on anemia management. Units using the computerized protocol spent a median of 3 hours per month on anemia management units using manual dosing spent a median of 6.5 hours per month.+
Montgomery (2000) (Montgomery et al. 686–690)
Design: RCT
N = 552 patients
Implementation: 00/0000
Study Start: 09/1996
Study End: 09/1998
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory careprobability of patients taking 2 drugs, probability of patients taking 3 drugsrisk chart group alone compared to computer support group had a lower probability of patients taking 2 drugs (OR 0.5, 95% CI 0.2 to 0.9) p <0.05) or 3 drugs (OR 0.3, 95% CI 0.1 to 0.6, p <0.05).+
Morrison (2006) (Morrison et al. 1033–1039)
Design: Time series
N = 3,864 patients
Implementation: 00/0000
Study Start: 02/2001
Study End: 02/2003
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
EHR/EMR system
Acute care/tertiary, 1171 Beds Academicmeperidine prescription rate, rate of patients receiving a concomitant laxative with an opioidrate of patients receiving a concomitant laxative with an opioid did not change with the introduction of the CDSS system (data for 5 groups, 24.7% of patients who needed a laxative, 27.8%, 32.1%, 26.8%, and 34.0%, all comparisons NS). Fewer patients received meperidine with the introduction of the CDSS system. For group 4 (CDSS and enhanced assessment compared with Group 1 control 44.2% vs. 25.4%, RRR 20%, p <0.05). For Group 5 vs. Group 1 the rate of meperidine use was even lower (44.4% vs. 11.9%, RRR 73%, p = 0.01).
Mullett (2001) (Mullett et al. e75)
Design: Before-after
N = 1,758 patients
Implementation: 02/1999
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated Hospital information system
Critical care units (CCU, ICU, NICU) Pediatric stand alone hospital, 232 Beds Inpatient hospital based, Academicper patient anti-infective doses, per patient number of anti-infectives, per patient anti-infective orders per course, mean subtherapeutic anti-infective days/100 patient days, mean excessive dosage anti-infective days/100 patient daysThe rate of per person use of anti-infective agents did not differ for PICU doses (12.8 vs. 13.4, NS), PICU number of doses (1.85 vs. 1.97, NS) but did differ for PICU anti-infective orders per patient-anti-infective course (1.56 vs. 1.38, p <0.01). The mean number of subtherapeutic risk days decreased (7.35 vs. 4.7, p <0.001) as did the mean excessive dosage risk days (8.45 vs. 6.1, p <0.001).+
Nash (2005) (Nash et al. 64–69)
Design: Time series
N = 39,440 doses
Implementation: 00/0000
Study Start: 00/0000
Study End: 00/0000
Medication safety reporting system
Integrated Hospital information system
Laboratory system
Acute care/tertiary, 1171 Beds Academicreduction in excessive dosing for the nursing intervention, reduction in excessive dosing for the pharmacist interventionThere was a reduction in the rate of excessive dosing in the participating units compared to the control unit in the nurse intervention (23% for baseline for the control group with 17% for the nurse intervention and 17% for the pharmacist interventions (p <0.05).+
Newby (2003) (Newby, Fryer, and Henry 210–213)
Design: Cross-sectional
N = 1,667 prescriptions
Implementation: 00/0000
Study Start: 10/0000
Study End: 11/0000
e-Rx
Integrated Pharmacy
Pharmacy, Stand alone (e.g. family run)rate of repeat orderingThe rate of repeat ordering was higher for all antibiotics if the original was written using an e-Rx system (40% for paper vs. 69% for initial e-Rx, adjusted OR 3.82, 95% CI 2.55 to 5.72, p <0.05). This same significant affect was seen for all 4 study antibiotics. The rate of filling prescriptions NS as reported by patients (69% if the prescription was on paper vs. 61% by e-Rx, NS).+
Niemi (2009) (Niemi et al. 389–397)
Design: Before-after
N = 5,076 patients
Implementation: 00/0000
Study Start: 10/2006
Study End: 03/2007
CDSS/CDS/CCDS/reminders
Integrated Billing/administration system, Imaging systems, Laboratory system, Pharmacy
Acute care/tertiaryantibiotic administration within four hours*, pneumonia vaccination status documentation*, appropriate pneumonia antibiotic selection*, ACE or ARB initiation*, provision of discharge instructions to patients*Compliance with the medication related indicators for pneumonia measures were NS; antibiotic administration within four hours (83% vs. 87%, RRR - 5%), pneumonia vaccination status documentation (82% vs. 92%, RRR-12%), appropriate pneumonia antibiotic selection (93% vs. 92%, RRR 1%). After implementation, heart failure medication related quality indicators measures were not significantly for ACE or ARB initiation (95% vs. 98%, RRR - 4%) but there was a significant increase in compliance with the provision of discharge instructions to patients (84% vs. 95%, RRR - 13%, p <0.01).
Niiranen (2008) (Niiranen and Yli- Hietanen 4330–4332)
Design: Time series
N = 18,818 patient followups
Implementation: 03/2005
Study Start: 04/2005
Study End: 12/2007
CDSS/CDS/CCDS/reminders
Integrated Laboratory system
Ambulatory care, Homeproportion of patient followups assigned by nurses, year 1 to 2, proportion of patient followups assigned by nurses, year 2 to 3In general, the share of patient followups assigned by nurses was similar in year 1 and 2 (56.7% vs. 55.1%, RRR 3%, NS), and increased significantly between year 2 and 3 (55.1% vs. 58.7%, RRR -7%, p <0.001).+
Novis (2010) (Novis et al. 648–654)
Design: Before-after
N = 800 patients
Implementation: 08/2007
Study Start: 03/2007
Study End: 03/2008
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Acute care/tertiarypercentage of patients receiving pharmacological prophylaxis, percentage of patients receiving sequential compression devices and pharmaprophy-laxisThe proportion of patients receiving the recommended pharmacological prophylaxis preoperatively more than doubled (14% to 36%, p <0.001) Overall, the percentage of at-risk patients receiving the recommended combined DVT prophylaxis of SCD and pharmacological prophylaxis increased nearly seven- fold (5% to 32%, p <0.001).+
Oliven (2005) (Oliven et al. 377–386)
Design: Cross-sectional
N = 1,350 patients
Implementation: 00/0000
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated Drug order database
EHR/EMR system, Hospital information system, Laboratory system
Acute care/tertiary, 88 Beds AcademicType 1 PEs per 100 hospitalization days, Type 2 PEs per 100 hospitalization daysThe incidence of Type 1 PEs was 5.21 and 1.36 per 100 hospitalization days in the HW dept and CDOE dept, respectively (p <0.0001). Type 2 PEs were more common, 7.20 and 3.02 per 100 hospitalization days in the HWdept and CDOEdept, respectively (p <0.0001), and about 75% of them were due to few drug laboratory interactions.+
Overhage (1996) (Overhage, Tierney, and McDonald 1551–1556)
Design: RCT
N = 24 practice teams
Implementation: 10/1991
Study Start: 10/1992
Study End: 03/1993
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
EHR/EMR system, Hospital information system, Laboratory system, Pharmacy
General Hospital Academicrates of compliance with preventive care recommend-ations*control teams complied with 24% of the reminders compared with 23% for intervention teams (p = 0.78) When preventive care measures were analyzed individually, 2 significant differences were seen in compliance (24-hour urine protein and angiotensin- converting enzyme [ACE] inhibitor) between control and intervention teams. Assumed to be due to chance with multiple testing and because they were in the opposite directions.
Overhage (1997) (Overhage et al. 364–375)
Design: RCT
N = 86 physicians on 6 services (services randomized)
Implementation: 00/0000
Study Start: 10/1992
Study End: 04/1994
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
EHR/EMR system, Laboratory system
General Hospital, Academicimmediate compliance with corollary ordering*, 24 hour compliance*, hospital stay compliance*Intervention physicians ordered the corollary orders required by the guidelines twice as often as control physicians did when measured by immediate compliance (46.3% vs. 21.9%, RRR - 111%, p <0.0001). Significant differences between study and control physicians also appear in 24 hour compliance (50.4% vs. 29.0%, RRR - 74%, p <0.0001) and hospital-stay compliance (55.9% vs. 37.1%, RRR 51%, p <0.0001).+
Overhage (2001) (Overhage et al. 361–369)
Design: RCT
N = 34 physicians
Implementation: 00/1984
Study Start: 09/1996
Study End: 02/1998
CPOE/POE system
Integrated Billing/administration system
CPOE/POE system, Imaging systems
Ambulatory care, Academicmean time spent in direct care per patient, minutes*, mean time spent in writing tasks per patient, minutes*Time spent in direct care with a patient in minutes remained the same in the control (paper- based) and CPOE groups (15.8 vs. 16.1, NS). Time spent on writing tasks in minutes remained the same between groups (6.2 vs. 6.9, NS).+
Ozdas (2006) (Ozdas et al. 188–196)
Design: Before-after
N = 540 patients
Implementation: 04/2003
Study Start: 08/2002
Study End: 09/2003
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated CDSS/CDS/CCDS/reminders
CPOE/POE system
EHR/EMR system
Acute care/tertiary, 630 Beds Academicrate of order set use for sensitive to AMI patients*, rate of order set use for confirmed AMI patients*There was a significant increase in ACS order set use after the implementation of the Admission Advisor for ‘sensitive to AMI’ admissions (60% vs. 70%, RRR -17%, p = 0.009), and a non-significant increase for confirmed AMI patients ( 46% vs. 64%, RRR - 39%, p = 0.07). For all suspected AMI admissions, ACS order set use yielded a significant increase in early aspirin ordering (77% vs. 91.2%, RRR -17%, p = 0.001) and an increase in trend toward significance in beta-blocker ordering (70% vs. 76%, RRR - 9%, p = 0.07). A similar non- significant trend in aspirin (89% vs. 97%, RRR - 9%, p = 0.07) and beta- blocker (81% vs. 88%, RRR - 9%, p = 0.18) ordering behavior associated with a confirmed diagnosis of AMI.
Palen (2006) (Palen et al. 389–395)
Design: RCT
N = 26,586 index dispensings
Implementation: 00/0000
Study Start: 11/2002
Study End: 10/2003
CDSS/CDS/CCDS/reminders
Integrated CDSS/CDS/CCDS/reminders
CPOE/POE system
EHR/EMR system, Pharmacy
Ambulatory carecompliance rateDifference between the control and intervention group physicians in the overall rate of compliance with ordering the recommended laboratory monitoring for prescribed study medications (NS). Laboratory monitoring was performed as recommended 56.6% of the time in the intervention group compared with 57.1% of the time in the control group (p = 0.31). Improved compliance favored the intervention group (71.2% vs. 62.3% [p = 0.003] for gemfibrozil; 75.7% vs. 73.9% [p = 0.05] for statins, 52.8% vs. 46% for colchicine [p = 0.05]; 42.9% vs. 0% for methotrexate [p = 0.03]).
Paterno (2009) (Paterno et al. 40–46)
Design: Cohort study
N = 71,350 alerts
Implementation: 00/1996
Study Start: 02/2004
Study End: 02/2005
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
Acute care/tertiary, 1633 beds in 2 hospitals Academiccompliance rate with DDI alerts:
-

