Medical News Stories
National Study on the Distribution, Causes, and Consequences of Voluntarily Reported Medication Errors Between the ICU and Non-ICU Settings
Study Question: Because ICU patients are more critical and usually have more complex medication regimens, are medication errors made in the ICU more likely to result in harm than those made in the non-ICU setting?
Study Description: This was a cross-sectional study of all hospital ICU and non-ICU medication errors reported to the MEDMARX system (a national database, participation in which is voluntary) between 1999 and 2005. Hospitals were characterized by size, inpatient doses per month, pharmacist availability, electronic ordering systems, hospital type, and automatic medication dispensing system.
Results: A total of 839,553 errors were reported from 537 hospitals. ICUs accounted for 55,767 (6.6%) of reported errors, with harm occurring more frequently compared to non-ICU floors (3.7% vs. 1.9%; OR 1.89 [1.62-2.17]). Medical and cardiac ICUs accounted for the largest proportion of submitted errors at 20% and 33%, respectively, and most occurred during the administration phase of the medication process (44%). The leading cause of harmful errors was described as “deficits in knowledge and performance” (53%) and, in the ICU, there were twice as many harmful errors directly attributable to calculations (9.8% vs. 5.3%; OR 1.82 [1.48-2.24]). Fifty-three percent of the time that errors were made in the ICU, the staff member determined responsible for the error(s) was not informed of their role in the error.
Conclusion: The ICU reported more harmful medication errors than non-ICU floors, most often occurring during medication administration.
Perspective: ICU patients are subject to more harmful errors due to the high-risk medications utilized in this setting. The results reinforce the importance of maintaining systems that ensure accurate calculations and highlight the need to more thoroughly address errors, beginning with communication with and education of those involved.