Table 2

Randomised controlled trials investigating double checking of medication administration

StudyCountrySample size/study durationSettingMethod of measuring errorsFindingsStudy quality
Kruse 199216 Australia129 234 oral, inhaled or topical administrations over 46 weeks3 wards of a geriatric assessment and rehabilitation unitChart review data supplemented by incident reports
  • The error rate per 1000 administrations was lower for double-checked administrations (2.12 (1.69 to 2.55)) compared with single checking (2.98 (2.45 to 3.51))

Fair
Modic 201623 USA5238 administrations of subcutaneous insulin to 266 patientsPatients with diabetes at a 1400-bed quaternary care hospitalIn double check group, anonymous self-report; in single check group, review of electronic medical records
  • The error rate for double-checked administrations was significantly lower (28.8%) compared with single checking (36.7%; p<0.001)

  • In multivariate regression, double-checked administrations were significantly associated with a lower odds of any type of error (OR 1.38 (1.23 to 1.55)), but not after adjustment for nurse, to account for correlated administrations (OR 1.18 (0.83 to 1.68))

Fair
Douglass 201825 USA43 pairs of ED and ICU nursesSimulated adult patient in a medical education centreDirect observation
  • 9% of nurses detected the weight-based dosage error in the single check group compared with 33% in the double check group (OR 5.0 (0.90 to 27.74))

  • 54% of nurses detected the wrong phial error in the single check group compared with 100% in the double check group (OR 19.9 (1.0 to 408.5))

  • Adherence to double checking was observed to be 100% (21 out of 21 nurses)

Good
  • ED, emergency department; ICU, intensive care unit.