overall,

-

severe alerts,

-

moderately severe alerts

71,350 alerts were reviewed, of which 39,474 occurred at the non-tiered site and 31,876 at the tiered site. Compliance with DDI alerts was significantly higher at the site with tiered DDI alerts compared to the non-tiered site (29% vs. 10%, p <0.001). At the tiered site, 100% of the most severe alerts were accepted, vs. only 34% at the non-tiered site (p <0.001); moderately severe alerts were also more likely to be accepted at the tiered site (29% vs. 10%, p <0.001).+
Patterson (1998) (Patterson 573–576)
Design: Before- after
N = 2,013 Patients (cases)
Implementation: 00/0000
Study Start: 00/0000
Study End: 01/1998
CDSS/CDS/CCDS/reminders
Integrated Hospital information system
Acute care/tertiary, 520 Beds Academicrate of DVT prophylaxis *The preintervention rate of deep vein thrombosis (DVT) prophylaxis was 85.2%. With the introduction of the computerized reminder, compliance with DVT prophylaxis increased to 99.3% (85.2% vs. 99.3%, p <0.001).+
Paul (2006) (Paul et al. 1238–1245)
Design: RCT
N = 3,529 patients in the RCT and 1,203 in the cohort study
Implementation: 00/0000
Study Start: 05/2004
Study End: 11/2004
CDSS/CDS/CCDS/remindersAcute care/tertiary, 424 Beds Academicappropriate antibiotic prescribing increasedAppropriate antibiotic prescribing increased for both intention to treat analyzes (64.5% vs. 72.7%, RRR 13%, p <0,05) and for per protocol analyzes (64.5% vs. 85.1%, RRR 32%, p <0.05). The cohort study showed similar increases in improved prescribing (57% vs. 70%, p <0.001).+
Peterson (2005) (Peterson et al. 802–807)
Design: Cohort study
N = 7,456 Medication orders
Implementation: 00/0000
Study Start: 10/2001
Study End: 05/2002
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
EHR/EMR system
Acute care/tertiary, Critical care units (CCU, ICU, NICU) 720 Beds Academicagreement with system recommended daily dose of psychotropic drugs for control vs. CPOE, incidence of dosing that was 10-fold greater than recommended for control vs. CDSSThe CDSS increased the prescription of the recommended daily dose (29% vs. 19%; RRR 58% p <0.001) reduced the incidence of dosing that was 10-fold greater than recommended (2.8% vs. 5.0%, RRR 48%; p <0.001).+
Peterson (2007) (Peterson et al. 2–40)
Design: RCT
N = 9,111 medication orders
Implementation: 00/0000
Study Start: 12/2005
Study End: 08/2006
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
EHR/EMR system
Acute care/tertiary
Critical care units (CCU, ICU, NICU)
Emergency department, Not specified Academic
ratio between prescribed and recommended dosesRatio between the prescribed dose and recommended dose showed that compared to controls the intervention group (reminders) received lower doses (3.0 vs. 2.5, p <0.001).+
Prescription in Ischaemic Stroke Management (PRISM) Study Group (2003) (Weir) et al. 143–153)
Design: RCT
N = 1,640 Patients
Implementation: 00/0000
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated Hospital information system
Unspecified Hospitalrelative risk reduction (RRR) in ischemic and hemorrhagic vascular eventsActual therapy prescribed vs. the option of ‘no antiplatelet or anticoagulant therapy. Estimated RRR(%) for the control and intervention in the first phase was 16.7 (13.2 to 23.7) vs. 16.3 (15.2 to 21.2) (not significantly different) For the second phase it was 16.3 (13.1 to 23.8) vs. 16.7 (13.5 to 22.9) (NS).
Quinn (2008) (Quinn et al. 160–168)
Design: RCT
N = 30 patients
Implementation: 00/0000
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Daibetes Management Tool
Integrated Web-based data analytics and therapy optimization tools
Ambulatory carechanges in medication (medication intensified)Patients using WDS were more likely to have physicians intensify diabetes medications (84.6% vs. 23.08%, p = 0.002).+
Quinzler (2009) (Quinzler et al. 30–35)
Design: Before- after
N = 20,031 prescribed drugs
Implementation: 00/2003
Study Start: 08/2006
Study End: 03/2007
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated CDSS/CDS/CCDS/reminders
Pharmacy
Acute care/tertiary 1680 Beds Academicproportion of prescriptions with inappropriate tablet splittingThe CDSS alert resulted in a significant reduction in prescriptions for inappropriate tablet splitting (2.7% vs. 1.4%, RRR 48%, p <0.001).+
Raebel (2005) (Raebel et al. 2395–2401)
Design: RCT
N = 10,169 dispensings
Implementation: 00/0000
Study Start: 09/2002
Study End: 12/2003
CDSS/CDS/CCDS/reminders
Integrated Laboratory system, Pharmacy
Ambulatory care, HMO pharmacypercentage of dispensings with baseline monitoringRecommended laboratory monitoring was completed in 74.7% (n=7,598) of dispensings at initiation of therapy. Compared to the usual care group, monitoring was higher in the intervention group (70% vs. 79%, RRR - 13%, p <0.001).+
Raebel (2007) (Raebel et al. 977–985)
Design: RCT
N = 59,680 patients
Implementation: 00/0000
Study Start: 05/2005
Study End: 05/2006
CDSS/CDS/CCDS/reminders
Pharmacy information system
Integrated EHR/EMR system
Ambulatory care, HMO pharmacynew dispensings of targeted medications*, dispensings of targeted medications considered inappropriate*In the analysis of all dispensings of targeted medications, there was a significant reduction of new dispensings of at least one targeted medication (2.2% vs. 1.8%, RRR 16%, p <0.002). For dispensings of targeted medications considered inappropriate, there was also a significant reduction with the use of the alerting system (1.5% vs. 1.1%, RRR 27%, p <0.001).+
Raebel (2007) (Raebel et al. 440–450)
Design: RCT
N = 11,100 women
Implementation: 00/0000
Study Start: 01/2003
Study End: 04/2003
CDSS/CDS/CCDS/reminders
Integrated Hospital information system
Pharmacy
