Inpatient medication reconciliation errors are common and can lead to serious adverse drug events. Most medication discrepancies are due to errors in taking an accurate medication history. Tools imbedded in the electronic medical record (EMR) are one mechanism by which the accuracy of medication reconciliation can be improved. We constructed an EMR-based admission navigation tool with medication reconciliation capabilities to reduce the potential for adverse drug events at an academic teaching hospital.
Purpose: To evaluate the impact of a new admission navigation tool in improving the accuracy of medication reconciliation on inpatient admissions to general medicine services.
Description:
An admission navigation tool was created and made available to 1 out of 4 general medicine teams as part of a 30-day pilot. Patients were randomly selected to participate in gold standard admission medication history by a trained nurse or pharmacist within 24 hours of admission. The gold standard history consisted of in-person or telephone review of medications with the patient, their families and/or outpatient pharmacy.
Patient charts were reviewed for discrepancies between the documented admission medication list and the gold standard. Accuracy was compared for the teams that used the admission navigation tool (intervention group) versus those that did not (control group). Errors identified were classified as class 1 (no potential for harm), class 2 (potential for mild-moderate harm), and class 3 (potential for serious harm).
There were 235 patients admitted to the general medicine service and 26 gold standard medication histories performed. In total, 421 medications were recorded on admission, with 59 medication errors noted (14%), including 33 omissions (55.9%), 14 errors in dosing or frequency (23.7%), 11 related to additional medications (18.6%), and 1 substitution (1.7%). There were 43 class 1 errors (72.9%), 16 class 2 errors (27.1%) and no class 3 errors.
Of the 13 patients in the control group, 45 errors were noted out of 187 medications (24.1%), including 14 class 2 errors and no class 3 errors. Of the 13 patients in the intervention group, 14 medication errors were noted out of 234 medications (6.0%), with 2 class 2 errors and no class 3 errors (Figure 1). Overall, the intervention showed an absolute risk reduction of 18.1% (95% CI 11.2-24.9, p<0.001) and a relative risk reduction of 75.1% (95% CI 56.1-85.9, p<0.001) in medication error rate, with an 88.6% relative risk reduction in class 2 errors (95% CI 50.4-97.4,p<0.001).
Conclusions: A high rate of medication discrepancies on admission to an academic medical center was discovered, with 1 in 5 of these errors potentially harmful. Implementation of an EMR-based admission navigator tool significantly reduced these errors.