Background: Patient understanding of prescription medication regimens is an important aspect of health literacy and inpatient medication reconciliation. There is a lack of data regarding patient knowledge of their own outpatient prescription medications. This study evaluated the ability of patients admitted to a community hospital medicine service to provide, from memory, their preadmission prescription medication list.Methods: All adult patients admitted to the inpatient medicine service of a community hospital during a three-month period were screened by the admitting provider. Patients were asked to recall, by memory, the names of their outpatient medications. Patients were not asked for dosing specifics or drug indication. The self-reported list of each patient was verified with family, pharmacy, or institutional care-provider. Patients were excluded if they were assessed by the provider as cognitively impaired, or if their medication history could not be verified. Subsequent readmissions for any included patient were excluded. Patient age, admission diagnosis, number of medications (grouped: 0, 1, 2-5, >5), and accuracy of medication history (yes or no) were recorded.

Results: 1323 consecutive patients were screened. 309 patients were excluded (164 cognitively impaired, 133 readmitted during study period, and 12 unable to verify). 1014 patients were enrolled; 46.9% patients were male (M=61) and 53.1% female (M=66). 49.5% of patients were able to correctly recall all outpatient prescription medications. Age increased with more medications (Figure 1) and was significantly higher in the medication number groupings of 2-5 (p<0.001) and >5 (p<0.001) compared to other groups (F(3,1010)=62.63, p<0.001). Percent accuracy decreased with more medications (Figure 1), but none of the dummy variables used for number of medications were significant predictors of accurate medication history; this remained true when controlled for age and sex. Increased age lowered the odds (p<0.001) and females had higher odds (p<0.01) of an accurate medication history controlling for all other variables. The overall model was significant (χ2(5)=413.21, McFadden R2=0.29, p<0.001).

Conclusions: Newly admitted inpatients perform poorly when asked to list their outpatient medications. Patient age and gender are related to patient inability to accurately recall all medications by name. Total prescribed medications increase, on average, with advancing age. As such, difficulties with prescription knowledge are amplified in the elderly.  This may directly impact the medication reconciliation process and the safe transition to inpatient care. Medication histories, particularly in older patients taking more than five medications, are suspect and should be verified by collateral sources.