|Cost-utility of risperidone compared with standard conventional antipsychotics in chronic schizophrenia
|Oh P I, Lanctot K L, Mittmann N, Iskedjian M, Einarson T R
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.
The use of risperidone (RIS) to ameliorate the negative symptoms of schizophrenia. Risperidone is a benzisoxazole derivative with combined serotonin and dopamine receptor blocking properties.
The study population comprised patients suffering from chronic schizophrenia. This was defined as at least two years of more or less continuous thought disturbance, with the disturbance including pronormal, active and residual phases.
The setting was not explicitly stated. The economic study was carried out in Canada.
Dates to which data relate
The effectiveness evidence was gathered from studies published between 1966 and 1995. The price year was 1997.
Source of effectiveness data
The effectiveness evidence was mainly derived from a review of published studies and experts' opinions. The authors also conducted a single study to derive the utility weights used in the analysis.
Link between effectiveness and cost data
The costing was carried out on a different sample of patients from that used in the effectiveness analysis.
A sample of 32 volunteers was used to derive the utility weights used in the cost-utility analysis. The sample was predominantly male (90.6%) with a mean age of 35.6 (+/- 8.8) years (range: 24 - 59). The mean duration of schizophrenia in this patient group was 13.3 (+/- 7.1) years (range: 1 - 33). The method used to select the sample was not reported. Power calculations to determine the sample size were not carried out.
The effectiveness analysis was based on a single group of patients who voluntarily provided the estimates of the outcomes. No follow-up was required.
Analysis of effectiveness
All 32 patients provided data on utilities. During 30-minute one-to-one interviews, the patients assessed the utility of different health states associated with schizophrenia. The method used was the standard gamble technique. The scenarios to be assessed were moderate paranoid delusions in a hospitalised patient, mild delusional symptoms in a hospitalised patient (one for each drug), mild delusional symptoms in a patient living in the community, and drug side effects including EPSs. An internal check was carried out, as the utility values were also collected on a visual analogue scale to confirm that the patients understood the task.
The average utility rating was 0.83 (+/- 0.06) for moderate symptoms;
the average utility rating was 0.91 (+/- 0.03) for mild symptoms with RIS;
the average utility rating was 0.90 (+/- 0.04) for mild symptoms with HAL, HAL-DEC and FLU;
the average utility rating was 0.84 (+/- 0.04) for mild symptoms with HAL, HAL-DEC or FLU plus EPSs;
the disutility of hospitalised patients was 0.07; and
the disutility for EPS was 0.06.
The internal check confirmed the validity of the results.
This part of the effectiveness analysis was used to calculate the overall quality-adjusted life-years (QALYs) associated with each drug treatment.
A decision analytic model was used to estimate the costs and utility values associated with the treatments. The typical patient considered in the model was a 40-year-old inpatient with a 10- to 20-year duration of paranoid schizophrenia. The patient also had an initial Brief Psychiatric Rating Scale score of 55 or a Positive and Negative Symptoms Scale for schizophrenics score of 90 (range: 60 - 120), indicating a moderate degree of psychotic symptoms. The patient started the therapy with the prescribed drugs. The possible events were then the development of tolerability (presence or absence of treatment-limiting side effects), the development of extra-pyramidal symptoms (EPSs), success or failure, discharge from hospital, and relapse. The time horizon of the model was one year.
Outcomes assessed in the review
The outcomes assessed in the review and used as model inputs were the probabilities of efficacy, dropout due to lack of efficacy, dropout due to adverse events, and EPSs with all drugs.
Study designs and other criteria for inclusion in the review
Only randomised controlled trials were included in the review. The inclusion criteria of the review were:
original clinical trial;
random allocation to treatment groups;
one group was administered either oral RIS (4 to 10 mg/day), HAL-DEC or FLU;
the trial was controlled by either placebo or active comparator medication;
the patients were women or men aged 18 years or over;
the patients were diagnosed with schizophrenia using standardised criteria;
acceptable criteria had been used to define chronic schizophrenia;
the length of the treatment was at least 4 weeks; and
appropriate standardised instruments were used to measure the outcomes.
