Adherence to Treatment With Antipsychotic Medication and Health Care Costs Among Medicaid Beneficiaries With Schizophrenia
Gilmer TP, Dolder CR, Lacro JP, Folsom DP, et al.;
Commented by , 20 Apr 2004
Aim of the study
Non-adherence to medication is one of the most important problems in the treatment of schizophrenia. Most patients who stop treatment relapse and schizophrenic relapses can have dramatic consequences for both patients and families. It is also a major hope that - due to their lower risk of extrapyramidal side-effects than high-potency conventional drugs - the “atypical” antipsychotics improve compliance.
However, neither the determinants of non-adherence nor the amount of non-compliance in every day care nor the impact of the use of "atypical" antipsychotics on compliance is clearly understood. We do also not know much about the costs associated with non-adherence. The authors therefore examined the relationship between adherence to treatment with antipsychotic medication and health expenditures. Furthermore, they tried to identify risk factors predictive of noncompliance.
Methods
The Medicaid eligibility and claims data from 1998 to 2000 for the County of San Diego (California) and pharmacy records were available for the assessment of adherence to treatment with antipsychotic drugs. Adherence was calculated according to the cumulative possession ratio which is the number of days medications were available for use divided by the number of days patients were eligible for Medi-Cal.
The following categories of one-year adherence were derived from the cumulative possession ratio: nonadherent (ratio=0.00–0.49), partially adherent (ratio=0.50–0.79), adherent (ratio=0.80–1.10), and excess medication fillers (ratio >1.10).
Excess filling means that the patients had requested more medication than necessary. Then the authors examined risk factors, hospitalizations, and costs associated with the different levels of adherence using regression models. One of these risk factors was the use of typical/atypical antipsychotic medication.
Results
The main result of the study was that only forty-one percent of the included patients with schizophrenia were adherent to treatment with their antipsychotic medications while 24% were nonadherent, 16% were partially adherent, and 19% were excess fillers. Patients who were adherent had lower rates of psychiatric hospitalization (14%) than nonadherent (35%) and partially adherent (24%) patients, but also than excess fillers (25%).
Similar results were found for the rates of medical hospitalization which were lower for compliant patients (7%) than for noncompliant patients (13%) or excess fillers (12%). Compliant patients also had clearly lower hospital costs than the other groups. However, as expected pharmacy costs were higher among compliant patients and excess fillers compared to nonadherent or partially adherent patients.
Important risk factors for non-compliance were African American or Latino background, living alone and a history of drug abuse. One of the most surprising findings was that compliance was not associated with the use of atypical antipsychotic medication with the exception of clozapine.
Discussion
The strength of this article is the assessment of adherence in a real world setting and using a big database. The high rates of non-compliance despite the widespread use of atypical antipsychotic medications in the US is alarming.
On the other hand there were also high rates of excessive filling of antipsychotic prescriptions in a substantial proportion of the patients and this excess filling was associated with higher rates of hospitalisation, as well. Such excess filling may result from unsatisfactory responses to antipsychotics requiring frequent changes of medication or situations in which drugs are sold, stolen or lost.
However, probably the most surprising finding was that there were no significant differences in terms of compliance between typical and atypical antipsychotics. Thus, the superiority in terms of tolerability of these drugs may not be as high as many of us think. Another reason might be a methodological problem. In this study patients were not randomly allocated to treatment.
Therefore, those who received typical medication might have been a selective group of patients who had already received typicals for a long time and who had tolerated them well.