|Cost-effectiveness and cost-utility analyses of finasteride therapy for the treatment of benign prostatic hyperplasia
|Canadian Coordinating Office for Health Technology Assessment
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.
Finasteride, 5 mg/day, for the treatment of benign prostatic hyperplasia (BPH) in elderly males.
Economic study type
Cost-effectiveness analysis and cost-utility analysis.
Males, aged above 60 and experiencing symptoms of BPH.
The practice setting was the hospital. The economic study was carried out in Ottawa, Ontario, Canada.
Dates to which data relate
Effectiveness and resource use data were derived from literature published between 1988 and 1995. The price date was not stated.
Source of effectiveness data
The estimate for final outcomes was based on a review of previously completed studies.
A decision tree was used to model the treatment options. Costs, effects and quality-adjusted-life-years (QALYs) were estimated over time horizons ranging from 2 to 15 years. A Monte-Carlo type simulation model was used to generate effectiveness probabilities. Average QALYs and costs were then estimated and compared with results from decision analysis.
Outcomes assessed in the review
The primary health outcome chosen by the authors was BPH symptom improvement, as measured by a symptom score, resulting from initiation with a particular therapy. The authors assumed that if an intervention were effective, then it would remain effective. The exception to this was TURP, where symptom scores can deteriorate.
Study designs and other criteria for inclusion in the review
Sources searched to identify primary studies
The sources used by the authors to search for data were not stated. However, the effectiveness estimates were taken from a BPH Clinical Practice Guideline that had reviewed 1,200 abstracts, extracted by searches on the Cumulated Index Medicus and MEDLINE.
Criteria used to ensure the validity of primary studies
The methodology employed in devising the BPH Clinical Practice Guideline was described.
Methods used to judge relevance and validity, and for extracting data
Number of primary studies included
Methods of combining primary studies
The BPH Clinical Practice Guideline had used a Confidence Profile Method to synthesise results. Estimates from 2 additional primary studies, including one randomised trial, were used to corroborate these results.
Investigation of differences between primary studies
Results of the review
The outcome probabilities (and 90% Confidence Intervals) associated with each treatment option can be summarised as follows:
Finasteride: mild or moderate symptoms 0.67 (0.54 - 0.78), no improvement 0.23, future TURP treatment 0.10 (0.09 - 0.12).
Watchful Waiting: mild or moderate symptoms 0.42 (0.31 - 0.55), no improvement 0.20 and future TURP treatment 0.38 (0.15 - 0.65).
TURP: mild 0.723 (0.54 - 0.90), mild with stress incontinence 0.021 (0.0175 - 0.025), mild with impotence 0.136 (0.034 - 0.324), no improvement 0.095, total urinary incontinence 0.010 (0.007 - 0.014) and death 0.015 (0.005 -0.033).
Measure of benefits used in the economic analysis
In the cost-effectiveness analysis, the effectiveness measure was used as the measure of benefits. In the cost-utility analysis, the measure of benefits was the quality-adjusted-life-year gained (QALY). A generic utility function of health status was derived from the Health Utility Index (Mark II), a multi-attribute health status classification system. Seven attributes are included: sensation, mobility, emotion, cognition, self-care, pain and fertility. Utility values were derived retrospectively from published patient-survey information and from the authors' opinion.
Direct costs were estimated from the perspective of the health care system. Costs and quantities were reported separately and included the cost of finasteride, medical labour, hospital and operating room use, diagnostic procedures and laboratory tests. The cost of subsequent therapy, including the treatment of side effects and re-treatment, was included in the analysis, but the cost of the initial examination, which was common to all three alternatives, was excluded. Costs incorporating provincial variations in prices and average length of stay were also specified. Cost information was taken from a range of Canadian federal and provincial agencies, including the Institute for Clinical and Evaluative Sciences (ICES) and the Canadian Institute for Health Information (CIHI). The date to which prices relate was not stated. All costs were discounted at 5%.
Statistical analysis of costs
A statistical analysis of costs was not performed.
Indirect costs were not included in the analysis.
To investigate the generalisability of the results, sensitivity analyses were conducted on effectiveness, treatment failure rate, cost estimate, utility values and the discount rate. High and low estimates of effectiveness for TURP, finasteride and WW were used in a multiway simple sensitivity analysis. In addition, two one-way analyses were conducted in which the effectiveness of WW was varied. High and low estimates of treatment failure rates for both finasteride and WW were used in a two-way simple sensitivity analysis. High and low estimates of the costs of the three treatment options were used in a multiway simple sensitivity analysis. Utility values were also varied. Utility for the 'moderate' health state was lowered whilst 'severe' state utility was kept constant. A two-way analysis was conducted in which 'moderate' and 'severe' utility values were lowered. The utility value for the 'mild-with-impotence' state was also lowered in a one-way analysis. Finally, the effects of setting the discount rate (for both costs and effects) at 0% and 10% were investigated.