Ambulatory care HMO pharmacythe proportion of pregnant women dispensed a category D or X medication*, the total number of first dispensings of targeted medicationsThe alerts resulted in a 47% reduction in the proportion of pregnant patients receiving category D or X drugs (p <0.001) Intervention patients received 238 dispensings of unique targeted medications and usual care patients received 361 dispensings (p = 0.03). The study was stopped primarily due to 2 false-positive alert types: Misidentification of medications as contraindicated in pregnancy by the pharmacy information system and misidentification of pregnancy related to delayed transfer of diagnosis information.+
Rasmussen (2005) (Rasmussen et al. 1137–1142)
Design: RCT
N = 253 patients
Implementation: 00/0000
Study Start: 00/2001
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated Internet based electronic diary
Ambulatory Care Academicgood compliance (use of medication always or almost always) Internet vs. specialist groupA significant improvement in compliance was observed for all groups, but good compliance was significantly higher (p <0.001) for both the Internet vs. the GP group and the specialist vs. the GP group. 4 of 4 measures of improve prescribing was noted in the internet group and the specialist group. The GP group also improved but to a lesser extent.+
Riggio (2009) (Riggio et al. 124–131)
Design: Before- after
N = 100 patients with heparin induced thrombocytopenia
Implementation: 06/2005
Study Start: 03/2004
Study End: 09/2006
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated Hospital information system
Acute care/tertiary, 728 Beds Academictime from platelet count criterion until heparin/enoxaparin stop* Time from platelet count criterion until 1st HIT laboratory test drawn* Time from platelet count criterion until direct thrombin inhibitor started*Counter to expectations, the time (in days) taken from alert to heparin stop order was significantly higher after implementation (1.3 vs. 2.9, p = 0.04). There were no significant differences in time (in days) from alert to lab test (2.3 vs. 3.0, NS), nor time to start of treatment with direct thrombin inhibitor (19.3 vs. 15.0, NS).
Riggio (2009) (Riggio et al. 1719–1726)
Design: Before- after
N = 2,151 discharge measures
Implementation: 00/2001
Study Start: 07/2005
Study End: 03/2008
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system, Hospital information system
Acute care/tertiary, 690 Beds Academicoverall compliance rate*CDSS yielded a 26% increase in overall compliance with the cardiac discharge measures, from 76.8% in the preintervention period to 96.8% (p <0.001) in the postintervention period.+
Rohrig (2008) (Rohrig et al. 63–68)
Design: Before- after
N = 156 patients
Implementation: 00/1999
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
EHR/EMR system
Critical care units (CCU, ICU, NICU) 14 bed unit Beds Academicrate of adequate treatment, rate of inadequate treatmentThe frequency of adequate treatment increased from an average 47.8% in the pre-period to 66.5% in the post-period (p <0.01). Rate of inadequate treatment decreased from 34.2% to 18.5%.+
Rollman (2002) (Rollman et al. 493–503)
Design: RCT
N = 200 Patients with documented major depression
Implementation: 00/0000
Study Start: 04/1997
Study End: 12/1998
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory careantidepressant prescribing rate (secondary)Prescribing antidepressants (continuous use of change in prescriptions) did not differ across the 3 groups at 3 or 6 months.
Rood (2005) (Rood et al. 172–180)
Design: RCT
N = 484 patients
Implementation: 04/2001
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated Hospital information system
Critical care units (CCU, ICU, NICU) 18 Beds Academicadherence to glucose measurement timing recommendations*, adherence to insulin dose advice*Rate of compliance with glucose measurement timing recommendations control- intervention- control (29% vs. 38% vs. 41% with period 2 and 3 greater than period 1, p = 0.05). During the intervention period, the rate for computerized group was higher than the control (36% vs. 40%, p = 0.05). Rate of compliance with insulin dose advice was higher in period 2 than 1, and then decreased significantly in period 3 (56% vs. 70% vs. 42%, p = 0.05). During the intervention period the rate for computerized group was higher than the control (64% vs. 77%, p = 0.05).+
Ross (2005) (Ross et al. 410–415)
Design: Cohort study
N = 190 physicians
Implementation: 00/0000
Study Start: 08/2001
Study End: 07/2002
e-Rx
Integrated Formulary
HMO pharmacyFormulary compliance ratio*, Absolute generic utilization ratio,* Adjusted generic utilization ratio *No differences between predominantly traditional prescribers and e-prescribers for formulary compliance (82.8% vs. 83.2%, p = 0.32) or absolute generic drug utilization (36.9% vs. 37.3%, p = 0.18) or adjusted generic drug utilization (74.3% vs. 74.7%, p = 0.27).
Rosser (1992) (Rosser et al. 911–917)
Design: RCT
N = 8,069 patients
Implementation: 00/0000
Study Start: 04/1985
Study End: 03/1986
CDSS/CDS/CCDS/remindersAmbulatory care, Academicrate of tetanus toxoid vaccination*The rates of tetanus toxin given were 3.2% in control, 22.8% in physician reminder, 24% in telephone reminder, and 30.6% in the letter reminder. The differences in the recorded vaccination rate between the randomized control group and the three reminder groups are as follows: 19.6% in the physician reminder group (95% CI 17.1% to 22.2%, p <0.00001), 20.8% in the telephone reminder group (CI 18.3% to 23.5%, p <0.00001) and 27.4% in the letter group (CI 24.8% to 30.2%, p <0.00001).+
Rubin (2006) (Rubin et al. 627–634)
Design: Observational study
N = 99 primary care physicians
Implementation: 01/2002
Study Start: 01/2002
Study End: 03/2004
CDSS/CDS/CCDS/reminders
Integrated Handheld, Stand- Alone
Ambulatory carechange in rate of antibiotic prescribing according to recommendations*, change in rate of adherence to NOT prescribe antibiotic recommendations*Adherence with CDSS recommendations increased from 79.3% in the first one- third of provider’s cases to 82.0% in the second two-thirds (an increase of 2.7%; p <0.016). Total adherence was higher with diagnoses for which an antibiotic was not indicated (84.8% vs. 75.7% for diagnoses warranting antibiotics), and providers showed a significant improvement in adherence over time for cases not requiring antibiotics (an increase of 2.7%; p <0.039).+