Primary studies were excluded if information on medications, patients, hospitalisation, dropouts and so on, could not be extracted.
Sources searched to identify primary studies
MEDLINE, EMBASE and International Pharmaceutical Abstracts were searched from 1966 to 1995. The search was restricted studies conducted in humans and reported in the English language. The keywords used were "schizophrenia" and "risperidone" (or "haloperidol decanoate" or "fluphenazine decanoate"). All papers and review articles were cross-referenced manually.
Criteria used to ensure the validity of primary studies
Studies not meeting the inclusion criteria were not considered.
Methods used to judge relevance and validity, and for extracting data
Two raters assessed the validity of the primary studies to be included in the review.
Number of primary studies included
Seventeen primary studies were used in the review of the effectiveness evidence.
Methods of combining primary studies
Data obtained from the primary studies were combined through a meta-analysis, using a random-effects model. As a result, 95% confidence intervals (CIs) were reported.
Investigation of differences between primary studies
The authors investigated the differences between the primary studies and ensured that the populations of the primary studies were similar.
Results of the review
The average efficacy rate was 0.67 (95% CI: 0.60 - 0.75) with RIS, 0.50 (95% CI: 0.32 - 0.68) with HAL, 0.46 (95% CI: 0.28 - 0.65) with HAL-DEC, and 0.39 (95% CI: 0.29 - 0.50) with FLU.
The dropout rate due to lack of efficacy was 0.14 (95% CI: 0.06 - 0.21) with RIS, 0.28 (95% CI: 0.09 - 0.47) with HAL, 0.02 (95% CI: 0 - 0.06) with HAL-DEC, and 0.02 (95% CI: 0 - 0.05) with FLU.
The dropout rate due to adverse events was 0.07 (95% CI: 0.05 - 0.08) with RIS, 0.06 (95% CI: 0.03 - 0.10) with HAL, 0.03 (95% CI: 0 - 0.07) with HAL-DEC, and 0.05 (95% CI: 0 - 0.11) with FLU.
The rate of EPS was 0.26 (95% CI: 0.20 - 0.33) with RIS, 0.48 (95% CI: 0.31 - 0.65) with HAL, not assessed with HAL-DEC, and 0.29 (95% CI: 0.07 - 0.51) with FLU.
Methods used to derive estimates of effectiveness
Expert opinion was used to assess some of the outcome measures not found in the published literature. These outcomes were the rates of discharge if symptoms improved and relapse within one year.
Estimates of effectiveness and key assumptions
The rate of discharge if symptoms improved was 0.81 (range: 0 - 1) and the relapse rate within one year was 0.16 (range: 0 - 1). It was also assumed that a successfully discharged patient would live in a supervised residential care environment, and any relapses were assumed to occur near the end of the evaluation period.
Measure of benefits used in the economic analysis
QALYs were used as the benefit measure in the economic analysis. The utility values derived from a sample of patients were combined with survival data derived from the review of the literature. The overall QALYs for each treatment were obtained using a decision model.
Discounting was irrelevant as all the costs were incurred over one year. The unit costs were reported separately from the quantities of resources used. The cost items included in the economic analysis were drug acquisition, psychiatric visits, community nursing visits, community social worker visits, community care manager visits, community residential care per day, hospital stay for acute admission and for prolonged admissions beyond 8 weeks, and relapse. The cost/resource boundary adopted was that of the government payer in Canada. The costs were estimated using actual data, which were mainly derived from provincial health insurance plans. The resource use was estimated on the basis of clinical practice in Canada. The price year was 1997.
Statistical analysis of costs
No statistical analysis of the costs was carried out.
The indirect costs were not included in the analysis.
One-way sensitivity analyses were carried out on all variables included in the decision model, to assess the robustness of the estimated cost-utility ratios. The 95% CIs were used for the effectiveness data, while all the costs were varied over a range of 50 to 150% of the baseline costs. A threshold analysis was also carried out. This calculated the point at which the conclusions of the analysis shifted from one strategy to the other, or became neutral.
Estimated benefits used in the economic analysis
The expected QALYs over one year were 0.87 with RIS, 0.83 with HAL, 0.84 with HAL-DEC, and 0.83 with FLU.
