OBJECTIVES : Seventeen Medical Homes (MHs) were established in the Local Health Authority (LHA) of Parma, Emilia-Romagna, Italy, between 2011 and 2016. We evaluated the association between MH implementation and potentially inappropriate medication (PIM) use as a quality indicator.

METHODS : We conducted a retrospective cohort study (01/01/2011-12/31/2017) using the Parma LHA administrative healthcare database. Patients were ≥ 65 years, lived in Parma for ≥1 year, had ≥1 medication dispensed, and had a documented primary care physician (PCP). PIM utilization was determined based on medications that should be always avoided in the 2011 and 2014 Maio criteria, an Italian modified version of the Beers criteria. MH exposure was determined by time-varying PCP affiliation with MHs (non-MH, pre-MH, and post-MH). Rates of PIM use were compared between exposure groups using repeated events Cox proportional hazards models and further examined based on MH complexity (high/medium or low).

RESULTS : Among 129,938 older adults included, 22% of their total person-time was under exposure to MHs. Compared with non-MH patients, MH patients had higher rates of PIM use before MH opening (HR=1.06; 95% CI: 1.04-1.08) but lower rates once MHs opened (HR=0.97; 95% CI: 0.96-0.98). The MH implementation was associated with an 8% decrease in PIM (post-MH vs. pre-MH HR=0.92; 95% CI: 0.89-0.93). After MH implementation, patients in both high/medium and low complexity MHs had lower use of PIMs but the reduction was greater for those in low complexity MHs (HR=0.86; 95% CI: 0.83-0.89) than in high complexity MHs (HR=0.93; 95% CI: 0.91-0.95).

CONCLUSIONS : Accounting for pre-existing differences in prescribing between MH and non-MH PCPs, the MH implementation was associated with a significant decrease in PIM use. Notably, this improvement was greater among patients in low complexity MHs. More research is needed to understand whether the reduction in PIP exposure will be sustained as MHs progress.