Safran (1995) (Safran et al. 341–346)
Safran (1993) (Safran et al. 224–228)
Design: RCT
N = 349 patients with HIV
Implementation: 00/0000
Study Start: 05/1992
Study End: 09/1993
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated EHR/EMR system
Ambulatory care, Academicmean response time to alerts* mean response times to reminders*Physicians who got alerts responded more quickly to them (mean 52 vs. 11 days, p <0.0001). Physicians who got reminders responded more quickly to them (mean 500 vs. 114 days, p = 0.0001).+
Schnipper (2008) (Schnipper et al. Symposium)
Design: Before- after
N = 30 clinicians
Implementation: 00/0000
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory careAntiplatelet prescribed or contraindication documented*, Beta- blocker prescribed *, Change in diabetic therapy if A1c >7.0 *Antiplatelet prescribed or contraindication documented improved from 3.2% in the preintervention to 31.0% in the postintervention period, p <0.001. Beta- blocker prescribed or contraindication documented was 4.2 % in the preintervention compared to 66.7% in the post period, p = 0.03. Change in diabetic therapy if A1c >7.0 was 10.7% in the pre-period and 16.9% in the post period, p = 0.11.+
Scotton (2009) (Scotton et al. 71–76)
Design: Before- after
N = 283 patients
Implementation: 12/2003
Study Start: 03/2003
Study End: 01/2006
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Acute care/tertiary, 606 Bedsproportion of cases with guideline violationsContrary to expectations, the prescribing guidelines were violated significantly less frequently during baseline (27.4%) than after implementation of the reminder (34.3%), p <0.01.
Segarra-Newnham (2003) (Segarra-Newnham 758–762)
Design: Before-after
N = 211 Patients
Implementation: 00/1997
Study Start: 00/1995
Study End: 07/2001
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory careVaccination rate for pneumococcal vaccine*, Vaccination rate for tetanus vaccine*Vaccination rates for enrolled before 1997 and after 1997 were 100% vs. 97% for pneumococcal vaccine (NS). However the vaccination rate for the same time period for tetanus vaccine was 100% vs. 61% due to national shortage of vaccine after 1997 (p <0.001).
Sellier (2009) (Sellier et al. 203–210)
Design: Time series
N = 942 prescriptions
Implementation: 00/0000
Study Start: 08/2006
Study End: 08/2007
CDSS/CDS/CCDS/reminders
Integrated Laboratory system, Pharmacy
Acute care/tertiary, 827 Beds Academicrate of inappropriate first prescriptions*, overall rate of inappropriate prescriptions, rate of cancellation of prescriptions if no eGFR lab result was availableThe rate of inappropriate first prescriptions did not differ significantly between intervention and control periods (19.9% vs. 21.3%, RRR 7%, p = 0.63); nor did the overall rate of inappropriate prescriptions (20.4% vs. 18.5%, RRR 9%, p = 0.37). The rate of cancellation of prescriptions if no eGFR lab result was available also did not differ between control and intervention periods (31.3% vs. 35%, RRR-12%, p = 0.62).
Shiffman (2000) (Shiffman et al. 767–773)
Design: Before-after
N = 9 physicians
Implementation: 00/0000
Study Start: 09/1996
Study End: 10/1998
CDSS/CDS/CCDS/reminders
Handheld
Ambulatory careAdherence rate with metered-dose inhaler/nebulization*, rate of systemic corticosteroid prescriptions*Adherence with metered-dose inhaler/nebulization rates did not differ between control and intervention (73% vs. 91%, NS), nor did rate of prescribing systemic corticosteroids (43% vs. 57%, NS).
Shojania (1998) (Shojania et al. 554–562)
Design: RCT
N = 396 physicians
Implementation: 00/0000
Study Start: 06/1996
Study End: 03/1997
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
EHR/EMR system, Imaging systems, Laboratory system, Pharmacy
Acute care/tertiary, 720 Beds Academicnumber of vancomycin orders/prescriber*, mean duration of treatment prescribed per physician*, mean number of days of vancomycin per course of treatment*The total number of orders for vancomycin for physicians in the control group was higher than in the intervention group (16.7 vs. 11.3 orders per physician, p = 0.04). Physicians in the intervention group prescribed vancomycin for 36% fewer days than physicians in the control group (26.5 vs. 41.2, p = 0.05). The number of days of vancomycin per course of treatment was also lower for the physicians in the intervention group, mean of 1.8 vs. 2.0 for the control group (p = 0.05).+
Shu (2001) (Bates et al. 965)
Design: Before-after
N = 44 Physicians (Interns)
Implementation: 11/01998
Study Start: 09/1998
Study End: 06/1999
CPOE/POE system
Integrated Hospital information system
Acute care/tertiary, 820 Bedstime spent ordering*The percentage of total time spent writing orders by medical interns between pre- CPOE and post-CPOE period increased from 2.1% to 9.0% (p <0.001).
Shulman (2005) (Shulman et al. R516–R521)
Design: Time series
N = 3,465 prescriptions over 4 times points
Implementation: 04/2002
Study Start: 09/2001
Study End: 12/2002
CPOE/POE system
Integrated Hospital information system
Critical care units (CCU, ICU, NICU) 22 (in the ICU) Beds Academicrate of ME*The proportion of MEs before CPOE was 6.7% and 4.8% after CPOE introduction (RRR 28%, p <0.04) The proportion of MEs with CPOE varied over time after its introduction (p <0.001). Evidence also indicated the strong linear trend of a declining proportion of MEs over time (p <0.001).+
Silveira (2007) (Delgado et al. 223–230)
Design: Before-after
N = 4,814 orders (treatment lines)
Implementation: 00/0000
Study Start: 00/0000
Study End: 00/0000
e-Rx
Integrated EHR/EMR system, Pharmacy
General Hospital, 53 beds in 2 wards of the hospital Bedsrate of errors:
-

