|Cost effectiveness of long-acting risperidone injection versus alternative antipsychotic agents in patients with schizophrenia in the USA
|Edwards N C, Locklear J C, Rupnow M F, Diamond R J
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 study compared the following treatment options for patients with schizophrenia:
oral risperidone (average dose 3.8 mg/day),
olanzapine (oral antipsychotic agent),
long-acting injectable haloperidol depot (typical antipsychotic agent), and
quetiapine, ziprasidone and aripiprazole (oral atypical antipsychotic drugs).
Further details of the drugs and specific doses were not reported.
Economic study type
The target patient population comprised community-dwelling patients with schizophrenia who had previously experienced a relapse necessitating hospitalisation. No further inclusion or exclusion criteria were reported.
The setting was the community. The economic study was carried out in the USA.
Dates to which data relate
Relevant published literature was accessed electronically in July 2004. Most of the effectiveness data were derived from studies published between 1993 and 2003. Further databases (e.g. the Consumer Health Sciences database) were accessed in 2003. The dates to which unpublished data (derived from various clinical trials) referred were not reported. The cost data were derived from various official sources published between 1992 and 2004. Most medical costs were reported for the price year 2003, while some costs (e.g. medication costs) were reported for the fiscal year 2004.
Source of effectiveness data
The effectiveness data were derived from a review and synthesis of completed studies, augmented with unpublished data from clinical trials and expert opinion where data were not available.
A decision analytic model was constructed to evaluate the cost-effectiveness of long-acting risperidone for patients with schizophrenia using Microsoft Excel 2002 (Microsoft Corp., Redmond). The time horizon of the model was 1 year. In the model, long-acting risperidone was compared with oral risperidone, olanzapine, quetiapine, ziprasidone, aripiprazole and the long-acting injectable haloperidol depot. The weight of each drug was estimated by its market share in the authors' setting. The authors reported that the structure of the model was based on that of a published study (Glazer and Ereshefsky 1996, see 'Other Publications of Related Interest' below for bibliographic details).
Outcomes assessed in the review
The input parameters used in the model were compliance rates, relapse rates, the frequency of relapse, the duration of relapse, and adverse event rates. Compliance rates for long-acting risperidone, oral risperidone, olanzapine, quetiapine, ziprasidone, aripiprazole, haloperidol depot and blended oral atypical (to compare long-acting risperidone with oral atypical antipsychotic drugs overall) were compared. Relapse rates and frequency of relapse (per relapsing patient) distinguished between those that did and did not require hospitalisation with atypical and typical antipsychotic agents. In terms of the adverse event rates, extra pyramidal side effects and weight gain experienced with the drug treatments were compared.
Study designs and other criteria for inclusion in the review
Sources searched to identify primary studies
Primarily, the authors searched PubMed for relevant medical literature. They also searched the Consumer Health Sciences database and unpublished data from clinical trials.
Criteria used to ensure the validity of primary studies
Methods used to judge relevance and validity, and for extracting data
Number of primary studies included
Overall, 19 primary studies provided effectiveness evidence.
Methods of combining primary studies
Compliance rates were adjusted to take the differential compliance of atypical and typical agents, as well as the differential compliance of long-acting injectable versus oral agents, into consideration. The compliance rates for long-acting risperidone were adjusted using modelled data. Relapse rates were adjusted to take the differential efficacy of atypical and typical agents into account.
Investigation of differences between primary studies
The authors do not appear to have investigated differences between the primary studies.
Results of the review
The results of the review were too numerous to report here, thus only the main results are presented.
The proportion of patients experiencing a relapse requiring hospitalisation in 1 year was 66% for haloperidol depot, 41% for oral risperidone, olanzapine, quetiapine, ziprasidone and aripiprazole, and 26% for long-acting risperidone.
The proportion of patients with an exacerbation not requiring hospitalisation was 60% for haloperidol depot, 37% for oral risperidone, olanzapine, quetiapine, ziprasidone and aripiprazole, and 24% for long-acting risperidone.
The mean number of days of relapse requiring hospitalisation per patient per year was 28 for haloperidol depot, 18 for oral risperidone, olanzapine, quetiapine, ziprasidone and aripiprazole, and 11 for long-acting risperidone. The mean number of days of exacerbation not requiring hospitalisation was 8 for haloperidol depot, 5 for oral risperidone, olanzapine, quetiapine, ziprasidone and aripiprazole, and 3 for long-acting risperidone.
