This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn.

Treatment of schizophrenia with risperidone versus standard antipsychotic therapy.

Treatment

Cost-effectiveness analysis.

146 patients were included; 52.1% were female, and 47.9% male. The mean age at index data was 40.8 years. The age range was between 17-84 years. All patients were intolerant to the standard neuroleptic therapy. The characteristics of the patient group were similar to the total population of patients receiving risperidone in Canada. 48.6 % of the patients were covered by SAP and 45.2 % were covered by Basic Health Care Program. Only 6.2 % of the patients were covered by 'other' health care programs.

Hospital and community. The study was carried out in Saskatchewan, Ontario, Canada.

Effectiveness, resources and price data were derived from five computerised databases (Health Insurance Registration File (HIRF), Prescription Drug Services Branch File(PDSBF), Hospital Services Branch File (HSBF), Medical Care Insurance Branch File (MCIBF) and Mental Health Services Branch File (MHSBF)) betweenJuly 1993 and June 1994.

Effectiveness data were derived from a single study.

Costing was retrospectively undertaken on the same patient sample as that used in the effectiveness study.

All patients who registered in the Saskatchewan Health Linkable Data File and had received 1 prescription for risperidone on or after 1 July of 1993 were included in the study. All 146 patients were intolerant to the standard neuroleptic therapy. 52.1 % were female, and 47.9 % male. The mean age at index data was 40.8 years. The age range was between 17-84 years. The mean observation period was 288 days. The observation period range was 189-354 days. No power calculation determined the sample size.

This was a before and after study. The patients were followed up for a maximum of one year. No loss to follow up or dropout rate were reported.

The effectiveness of the clinical study was based on intention to treat. The number of hospital admissions, physicians visits, visits to mental health services and length of hospital stay were measured as proxies for health outcomes.

The total number of hospitalisations was reduced 60.3% in theafter' period which meant an average decrease of 1.4 admissions per patient per year. The length of hospital stay (total cumulative days of hospitalisation) in theafter' period was decreased by 58.2% compared to the 'before' period, resulting in a decrease of 29 days of hospitalisation per patient per year. The decrease in hospital admissions was greater in the Basic Health Care Program with 2.1 fewer admissions per patient per year compared to Saskatchewan Assistance Plan (SAP)with 0.8 fewer admissions per patient per year. After the initialisation of risperidone the number of physician visits reduced from 2127 to 1654, of which 1348 and 1042 visits respectively were related to schizophrenia.

Overall, a decrease of 1.8 total visits per patient per year was observed in theafter' period. This decrease was primarily due to patients under the SAP plan (n=60) where a decrease of 10.8 visits per patient per year was realised. In contrast, an increase of 9.9 visits per patient per year was noted in the Basic Health Care group (n=46).

Not stated.

The effectiveness of the clinical study was based on intention to treat. The number of hospital admissions, physicians' visits, visits to mental health services and length of hospital stay were measured as proxies for health outcomes.

Direct health service costs were used. Some costs and resources were reported separately. Data were derived from the databases listed above: HIRF, PDSBF, HSBF, MCIBF and MHSBF. From these databases the authors calculated the cost of a hospital stay, the cost of a visit to the Mental Health Service Branch and the cost of prescriptions. Costs of physicians and drug dispenses were obtained from actual billing data. The method of estimation of costs of hospital stay was not stated. Cost data referred to 1993 and 1994. No discounting wasapplied.

The Wilcoxon signed rank test wasused and p values were reported.

Canadian Dollars (Can$)

No sensitivity analysis was carried out.

The total number of hospitalisations was reduced 60.3% in theafter' period which meant an average decrease of 1.4 admission per patient per year. The length of hospital stay (total cumulative days of hospitalisation) in theafter' period was decreased by 58.2% compared to the 'before' period, resulting in a decrease of 29 days of hospitalisation per patient per year. Decrease in hospital admissions was greater in the Basic Health Care Program with 2.1 fewer admissions per patient per year compared to SAP with 0.8 fewer admission per patient per year. After the initialisation of risperidone the number of physician visits reduced from 2127 (1348 related to schizophrenia) to 1654 (1042 related to schizophrenia).

Overall, a decrease of 1.8 total visits per patient per year was observed in theafter' period. This decrease was primarily due to patients under the SAP plan (n=60) where a decrease of 10.8 visits per patient per year was realised. In contrast, an increase of 9.9 visits per patient per year was noted in the Basic Health Care group (n=46).

The cost of physicians' services was Can$123,543 before risperidone and Can$92,131 after. The cost of psychiatrist visits was Can$76,867 in thebefore' period and Can$53,068 in theafter' period. Cost of risperidone in theafter' period was Can$150,145. The cost of depot antipsychotics was Can$13,060 in thebefore' period and Can$6,708 in theafter' period. The oral antipsychotics' cost was Can$25,196 in thebefore' period and Can$11,397 in theafter' period. The cost of antiparkinson medication was Can$6,295 in the before' period and Can$6,315 in theafter' period. All drugs' cost Can$92,992 before the administration of risperidone and Can$227,965 after.

The overall annual cost saving was Can$7,925 per patient per year after the initiation of risperidone therapy. The treatment with risperidone was the dominant strategy.

The use of risperidone was associated with an overall decrease in health care resource utilisation (Can$7,925 per patient per year). The reduction in hospital admissions, length of stay, and physician visits offset almost 8-fold the cost of risperidone therapy.

This study was based on a selected group of schizophrenic patients who failed, or became intolerant to, previous antipsychotic therapy. The before-and-after study design may also involve bias as differences in health outcomes might be due to the course of the disease rather than to the different drug treatments. Further, the health outcome measurement was based only on a limited number of dimensions; neither side effects nor quality of life were considered. Costs of physician visits and drug dispenses were obtained from billing data which may or may not reflect real costs. Some other costs were estimated by the authors in an unspecified way. No indirect costs were taken into account. Cost results came from a specific Canadian setting which also limits the generalisability of the study. A sensitivity analysis of costs would have been useful to overcome this limitation.

The study cast light on the need for the integrated clinical and financial Management Information System. This could be a powerful tool forcost-effectiveness analysis and for health policy-makers.

Janssen Pharmaceutica, Canada.