medication data,

-

dose,

-

administration frequency/time,

-

route of administration,

-

nursing transcription

The EP system was associated with a lower rate of errors compared with the manual system for medical data (38% vs. 8%, RRR 79%, p <0.05), dosage (29% vs. 2%, RRR 92%, p <0.05), administration frequency/time (6% vs. 1%, RRR 83%, p <0.05) and route of administration (17% vs. 0%, RRR 99%, p <0.05). Nursing transcription errors were increased (18% vs. 21%, RRR 17% p <0.05) while drug interaction (2% vs. 3%) and treatment duration errors (1% vs. 1%) remained the same.+
Sintchenko (2005) (Sintchenko et al. 398–402)
Design: Before-after
N = not reported n/a
Implementation: 10/2002
Study Start: 04/2002
Study End: 03/2003
CDSS/CDS/CCDS/reminders
Integrated Laboratory system
Acute care/tertiary, Critical care units (CCU, ICU, NICU) 800 (18 bed ICU) Beds Academicantibiotic consumption (defined daily doses/1,000 patient days)*Consumption of antibiotics in defined daily doses/1,000 patient days decreased significantly after implementation of the hand-held decision support tool (1,767 vs. 1,458, p = 0.04).+
Small (2008) (Small, Barrett, and Price 181–187)
Design: Cross-sectional
N = 1,941 prescriptions
Implementation: 01/2003
Study Start: 01/2005
Study End: 09/2005
CPOE/POE systemAcute care/tertiary, Academicerror rate*, types of errors, severity of errors, error rates among prescribersFor error rates using computerized vs. spreadsheets indicated a relative risk reduction of 42% (20% vs. 12%, RRR 42%, p <0.0001) The distribution of type of error differed significantly according to prescription method (p <0.001) and the distribution of severity of errors also differed significantly according to prescribing method (p <0.001).+
Smith (2006) (Smith et al. 1098–1104)
Design: Time series
N = no sample size given number of dispensings
Implementation: 09/2000
Study Start: 10/1999
Study End: 12/2002
CDSS/CDS/CCDS/reminders
Integrated CDSS/CDS/CCDS/reminders
CPOE/POE system
EHR/EMR system
Ambulatory carenumber of dispensing of non-preferred drugs/10,000 population in elderly patients, number of dispensing of preferred drugs/10,000 population in elderly patients, number of dispensing of non-preferred drugs/10,000 population in non-elderly patientsFollowing the implementation of the drug- specific alerts, a large and persistent reduction (5.1 prescriptions per 10,000, p = 0.004) a 22% relative decrease from the month before alert implementation, in the exposure of elderly patients to nonpreferred medications was observed. We found no evidence of a decrease in use of nonpreferred agents for nonelderly patients. There was an upward, though non- significant trend in the use of preferred agents in elderly patients following the intervention (p = 0.66).
Sobieraj (2008) (Sobieraj 1755–1760)
Design: Before-after
N = 101 patients
Implementation: 03/2007
Study Start: 07/2006
Study End: 00/0000
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated CPOE/POE system
Acute care/tertiary, 819 Beds Academiccompliance with ordering VTE prophylaxisThe addition of alerts for patients at risk of VTE and an education program resulted in a significant improvement in compliance with ordering VTE prophylaxis (49% vs. 93%, RRR -90%, p <0.001).+
Spencer (2005) (Spencer et al. 416–419)
Design: Before-after
N = 5,063 medication errors
Implementation: 10/2002
Study Start: 01/2002
Study End: 05/2003
CPOE/POE systemAcute care/tertiary, 688 Beds Academicreported errors per dischargeImplementation of CPOE on the two units was associated with a significant increase in reported errors, from 0.068 per discharge before CPOE implementation to 0.088 per discharge afterward (p = 0.01).+
Steele (2005) (Steele et al. e255)
Design: Before- after
N = 54,206 patient visits
Implementation: 12/2002
Study Start: 08/2002
Study End: 04/2003
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
EHR/EMR system, Laboratory system
Ambulatory carepercentage of:
-

time provider ordered the rule-associated laboratory test (for which alert was triggered and message displayed),

-

times medication order triggered but not completed (for an abnormal laboratory value),

-

times the provider ordered the rule- associated laboratory test (for alert that was triggered for a missing laboratory test)