Methods used to derive estimates of effectiveness
Effectiveness data that were not available in the literature were based on expert opinion. Modified Delphi panel techniques were used to elicit input from clinical experts.
Estimates of effectiveness and key assumptions
The duration of relapse not requiring hospitalisation was 5.0.
Measure of benefits used in the economic analysis
The health benefit measures used were the number of relapses averted per patient per year and the number of relapse days averted per year. These measures were derived from the outcomes.
Adopting the health care system perspective, the direct costs included in the analysis were for inpatient care (hospitalisation, day hospital and emergency room), outpatient care (physician office visit, mental health clinic visit, home health care, social or group therapy meetings, nutritionist) and medications (long-acting risperidone, oral risperidone, olanzapine, quetiapine, ziprasidone, aripiprazole blended oral atypical agent, haloperidol depot and benzodiazepine). The costs were derived from actual published data, while the estimation of the quantities of resources used was based on expert opinion. All unit costs were reported. However, the authors seem to have reported medical costs for the year 2003, while the costs of medication and injection administration, medication utilisation and dose distribution for long-acting risperidone were reported for the year 2004. Since the time horizon of the model was 1 year, discounting was not relevant.
Statistical analysis of costs
The costs were treated deterministically.
The indirect costs were not included in the analysis.
A sensitivity analysis was carried out to test variability in the data. All input parameters of the model were investigated. Although the type of sensitivity analysis was not explicitly stated, the authors seem to have carried out a one-way sensitivity analysis. The authors used 95% confidence intervals (CIs) available in the literature, and ranges based on clinical input where the 95% CI were not available.
Estimated benefits used in the economic analysis
Compared with oral atypical antipsychotic agents, long-acting risperidone resulted in a 15.3% reduction in relapse rate requiring hospitalisation and a 12.9% reduction in relapse rate not requiring hospitalisation.
The incremental benefits of long-acting risperidone compared with oral atypical antipsychotic agents were 0.6 relapses averted per patient per year, 6.5 days of relapse requiring hospitalisation saved per patient per year, and 1.7 days of relapse not requiring hospitalisation saved per patient per year.
The total treatment costs were reported per patient per year.
The total cost was $20,769 for long-acting risperidone, $20,929 for oral risperidone, $22,194 for olanzapine, $21,276 for quetiapine, $21,028 for ziprasidone, $21,837 for aripiprazole, $21,493 for blended oral atypical agent, and $28,992 for haloperidol depot.
Using long-acting risperidone rather than an oral atypical antipsychotic agent resulted in $161 of health care savings per patient per year compared with oral risperidone, $259 compared with ziprasidone, $508 compared with quetiapine, $1,068 compared with aripiprazole, and $1,425 compared with olanzapine.
Compared with the class of oral atypical antipsychotic agents overall (weighting oral atypical agents by market share), long-acting risperidone resulted in $724 of health care savings per patient per year.
Synthesis of costs and benefits
The costs and benefits were combined using an incremental cost-effectiveness ratio.
Long-acting risperidone was found to be dominant, i.e. more effective and less costly than all other treatment options.
The sensitivity analysis demonstrated the most influential parameters of the model. If the relapse rate requiring hospitalisation of compliant patients was changed to the upper 95% CI, long-acting risperidone became more costly in comparison with an oral atypical antipsychotic agent or oral risperidone. The model was sensitive when:
the hospitalisation relapse rate of partially compliant patients and non-compliant patients was changed to the lower 95% CI;
the compliance rate of non-compliant patients was changed to the lower 95% CI;
the compliance rates of compliant patients was changed to the upper 95% CI;
the frequency of relapse requiring hospitalisation and the frequency of relapse not requiring hospitalisation were changed to the upper 95% CI; and
the relapse rate not requiring hospitalisation of compliant patients was changed to the upper 95% CI.
The model was also sensitive when the duration of relapse requiring hospitalisation took the minimum value in literature (i.e. 18.2 days). In all the above cases, long-acting risperidone proved to be more costly than oral atypical antipsychotic agents or oral risperidone, and incremental cost-effectiveness ratios were calculated.