Estimated benefits used in the economic analysis
The probability of symptom score improvements was reported in the effectiveness results. For the cost-utility analysis, incremental QALYs were calculated. Negative values are in parentheses. If the pre-treatment symptom severity was 'moderate', then the incremental QALYs resulting from finasteride over WW varied from 0.0163 with a 2-year horizon to 0.0911 with a 15-year horizon. For finasteride over TURP the figures were 0.0290 with a 2-year horizon and 0.1618 with a 15-year horizon. If the pre-treatment symptom severity was 'severe', then the incremental QALYs resulting from finasteride over WW varied from (0.0283) with a 2-year horizon to (0.1582) with a 15-year horizon. For finasteride over TURP the figures were (0.1198) with a 2-year horizon and (0.6686) with a 15-year horizon. The incremental QALYs resulting from TURP over WW were also reported. Treatment side effects were considered in the economic analysis, in terms of their effect on resource use, but they were assumed to have no effect on quality of life. The duration of benefits was considered up to fifteen years after the initial treatment. A discount rate of 5% was applied to both measures of benefit.
The national average cost for selected time horizons, were as follows:TURP was estimated to cost Can$5,676 with a 2-year horizon, increasing to Can$7,066 with a 15-year horizon. Finasteride was estimated to cost Can$2,028 with a 2-year horizon, increasing to Can$7,294 with a 15-year horizon. WW was estimated to cost Can$2,301 with a 2-year horizon, increasing to Can$3,264 with a 15-year horizon.
The incremental costs for selected time horizons, were as follows: for finasteride over WW, this cost increased from Can$(273) with a 2-year horizon to Can$4,029 with a 15-year horizon. For finasteride over TURP, this cost rose from Can$(3,648) with a 2-year horizon to Can$228 with a 15-year horizon.
The incremental cost of TURP over WW was also reported.
Synthesis of costs and benefits
The incremental cost-effectiveness ratios (ICERs), for selected time horizons, were as follows:for finasteride over WW the ICER ranged from Can$(30,290) with a 2-year horizon to Can$841,517 with a 15-year horizon. For finasteride over TURP the ICER fell from Can$37,156 with a 2-year horizon to Can$(4,370) with a 15-year horizon. The incremental ICER of TURP over WW was also reported.
The ratios for incremental cost per QALY, for selected time horizons, were as follows:
If pre-treatment symptoms were moderate:for finasteride over WW, the ratio increased from Can$(16,755) with a 2-year horizon to Can$44,222 with a 15-year horizon. For finasteride over TURP, the ratio increased from Can$(125,837) with a 2-year horizon to Can$1,406 with a 15-year horizon.
If pre-treatment symptoms were severe: for finasteride over WW, the ratio increased from Can$9,651 with a 2-year horizon to Can$(25,472) with a 15-year horizon. For finasteride over TURP, the ratio increased from Can$30,462 with a 2-year horizon to Can$(340) with a 15-year horizon.
The incremental cost-per QALY of TURP over WW was also reported. In general, the Monte-Carlo type simulation corroborated these findings, although the cost of finasteride therapy was found to be higher than the result from the decision analysis. The cost-effectiveness of finasteride was found to be highly sensitive to the effectiveness estimate for all therapies, the effectiveness estimate for WW in particular, the re-treatment rates and to the cost of WW and TURP, relative to the cost of finasteride.
The authors concluded that the cost-effectiveness of initiating treatment for BPH with finasteride, rather than WW or TURP, was dependent on the choice of time horizon. Finasteride was more cost-effective than WW if the time horizon was less than 4 years; finasteride was more cost-effective than TURP if the time horizon was less than 14 years.'Cost-effective' is defined here as being less costly and less efficacious. The cost-utility of finasteride was found to depend on the choice of time horizon, but also upon symptom severity at the start of therapy. The incremental cost-utility ratio of finasteride, relative to either comparator, was higher for patients with moderate, rather than severe, symptoms. Finasteride was cost-saving, relative to WW, if these patients had a life span of 3 years or less and was cost-saving, relative to TURP, if these patients had a life span of 14 years or less.
The comparators used in the analysis were chosen to represent current practice. You, the user of the database, should judge whether these are widely used technologies in your own setting.
The cost-effectiveness and cost-utility analyses were clearly and comprehensively reported, although the base year for costs was not stated. The authors noted that their results are highly sensitive to the choice of time horizon, the baseline symptom severity of patients, the effectiveness estimate for WW and to the incremental cost of finasteride. They address the issue of generalisability of their results, noting that the clinical and economic data used in the analysis, as well as clinical practice patterns, may vary in different settings. The unreliability of the data, in particular the limited effectiveness data for finasteride and the reprocessed information used to generate utility data, is also acknowledged.
Although the authors conducted sensitivity analyses on a range of variables, they did not specifically explore their admittedly "tenuous" assumption that the effectiveness of finasteride would be maintained, as long as it was taken. Had this been done, the sensitivity of the cost-effectiveness of finasteride to assumptions concerning its efficacy could have been made explicit. This might have been useful to inform both contemporary decisions and future clinical trial designs.
Implications of the study
When long-term efficacy data for finasteride becomes available, a better measure of the cost-effectiveness of finasteride, relative to alternative therapies, could be estimated.
Canadian Coordinating Office for Health Technology Assessment. Cost-effectiveness and cost-utility analyses of finasteride therapy for the treatment of benign prostatic hyperplasia. Ottawa, ON, Canada: Canadian Coordinating Office for Health Technology Assessment. 1995
Subject indexing assigned by CRD
Aged; Cost-Benefit Analysis; Finasteride /therapeutic use /economics; Male; Prostate /drug effects; Prostatectomy; Prostatic Hyperplasia /drug therapy; Quality-Adjusted Life Years
Date bibliographic record published
Date abstract record published