Medication orders for which an alert was presented shows an increase in the percentage of time the provider ordered the rule-associated laboratory test (38.5% vs. 51.1%, p, 0.001). When alert was for an abnormal laboratory value, percentage of times medication order triggered but was not completed increased from 5.6% at baseline to 10.9% during the intervention (p = 0.03). The largest effect was noticed when the alert was triggered for a missing laboratory test, the percentage of times the provider ordered the rule-associated laboratory test increased from 43.0% at baseline to 62.0% (p <0.001). All other outcomes did not have statistically significant change.+
Stone (2009) (Stone et al. 960–967)
Design: Before- after
N = 18,884 procedures
Implementation: 05/2007
Study Start: 12/2006
Study End: 05/2008
CPOE/POE systemUnspecified Hospitalmedication error rates, Mean total time from placement of order to nurse receiptMedication error rates did not decrease significantly from preimplementation to 6 or 12 months postimplementation (0.22% vs. 0.16 % vs. 0.21%, p = NS). Mean total time from placement of order to nurse receipt before implementation was significantly reduced (41.2 minutes vs. 27 seconds, p <0.01).
Tamblyn (2003) (Tamblyn et al. 549–556)
Design: RCT
N = 12,560 Patients
Implementation: 00/0000
Study Start: 01/1997
Study End: 02/1998
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory carerate of initiation of inappropriate drugs per 1,000 visits, Rate of discontinuation of inappropriate drugs per 1,000During the study the number of new potentially inappropriate prescriptions per 1,000 visits was lower (52.2 v 43.8) in the CDS group than in the control group (RR 0.82, 95% CI 0.69 to 0.98). The rate of discontinuation of inappropriate drugs per 1,000 was not different: 67.4 vs. 71.4, (RR 1.06, 95% CI 0.089 to 1.26).+
Tamblyn (2010) (Tamblyn et al. 176–188)
Design: RCT
N = 2,293 patients
Implementation: 00/0000
Study Start: 04/2006
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system, Insurance, provincial beneficiary and prescription databases
Ambulatory carerate of drug profile review, Changes in therapySignificant increase in drug profile review in the intervention compared to the control group (44.5% vs. 35.5%; p <0:001). There was no statistically significant difference between the intervention and control group in the proportion of patients who had increases in therapy (28.5% vs. 29.1%; OR 0.98; p = 0.86).+
Tang (1999) (Tang et al. 115–121)
Design: Time series
N = 2,484 patient visits
Implementation: 07/1996
Study Start: 10/1995
Study End: 01/1998
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory care, Academiccompliance rates with vaccination guidelines*- computer users compliance rates with vaccination guidelines*- paper usersCompliance rates did not increase in the first year for either group. For the computer users, compliance rates steadily increased year 2 to year 3 to year 4 (38.7% vs. 60.9%, RRR -57%, p = 0.001; 60.9% vs. 68.2%, RRR -12%, p = 0.02). For the paper group, year 2 to 3 saw a significant increase (28.5 vs. 37.0, p = 0.02), but year 3 to 4 saw a significant decrease (37.0% vs. 30.6%, p = 0.03). No comparisons between paper and computer were performed by the authors.+
Tang (2009)162
Design: Before- after
N = 1,762 patients
Implementation: 01/2005
Study Start: 09/2004
Study End: 06/2007
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
EHR/EMR system, Laboratory system, Pharmacy
Ambulatory careoverall compliance rate, pregnancy test ordering, Charting of cumulative dose, liver function and lipid profile test orderingIntroduction of e-isotretinoin chart resulted in marked improvement in physician compliance to all steps of the isotretinoin prescription process, with the overall compliance rate increasing from 57.5% to 97.8% (p <0.05) in the first year post- implementation. Of the female patients, 100% were tested for pregnancy prior to starting isotretinoin therapy, an increase of 66.0% compared with the pre- implementation period (p <0.05). Charting of cumulative dose improved (an increase of 13.5% to 99.5%, p <0.05) so did liver function tests and lipid investigations (an increase of 3.8% to 100%). The results demonstrated close to 100% compliance with charting of cumulative dose of isotretinoin, pregnancy testing, liver function and lipid profile tests. The results sustained for more than 2 years from January 2005 to June 2007 [no analysis given past 1 year].+
Teich (2000)163
Design: Time series
N = not reported orders
Implementation: 10/1993
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated Hospital information system, Imaging systems, Laboratory system
Acute care/tertiary, 720 Beds AcademicH2-blocker orders, variability in dosages, frequency of administration and exceeding maximum dosages, proportion of orders for 3x IV ondansetron, compliance with heparin ordered consequent to bed restStudy 1:
Nizatidine was used for <20% of all oral H2- blocker orders before implementation of the alert, vs. >95% after wards (p < 0.001); this was sustained for year 1 and 2. The use of IV ranitidine increased from 0% before the intervention to 71% of intravenous H2- blocker orders (32/45) in the first week and to 97% or more from the fourth week onward.
Study 2: Variability in standard deviation dosages across medications reduced by 11% following implementation of the dosage guidance application (p <0.001). Maintained over 3 years followup. Standard deviation of frequency of administration reduced by 30% post- implementation (p <0.001) and proportion of orders exceeding maximum dose decreased significantly from 2.1% to 0.56% post- implementation (p <0.001).
Study 3: Orders for 3x IV ondansetron increased significantly after the preferred order was highlighted in the dosing list (5.9% vs. 93.5%, RRR - 1485%, p <0.001). Study 4: heparin ordering with bed rest increased from 23.9% to 46.9%.
+
Terrell (2009)164
Design: RCT
N = 5,162 Patients
Implementation: 00/0000
Study Start: 01/2005
Study End: 07/2007
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
Acute care/tertiary, 450 Beds Academicproportion of ED visits by seniors with an inappropriate medication, proportion of medications that were potentially inappropriate was also reducedThe decision support reduced the proportion of ED discharges that resulted in potentially inappropriate prescriptions (3.9% vs. 2.6%; p = 0.02; OR 0.55, 95% CI 0.34 to 0.89). The proportion of medications that were potentially inappropriate was also reduced, from 5.4% to 3.4% (p =.006; OR 0.59, CI 0.41 to 0.85).+
Terrell (2009)165
Design: RCT
N = 5,162 patient visits to 63 physicians
Implementation: 00/0000
Study Start: 01/2005
Study End: 07/2007
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
EHR/EMR system
Acute care/tertiary, 450 Beds Academicvisits with an inappropriate medication prescription*, prescriptions that were inappropriate, n (%)Primary Outcome: Decision support significantly reduced the proportion of ED discharges that resulted in a potentially inappropriate prescription (3.9% vs. 2.6%; p = 0.02; OR 0.55, 95% CI 0.34 to 0. 0.89. This difference represents an absolute RR of 1.3% (95% CI 0.4 to 2.3).
Secondary Outcome: When analyzed as a percentage of all medications prescribed by physician subjects, the proportion of medications that were potentially inappropriate was significantly reduced, from 5.4% to 3.4% (p = 0.006; OR 0.59, 95% CI 0.41 to 0.85), with an absolute reduction of 2.0% (95% CI 0.7 to 3.3).
+
Tierney (2003)166
Design: RCT
N = 706 patients, 20 pharmacists, 94 physicians and 1 nurse practitioner
Implementation: 00/0000
Study Start: 01/1994
Study End: 05/1996
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system, Pharmacy
Ambulatory care, Outpatient hospital based Academiccompliance with cardiac care suggestions*Neither the physician nor the pharmacist intervention had any significant effect on whether patients’ cardiac care was compliant with the suggestions (p >0.8 across the 4 intervention groups by analysis of variance, with p > 0.7 and p >0.4 when testing the physician and pharmacist interventions separately).
Tierney (2005)167
Design: RCT
N = 706 patients
Implementation: 00/0000
Study Start: 01/1994
Study End: 05/1996
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system
EHR/EMR system, Pharmacy
Ambulatory care, Pharmacy Outpatient hospital based Academicadherence to the care suggestions*There were no differences between the four study groups in either adherence to the care suggestions, combined or individually (32% control, 32% physician intervention, 32% pharmacist intervention, 37% both interventions, NS).
Upperman (2005)168
Design: Before- after
N = Not reported ADE/1,000 doses
Implementation: 00/2002
Study Start: 01/2002
Study End: 0/0000
CPOE/POE system
Integrated EHR/EMR system