When all the cost parameters were varied by +/- 25%, the results were only sensitive to the cost of hospitalisation. When the hospitalisation cost was decreased by 10 or 25%, long-acting risperidone became a more costly option than oral risperidone; when it was decreased by 25% long-acting risperidone became more costly than all other treatment options.
The use of long-acting risperidone is predicted to result in better clinical outcomes and lower total health care costs than its comparators (i.e. oral risperidone, olanzapine, quetiapine, ziprasidone, aripiprazole and haloperidol depot). Long-acting risperidone may, therefore, be a cost-saving therapeutic option for patients with schizophrenia.
CRD COMMENTARY - Selection of comparators
The selection of the comparators was explicitly justified. You should decide if they represent widely used technologies in your own setting.
Validity of estimate of measure of effectiveness
It appears that a systematic review of the literature was undertaken. The study designs, other inclusion criteria for the review, and possible differences between the primary studies were not investigated. However, the authors conducted an extensive sensitivity analysis to test variability in the data, which enhances the generalisability of the results. Where effectiveness data were not available in the literature, the authors used expert opinion. A Delphi panel of two experts was assembled to derive the estimates of effectiveness, but the authors did not report how the members of the Delphi panel were selected.
Validity of estimate of measure of benefit
The authors used the number of relapses averted per patient per year and the number of relapse days averted per year as the measures of benefit in the economic analysis. Theses outcome measures did not provide a sense of the wider non-health benefits associated with treatment. It would appear that the benefits were not discounted given the short time period of the study analysis.
Validity of estimate of costs
Adopting the health care system perspective it seems that all the relevant categories of costs were included in the analysis. The unit costs were reported, thus enhancing the reproducibility of the results to other settings. However, as already noted, not all costs were reported for the same price year and this factor should be taken into account in any future reflation exercise. Resource use was mainly derived on the basis of expert opinion. Statistical analyses of the resource quantities or costs were not conducted (i.e. the costs were treated deterministically). However, the robustness of the estimates used was investigated in a sensitivity analysis using appropriate ranges. This facilitates the interpretation of the study findings. Discounting was not necessary, as the costs were incurred during less than 2 years.
The authors did not compare their findings with those from other studies, so it is not known how far their results agree with other published results. However, this might have been due to the lack of published studies comparing all available treatment options. The structure of the model was straightforward, and the authors felt that it could be easily adapted to other settings to reflect different practice patterns and costs. The authors do not appear to have presented their results selectively.
The authors reported a number of limitations to their study. First, the indirect costs were not included in the analysis, although the inclusion of such costs would most probably strengthen the conclusions. Second, research published after July 2004 was not included in the base-case analysis. However, the authors compared the results of the sensitivity analysis with those of some studies published after the time of their PubMed search, and found consistency in their findings. The authors acknowledged that some effectiveness estimates were based on expert opinion and information contained in databases, and not on peer-reviewed published literature, owing to the lack of available studies. On the other hand, they reported that selection bias of the patient population might have affected the results of some peer-reviewed studies on compliance for depot antipsychotic agents.
Implications of the study
The authors did not make any explicit recommendations for changes in policy or practice, or for the need for future research. Their discussion, however, highlighted some areas where more information is needed.
Source of funding
Supported by Janssen Medical Affairs LLC.
Edwards N C, Locklear J C, Rupnow M F, Diamond R J. Cost effectiveness of long-acting risperidone injection versus alternative antipsychotic agents in patients with schizophrenia in the USA. PharmacoEconomics 2005; 23(Supplement 1): 75-89
Other publications of related interest
Glazer WM, Ereshefsky LA. Pharmacoeconomic model of outpatient antipsychotic therapy in "revolving door" schizophrenic patients. J Clin Psychiatry 1996;57:337-45.
Subject indexing assigned by NLM
Antipsychotic Agents /administration & Cost of Illness; Cost-Benefit Analysis; Decision Trees; Hospitalization /economics /trends; Humans; Injections, Intravenous; Patient Compliance; Risperidone /administration & Schizophrenia /drug therapy /economics; United States; dosage /economics /therapeutic use; dosage /economics /therapeutic use
Date bibliographic record published
Date abstract record published