Medication Complexity, Medication Number, and Their Relationships to Medication Discrepancies.

ABSTRACT

BACKGROUND: Medication reconciliation to identify discrepancies is a National Patient Safety Goal. Increasing medication number and complex medication regimens are associated with discrepancies, nonadherence, and adverse events. The Medication Regimen Complexity Index (MRCI) integrates information about dosage form, dosing frequency, and additional directions. OBJECTIVE: This study evaluates the association of MRCI scores and medication number with medication discrepancies and commissions, a discrepancy subtype. METHODS: This was a retrospective cohort study of a convenience sample of 104 ambulatory care patients seen from April 2010 to July 2011 within the Department of Veterans Affairs. Primary outcomes included any medication discrepancy and commissions. Primary exposures included MRCI scores and medication number. Multivariable logistic regression models associated MRCI scores and medication number with discrepancies. Receiver operating characteristic (ROC) curves provided discrepancy thresholds. RESULTS: For the 104 patients analyzed, the median MRCI score was 25 (interquartile range [IQR] = 14-43), and the median medication number was 8 (IQR = 5-13); 60% of patients had any discrepancy, whereas 36% had a commission. In adjusted analyses, patients with MRCI scores ≥25 or medication number ≥8 were more likely to have commissions (odds ratio [OR] = 3.64, 95% CI = 1.41-9.41; OR = 4.51, 95% CI = 1.73-11.73, respectively). The unadjusted ROC threshold for commissions was 36 for MRCI (sensitivity, 59%; specificity, 82%) and 9 for medication number (sensitivity 68%; specificity 67%). CONCLUSION: Patients with either MRCI scores ≥25 or ≥8 medications were more likely to have commissions. Given equal performance in predicting discrepancies, the efficiency and simplicity of medication number supports its use in identifying patients for intensive medication review beyond medication reconciliation.