Antipsychotic adherence patterns and health care utilization and costs among patients discharged after a schizophrenia-related hospitalization

BACKGROUND This study aimed to assess antipsychotic adherence patterns and all-cause and schizophrenia-related health care utilization and costs sequentially during critical clinical periods (i.e., before and after schizophrenia-related hospitalization) among Medicaid-enrolled patients experiencing a schizophrenia-related hospitalization. METHODS All patients aged ≥ 18 years with a schizophrenia-related inpatient admission were identified from the MarketScan Medicaid database (2004-2008). Adherence (proportion of days covered [PDC]) to antipsychotics and schizophrenia-related and all-cause health care utilization and costs were assessed during preadmission (182- to 121-day, 120- to 61-day, and 60- to 0-day periods; overall, 6 months) and postdischarge periods (0- to 60-day, 61- to 120-day, 121- to 180-day, 181- to 240-day, 241- to 300-day, and 301- to 365-day periods; overall, 12 months). Health care utilization and costs (2010 US dollars) were compared between each adjacent 60-day follow-up period after discharge using univariate and multivariable regression analyses. No adjustment was made for multiplicity. RESULTS Of the 2,541 patients with schizophrenia (mean age: 41.2 years; 57% male; 59% black) who were identified, approximately 89% were "discharged to home self-care." Compared with the 60- to 0-day period before the index inpatient admission, greater mean adherence as measured by PDC was observed during the 0- to 60-day period immediately following discharge (0.46 vs. 0.78, respectively). The mean PDC during the overall 6-month preadmission period was lower than during the 6-month postdischarge period (0.53 vs. 0.69; P < 0.001). Compared with the 0- to 60-day postdischarge period, schizophrenia-related health care costs were significantly lower during the 61- to 120-day postdischarge period (mean: $2,708 vs. $2,102; P < 0.001); the primary cost drivers were rehospitalization (mean: $978 vs. $660; P < 0.001) and pharmacy (mean: $959 vs. $743; P < 0.001). Following the initial 60-day period, both all-cause and schizophrenia-related costs declined and remained stable for the remaining postdischarge periods (days 121-365). CONCLUSIONS Although long-term (e.g., 365-day) adherence measures are important, estimating adherence over shorter intervals may clarify the course of vulnerability to risk and enable clinicians to better design adherence/risk-related interventions. The greatest risk of rehospitalization and thus greater resource utilization were observed during the initial 60-day postdischarge period. Physicians should consider tailoring management and treatment strategies to help mitigate the economic and humanistic burden for patients with schizophrenia during this period.