Skip to main content
. 2022 Dec 13;5:179. doi: 10.1038/s41746-022-00739-x

Table 2.

Prescribing crude error rates and adjusted incidence rate ratios (IRRs) for the SWCRCT (short-term impact of eMM).

Control (Paper) Intervention (first 70 days of eMM) IRR (95% CI)a (eMM vs Paper) P valuea
All clinical prescribing errors
Number of patients 1686 2096
Number of medication orders 9635 16,734
Number of clinical prescribing errors 1399 2942
Error rateb per 100 orders (95% CI) 14.52 (13.78–15.30) 17.58 (16.96–18.23) 1.05 (0.92–1.21) 0.45
Potential ADEs (clinical prescribing errors with a potential severity score ≥3)
Number of potential ADEs 353 812
Error rateb per 100 orders (95% CI) 3.66 (3.30–4.07) 4.85 (4.53–5.20) 1.62 (1.28–2.04) <0.001
Clinical prescribing errors involving high-risk medications
Number of patients with high-risk medication orders 799 1228
Number of high-risk medication orders 1200 2274
Number of clinical prescribing errors involving high-risk medications 232 390
Error rateb per 100 high-risk medication orders (95% CI) 19.33 (17.00–21.99) 17.15 (15.53–18.94) 0.88 (0.66–1.16) 0.35
All procedural errorsc
Number of patients with procedural errors reviewed 1162 1429
Number of medication orders 6698 11,441
Number of procedural errors 7652 7699
Procedural error rateb per 100 orders (95% CI) 114.24 (111.71–116.83) 67.29 (65.81–68.81) 0.65 (0.61–0.70) <0.001

aAdjusted incidence rate ratio (IRR), 95% CIs and P values from generalised linear mixed models adjusting for patient age, study time, and clustering by ward (unit of randomisation) and patient. ADE is adverse drug event.

bCrude error rates and 95% CIs are from simple Poisson models.

cProcedural errors were assessed from a random sample comprising 68.5% of patients during the entire SWCRCT period. Details of changes in procedural errors by error type are in Supplementary Tables 1 and 2.