BACKGROUND: Patients transitioning to a nursing home from their home or other facility are at high risk for medication errors. OBJECTIVE: Our aim was to describe characteristics of medication errors occurring during transitions to nursing homes, to compare characteristics of transition errors with errors not involving a transition, and to evaluate the impact of these errors on patient harm. METHODS: This was a cross-sectional analysis of individual medication error incidents reported by North Carolina nursing homes to the Medication Error Quality Initiative during fiscal years 2007 through 2009. Bivariate associations between errors in transition with patient factors, error-related factors, reported causes of errors, and impact on patients were tested using a χ(2) test. Multivariate logistic regression explored whether medication errors during transitions were more harmful than errors not occurring during transitions. Patient-related factors included in the model were age, sex, and cognitive ability. Error-related factors were primary type of error, process phase when error began, primary personnel involved, and an indicator for repeat error. RESULTS: A total of 27,759 individual medication error incidents were reported over a 3-year period in North Carolina nursing homes. Of these errors, 2919 incidents (11%) involved a patient transitioning to a nursing home. Errors involved in transitions were found to have higher odds of patient harm compared with errors not involved in transitions (odds ratio = 1.85; 95% CI, 1.30-2.63). Staff communication, order transcription, medication availability, pharmacy issues, and name confusion were particularly important contributors to medication errors during transitions (P < 0.05 for comparison with nontransition errors). CONCLUSIONS: Transitions across care settings introduce risk for patient harm, and medication errors are an important area for improvement during transitions.
BACKGROUND:Patients transitioning to a nursing home from their home or other facility are at high risk for medication errors. OBJECTIVE: Our aim was to describe characteristics of medication errors occurring during transitions to nursing homes, to compare characteristics of transition errors with errors not involving a transition, and to evaluate the impact of these errors on patient harm. METHODS: This was a cross-sectional analysis of individual medication error incidents reported by North Carolina nursing homes to the Medication Error Quality Initiative during fiscal years 2007 through 2009. Bivariate associations between errors in transition with patient factors, error-related factors, reported causes of errors, and impact on patients were tested using a χ(2) test. Multivariate logistic regression explored whether medication errors during transitions were more harmful than errors not occurring during transitions. Patient-related factors included in the model were age, sex, and cognitive ability. Error-related factors were primary type of error, process phase when error began, primary personnel involved, and an indicator for repeat error. RESULTS: A total of 27,759 individual medication error incidents were reported over a 3-year period in North Carolina nursing homes. Of these errors, 2919 incidents (11%) involved a patient transitioning to a nursing home. Errors involved in transitions were found to have higher odds of patient harm compared with errors not involved in transitions (odds ratio = 1.85; 95% CI, 1.30-2.63). Staff communication, order transcription, medication availability, pharmacy issues, and name confusion were particularly important contributors to medication errors during transitions (P < 0.05 for comparison with nontransition errors). CONCLUSIONS: Transitions across care settings introduce risk for patient harm, and medication errors are an important area for improvement during transitions.
Authors: David G Stevenson; Stacie B Dusetzina; A James O'Malley; Susan L Mitchell; Barbara J Zarowitz; Michael E Chernew; Joseph P Newhouse; Haiden A Huskamp Journal: Med Care Date: 2014-10 Impact factor: 2.983