Acute care/tertiary, Pediatric stand alone hospital, AcademicADE rates per 1,000 before and after CPOE implementationAll ADEs before CPOE were 0.3 per 1,000 doses, whereas after CPOE ADEs were 0.37 per 1,000 doses (p = 0.3).+
Uttaro (2007)169
Design: Cohort study
N = 23 psychiatrists
Implementation: 01/2004
Study Start: 01/2004
Study End: 03/2005
CDSS/CDS/CCDS/reminders
Stand-Alone, New York State Office of Mental Health intranet, pharmacology resources
Other specialty hospital (rehab, oncology)percentage of caseloads on 2 or more antipsychotics*, overall percentage of patients on 2 or more antipsychoticsOverall, there were moderately large reductions for most psychiatrists in the percentage of caseloads on 2 or more antipsychotics (56% vs. 36%, RRR 36%, p <0.01). There were significantly greater reductions in March 2005 for psychiatrists who had higher percentages of their caseloads on two or more concurrent antipsychotics in January 2004. The overall percentage of patients on 2 or more antipsychotics dropped significantly (54% vs. 36%, RRR 33%, p <0.01).+
van Doormaal (2009)170
Design: Time series
N = 1,195 patients
Implementation: 00/0005
Study Start: 07/2005
Study End: 05/2008
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated Barcoding system, Pharmacy
Acute care/tertiary, 1,300 (Groningen); 600 (Tilburg and Waalwijk) Beds Academicmedication errors (ME)* preventable adverse drug events (pADEs)*During the baseline period, 55% of all medication orders (MOs) contained at least one or more MEs, whereas during the postintervention period this was 17%; a significant immediate absolute reduction of 40.3% (95% CI: −45.13% to 35.48%). In the baseline period, 15.5% of admitted patients experienced one or more pADE, as opposed to 7.3% in the postintervention period. Decrease could not be attributed to CPOE/CDSS. The immediate change was NS (−0.42%, 95% CI: −15.52% to 14.68%) because of the observed underlying negative trend during the pre- CPOE period of −4.04% [95% CI: −7.70% to 0.38%] per month.+
Van Wyk (2007)171
Design: RCT
N = 87,860 Patients
Implementation: 00/0000
Study Start: 05/2004
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system
Ambulatory carePercentage of patient treatedOf the patients requiring treatment, 66% were treated in the alerting arm, 40% in the on-demand arm, and 36% in the control arm. After adjustment for differences between arms, the likelihood of being treated was 40% higher in the alerting arm (adjusted RR = 1.40; 95% CI 1.15 to 1.70) and 19% higher (NS) in the on- demand arm in comparison to the control arm (adjusted RR = 1.19; 95% CI 0.94 to 1.50). A similar pattern was shown for the need for screening within the 3 groups.+
Varkey (2007)172
Design: Cross- sectional
N = 4,527 prescriptions
Implementation: 00/0000
Study Start: 00/1996
Study End: 00/2002
CPOE/POE system
Integrated CDSS/CDS/CCDS/reminders
Ambulatory care, Other institution basedfrequency of intercepted prescription errors*Statistically significant decrease in frequency of intercepted prescription errors among handwritten and computerized prescriptions was observed (7.4% vs. 4.9%, p = 0.0048).+
Voeffray (2006)173
Design: Before- after
N = 2,445 prescriptions
Implementation: 00/0000
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated Pharmacy
Acute care/tertiary, 850 Beds Pharmacy, Inpatient hospital based, AcademicRate of error*The average monthly error rate was 15% (95% CI 13% to 18%). After introduction of the CPOE system, the average monthly error rate (which included both computer orders and handwritten, amounted to 13% (95% CI 10% to 16%). This decrease in rate was not statistically different from the rate observed in the first period (p = 0.36). Postimplementation errors in the computerized group only was 0.6% (95% CI 0.3% to 1.4%).
Walsh (2008)174
Design: Time series
N = 627 admissions
Implementation: 04/2002
Study Start: 09/2001
Study End: 05/2003
CPOE/POE system
Integrated CDSS/CDS/CCDS/reminders
Critical care units (CCU, ICU, NICU) General Hospital, 59 Pediatric beds Bedsrates non-intercepted serious medical errors*The rates of errors did not differ for all errors (44.7 before vs. 50.9 errors per 1,000 patient days after COPE, NS), non-intercepted serious medical errors (23.1 before vs. 20.6 per 1,000 patient days after CPOE, NS), or serious medical errors (31.7 before vs. 33.0 per 1,000 patient days after CPOE, NS).
Were (2009)175
Design: Before-after
N = 40 Patients
Implementation: 00/0000
Study Start: 12/2007
Study End: 04/2008
CDSS/CDS/CCDS/reminders
CPOE/POE system
Integrated EHR/EMR system, Imaging systems, Laboratory system
Acute care/tertiary, 264 Beds Academicacceptance of all recommendations, rate of acceptance of pharmacological recommendationsMore recommendations were implemented in the reminders group (59% vs. 78%, RRR 32%, p = 0.01) The rate of acceptance of pharmacological recommendations was similar (51% vs. 77%).+
Wilkes (2009)176
Design: Before- after
N = 84 patients
Implementation: 06/2005
Study Start: 06/2005
Study End: 05/2006
CDSS/CDS/CCDS/reminders
Integrated EHR/EMR system, Laboratory system
Acute care/tertiary, Pediatric stand alone hospital, 418 Bedsprescription rate among eligible patients, prescription rate -off-labelThe rate of oseltamivir prescription did not change significantly for patients eligible for the drug (40% vs. 25%, RRR 38%, p = NS), or for off-label prescribing for patients not eligible for the drug (4% vs. 5%, RRR -24%, p = NS) following the implementation of a computerized reminder.
Wrona (2007)177
Design: Observational study
N = 536 PCA patients
Implementation: 00/2003
Study Start: 01/2003
Study End: 03/2004
CPOE/POE system,
Integrated EHR/EMR system, Imaging systems, Laboratory system
Pediatric stand alone hospitalrates of respiratory monitoring rates of oxygen saturation monitoringCompared to the control group of ‘no order set’, patients in the Acute Pain Team Service had a higher rate of respiratory monitoring (43% vs. 66.3%, RRR - 54%, p <0.05) and oxygen saturation monitoring (86.1% vs. 98.6%, RRR - 15%, p <0.05). Compared to the control group of ‘no order set’, patients in the prescriber initiated PCA had higher respiratory rate monitoring (43% vs. 57.8%, RRR - 34%, p <0.05). No other comparisons were significant.+
Xamplas (2010)178
Design: Before-after
N = 96 patients
Implementation: 02/2008
Study Start: 00/0000
Study End: 00/0000
CDSS/CDS/CCDS/reminders
Integrated CPOE/POE system, e-MAR, Pharmacy
Acute care/tertiary, 465 Beds Inpatient hospital based, AcademicPiperacillin–tazobactam days per 1,000 patient- days*, Piperacillin– tazobactam doses per 1000 patient-days*While the number of piperacillin– tazobactam days per 1,000 patient days did not significantly change (124 ± 6.3 vs. 121 ± 12.6, p = 0.389) during the preintervention and postintervention periods, there was a significant reduction in the number of piperacillin– tazobactam doses per 1,000 patient- days during the postintervention period (457 ± 33.3vs. 341 ± 35.7, p <0.001).+
Yu (2009)179
Design: Cross-sectional
N = 3,364 hospitals
Implementation: 00/0000
Study Start: 07/2003
Study End: 06/2004
CPOE/POE systemUnspecified Hospital, Not specified11 medication quality indicators*Among the 11 medication- related measures for acute myocardial infarction, heart failure and pneumonia, the mean performance on 6 cardiovascular- related measures was higher among CPOE hospitals (p <0.001) vs. the comparison (nonCPOE) hospitals. Also, for one pneumonia measure, administering “Antibiotics within 4 hours of arrival for patients with pneumonia,” performance was lower for hospitals with full CPOE implementation (p <0.001). Four quality indicators were not significantly different among the groups; 3 for pneumonia and administration of thrombolytic agent within 30 minutes for AMI. The differences are maintained when hospital teaching status and ownership and number of beds are taken into account.+
Zanetti (2003)180
Design: RCT
N = 273 patients
Implementation: 00/0000
Study Start: 03/2000
Study End: 06/
2000
CDSS/CDS/CCDS/reminders
Integrated Hospital information system
Acute care/tertiary, Academicappropriate redosing of antibiotics*More patients in the alarm plus reminder group received appropriate redosing of antibiotics after > 240 minutes in surgery (adjusted OR 3.31, 95% CI 1.97 to 5.56, p <0.0001).+
Zhan (2006)181
Design: Mixed methods
N = 138,922 number of errors/100,000 doses
Implementation: 00/0000
Study Start: 01/2003
Study End: 12/2003
CPOE/POE systemUnspecified Hospitalnumber of errors reported per 100,000 doses-inpatients*, number of errors reported per 100,000 doses-outpatients*The number of errors reported per 100,000 doses was not different among non-CPOE (n=339) and CPOE facilities (n=120) for inpatients (mean 56 vs. 55, p = 0.9) or outpatients (mean 60 vs. 57, p = 0.8).