The expected one-year costs were Can$69,855 with RIS, Can$76,353 with HAL, Can$78,388 with HAL-DEC, and Can$82,264 with FLU.
Synthesis of costs and benefits
An incremental cost-utility analysis was carried out to combine the costs and benefits of each treatment. However, the incremental cost-utility ratios were not calculated as the RIS-based treatment was dominant (more effective and less costly) over the other interventions. The dominance of RIS was not affected by variations in the cost data, while the QALYs were more sensitive to variations in the efficacy rates and utility values.
Risperidone (RIS) proved to be a cost-effective treatment for patients with chronic schizophrenia. It was associated with less costs and higher effectiveness than other currently used treatments.
CRD COMMENTARY - Selection of comparators
The authors justified their choice of the comparators. HAL, HAL-DEC and FLU represented widely used antipsychotics in Canada. You should assess whether they represent widely used treatments in your own setting.
Validity of estimate of measure of effectiveness
The effectiveness measures were mainly estimated from a review of primary studies. The search methods and the inclusion criteria were reported in detail. Only randomised controlled trials providing good-quality data were considered. A meta-analysis was conducted to combine the estimates from the primary studies. Confidence intervals were provided and were used in the sensitivity analyses. Some assumptions made by experts were also required, but this was due to the lack of reliable published data. The ranges were used for the sensitivity analysis. The utility values were derived from a small primary collection of data, obtained from patients who voluntarily participated in the study. The authors noted that measurements carried out on a larger sample of patients would have been preferable. Uncertainty around these estimates was dealt with by carrying out sensitivity analyses. The estimated cost-utility ratios were robust to both the assumptions and utility values. The authors acknowledged that adherence to prescribed regimens was not considered in the model, but the compliance rates were expected to be higher with RIS than with the other treatments, thus supporting the dominance of RIS.
Validity of estimate of measure of benefit
QALYs were used as the benefit measure in the economic analysis. No discounting of the benefits was required, as the time horizon of the analysis was one year. The overall benefits were obtained using a decision model. The use of QALYs allows the comparison of the treatment with other interventions funded in the health care system.
Validity of estimate of costs
The perspective adopted in the study was stated and all the relevant categories of costs appear to have been included in the analysis. The authors stated that a societal perspective was not adopted, as estimates of the indirect costs were unavailable. A complete breakdown of the costs was provided. Also, the unit costs were reported and the price year was given. This enhanced transparency and generalisability, and simplified any reflation exercise to other settings. The costs were treated deterministically, but sensitivity analyses were carried out, varying any cost parameter used in the model. The dominance of RIS was not sensitive to wide variations of the unit costs. The cost estimates were specific to the Canadian setting, but they were reported in full.
The authors compared their findings with those from other studies, which confirmed the results of the present analysis. The issue of the generalisability of the study results to other settings was addressed. Several sensitivity analyses were carried out and the unit costs were reported separately for the resources used, thus enhancing the external validity of the analysis. The results of the sensitivity analyses were presented in detail. A population of patients with chronic schizophrenia was considered in the study and this was reflected in the conclusions of the authors. The authors noted that their findings should not be generalised to the whole population of patients suffering from schizophrenia, especially those patients in the early stages of the disease.
Implications of the study
The authors point out that further comparisons between RIS and other atypical antipsychotics should be carried out. In addition, local systems and networks should be activated in order to reduce hospitalisation and increase cost-savings. This was generally a well-designed and clearly reported study.
Source of funding
Supported by Janssen-Ortho Canada.
Oh P I, Lanctot K L, Mittmann N, Iskedjian M, Einarson T R. Cost-utility of risperidone compared with standard conventional antipsychotics in chronic schizophrenia. Journal of Medical Economics 2001; 4: 137-156
Subject indexing assigned by CRD
Adult; Antipsychotic Agents /administration & Canada; Cost-Benefit Analysis; Fluphenazine /administration & Haloperidol /administration & Humans; Patient Compliance; Risperidone /administration & Schizophrenia /drug therapy; Treatment Outcome; dosage; dosage; dosage; dosage /economics
Date bibliographic record published
Date abstract record published