Medication administration errors in assisted living: scope, characteristics, and the importance of staff training

Abstract

Objectives: To compare rates of medication errors committed by assisted living staff with different training and to examine characteristics of errors.

Design: Observation of medication preparation and passes, chart review, interviews, and questionnaires.

Setting: Stratified random sample of 11 assisted living communities in South Carolina (which permits nonnurses to administer medications) and Tennessee (which does not).

Participants: All staff who prepared or passed medications: nurses (one registered nurse and six licensed practical nurses (LPNs)); medication aides (n=10); and others (n=19), including those with more and less training.

Measurements: Rates of errors related to medication, dose and form, preparation, route, and timing.

Results: Medication preparation and administration were observed for 4,957 administrations during 83 passes for 301 residents. The error rate was 42% (20% when omitting timing errors). Of all administrations, 7% were errors with moderate or high potential for harm. The odds of such an error by a medication aide were no more likely than by a LPN, but the odds of one by staff with less training was more than two times as great (odds ratio=2.10, 95% confidence interval=1.27-3.49). A review of state regulations found that 20 states restrict nonnurses to assisting with self-administration of medications.

Conclusion: Medication aides do not commit more errors than LPNs, but other nonnurses who administered a significant number of medications and assisted with self-administration committed more errors. Consequently, all staff who handle medications should be trained to the level of a medication aide.