The HIT system studied is in bold, followed by the systems that it was integrated with. The outcome column indicates whether at least 50% of the relevant outcomes abstracted were positively impacted by the MMIT (+) or not (−).

*

indicates outcomes noted as being the primary outcome by the paper’s authors

Abbreviations: A1c = hemoglogin A1c; ACE = Angiotensin Converting Enzyme; ACEI = Angiotensin-Converting Enzyme Inhibitor; ADEs = Adverse Drug Events; ALT = Alanine Aminotranceferase; AMI = Acute Myocardial Infarction; AR = Absolute Reduction; ARB = Angiotensin-II-Receptor Blocker; ARI = Acute respiratory infection; AST = Aspartate Aminotransferase; CC = Care Considerations; CCDS = Computerized Clinical Decision Support; CDS = Clinical/Computerized Decision Support; CDSS = Clinical Decision Support System; CHF = Congestive Heart Failure; CI = Confidence interval; CIT = Clinical information technology; COPD = Chronic Obstructive Pulmonary Disease; CPG = Clinical Practice Guidelines; CPOE = Computerized Provider Order Entry; DDI = Drug-drug Interaction; DS = Decision Support; DSS = Decision Support System; ED = Emergency Department; EHR = Electronic Health Record; e-MAR = Electronic Medication Administration Record; EMR = Electronic Medical Records; EP = Electronic Prescribing; e-RX = Electronic Prescribing; e-TAR = Electronic Treatment Authorization Request; GP = General Practitioner; h = Hour; HIT = Health Information Technology; HIV = Human Immunodeficiency Virus; hr = Hour; hrs = Hours; ICU = Intensive Care Unit; K = Potassium; LVSD = Left Ventricular Systolic Dysfunction; ME = Medication Error; Mg = Magnesium; min = Minute; MMR = Measles; Mumps and Rubella; N = Sample Size; n/a = Not Applicable; Np = Nurse Practitioner; NR = not reported; NS = NS; NSAID = Nonsteroidal anti-inflammatory drug; NSAIDS = Nonsteroidal anti-inflammatory drugs; OR = Odds ratio; OSUH = Ohio State University Health System; p = Probability; PCA = Patient-Controlled Analgesia; PDA = Personal Digital Assistants; PICU =;Pediatric Intensive Care Unit; POE = Provider Order Entry; PONV = Postoperative Nausea and Vomiting; PRN = pro re nata; RCT = Randomized Controlled Trial; RR = Relative Risk; RRR = Relative Risk Reduction; RV = rule violation; UDDS = Unit Dose Drug Dispensing System; UTI = Urinary tract Infection; vs. = Versus; VTE = Venous Thromboembolism

indicates outcomes noted as being the primary outcome by the paper’s authors

From: Appendix C, Evidence Tables

Cover of Enabling Medication Management Through Health Information Technology
Enabling Medication Management Through Health Information Technology.
Evidence Reports/Technology Assessments, No. 201.
McKibbon KA, Lokker C, Handler SM, et al.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.