Medication discrepancy rates among Medicaid recipients at hospital discharge

To compare the rate of discrepancies per patient for medications changed during hospitalization in patients with and without prescription provider comments at hospital discharge. Secondary objectives included comparing 35-day readmission rates, describing the overall medication discrepancy rate stratified by age group and type of discrepancy, collecting average number of medication changes, and reporting percentage change in admission medications at discharge.

This single-center prospective cohort included NC Medicaid recipients discharged from East Carolina University Family Medicine service from November 1, 2015, to January 31, 2016. Patients were assigned to a group based on presence or absence of provider comments on discharge prescriptions. Outpatient pharmacy claims were compared with the discharge summary to identify medication discrepancies. Medication discrepancy rates between groups were to be compared by means of independent-samples t test. Medication discrepancy rates were compared according to 35-day readmission status, age group, and type of discrepancy by means of independent-samples t tests and analysis of variance. Descriptive statistics were used for other secondary outcomes.

Of 118 patients included, only 1 had provider comments. Therefore, a medication discrepancy rate comparison was not performed. Patients had a mean of 4 medication changes made to their regimen and 21.3% change in admission medications. Sixty-one percent of patients had at least 1 medication discrepancy, with an overall rate of 1.19. Patients readmitted within 35 days had a significantly greater medication discrepancy rate than those not readmitted (1.63 vs. 1.05, respectively; P = 0.044). Patients 18-49 years of age had the highest discrepancy rate and those older than 80 years of age the lowest (1.58 and 0.50, respectively). New or changed discharge medication not filled accounted for 69% of discrepancies.

Although medication discrepancies were common, use of provider comments was rare. Future studies should address more effective ways to communicate pertinent information to community pharmacists and methods to improve adherence in obtaining new medications.