Testing a theoretical model for severe medication errors Public Deposited

Downloadable Content

Download PDF
Last Modified
  • March 21, 2019
  • Chang, Yun Kyung
    • Affiliation: School of Nursing
  • There were three purposes of this dissertation. The first was to develop a theoretical framework for severe medication errors in acute care hospitals by integrating the human error model and the organizational learning model. The second purpose was to examine direct effects between nursing units' error-producing conditions and severe medication errors. The third purpose was to test the moderating role of learning climate in the relationships between error-producing conditions and severe medication errors in acute care hospitals. The human error model guided the identification of error-producing conditions, which included work environment factors, team factors, person factors, patient factors and medication-related support services. The organizational learning model was used to introduce the concept of learning climate and hypothesize its moderating role between error-producing conditions and medication errors. It was hypothesized that when learning climate was positive, the relationship between error-producing conditions and medication errors would be weaker, and when learning climate was negative the relationship between error-producing conditions and medication errors would be stronger. Using data derived from the Outcomes Research in Nursing Administration II study, which included 286 nursing units at 146 hospitals across the U.S., results supported a positive relationship between RN hours and medication error rates and an asymptotic relationship between nursing education levels and medication error rates such that nursing units with 40-50% of BSN-prepared nurses had the lowest rate of severe medication errors. The study also found that nursing units with a positive learning climate had higher rates of severe medication errors than those with a negative learning climate, implying potential under-reporting and under-detection of medication errors. These results suggest several policy and practice implications. From a policy perspective, 40-50% of BSN-prepared nurses at the nursing unit may be the optimal proportion to minimize severe medication errors. Further, development of a better measurement of medication errors, which is free from under-reporting or under-detecting problems, is warranted. From a practice perspective, nursing units should identify error-producing conditions present on their units that contribute to medication errors and make an effort to create a positive learning climate by emphasizing open communication, revealing and thinking about medication errors.
Date of publication
Resource type
Rights statement
  • In Copyright
  • Mark, Barbara
Degree granting institution
  • University of North Carolina at Chapel Hill
  • Open access

This work has no parents.