|
Cost utility analysis of survival with epoetin-alfa versus placebo in stage IV breast cancer |
Martin S C, Gagnon D D, Zhang L, Bokemeyer C, Van Marwijk Kooy M, van Hout B |
|
|
Record Status 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. Health technology The comparison of epoetin-alpha, a recombinant human erythropoietin, with placebo among patients with Stage IV breast cancer.
Study population To be included in the original study by Littlewood et al. (see Other Publications of Related Interest), the participants were required to be at least 18 years of age with a solid or non myeloid haematological malignancy, to be anaemic (as defined by the authors) and to be receiving or due to receive nonplatinum-containing chemotherapy for a further 3 to 6 cycles.
Setting The setting was secondary care. The trial was described as multi-centred, primarily European, and did not include any sites in the USA.
Dates to which data relate The effectiveness results and resource use came from the original study, which was published in 2001. The prices used were for the year 2000.
Source of effectiveness data The effectiveness data were derived from a single study.
Link between effectiveness and cost data Survival data were collected prospectively 12 months after the original study was completed. The costs were calculated retrospectively, using UK treatment patterns from data obtained from physician surveys and nationally compiled data.
Study sample The sample size was not determined in advance. Instead, the sample was formed by selecting a sub-set of 55 patients with breast cancer from the 375 patients in the original trial who had been diagnosed with a variety of different cancers. The authors acknowledged that the sample selection method was of concern and that the study was not powered to evaluate the outcome of survival. The intervention group comprised 36 patients (65.5%) and the placebo group 19 patients (34.5%).
Study design The original study was a randomised controlled trial (RCT), stratified by baseline haemoglobin level and by tumour type. The trial was described as multi-centred. The authors stated that the trial was double-blind and that the patients were randomised 2:1 to treatment and placebo. Survival follow-up was recorded for 12 to 36 months post-study. Further details are provided in the original study (see Other Publications of Related Interest).
Analysis of effectiveness The primary end point of the original trial was the proportion of patients requiring a blood transfusion. The secondary end points were change in haemoglobin level and quality of life. The protocol was subsequently amended to collect and assess survival data prospectively. Survival was the outcome measure for the current study. It was not stated whether the intervention and placebo groups were comparable.
Effectiveness results The effectiveness results were not reported. The original study by Littlewood et al. (see Other Publications of Related Interest) reported Kaplan-Meier estimates. These showed a trend in overall survival favouring the intervention treatment, (log-rank p=0.13).
Clinical conclusions The authors of the original study were encouraged by the survival results and proposed that a further trial be carried out specifically to investigate their findings. The current economic evaluation was based solely on these survival results.
Measure of benefits used in the economic analysis The measures of benefits used were the expected life-years (LYs) and the quality-adjusted life-years (QALYs). Patient utilities were not collected in the follow-up period, so the utility weights were taken from another study (Brown et al., see Other Publications of Related Interest). These values were obtained from 30 UK oncology nurses using the standard gamble technique, for a number of health states associated with advanced breast cancer. These corresponded to four treatment phases in the trial, that is, active treatment, follow-up care, supportive care, and terminal care. The average length of time the patients in each group spent in each phase was calculated and summed to give the total LYs. The benefits were discounted. The uncertainty of the net benefits was investigated using a bootstrapping procedure, by randomly assigning utility values at each iteration.
Direct costs The cost boundary adopted was that of the health service. The costs included blood transfusions, epoetin-alpha usage, chemotherapy duration, diagnostic tests and procedures, radiotherapy, drug therapy, palliative surgical procedures, inpatient and ambulatory care services, pain control and palliative care, patient monitoring and terminal care. The unit costs of transfusions and epoetin-alpha usage were applied directly to individual patient use. All other unit costs were estimated using a study based on the opinions of 17 of 56 physicians invited to describe treatment and resource use in the UK for Stage IV breast cancer patients. The costs were then derived from nationally compiled data sources, while the drug costs were obtained from the British National Formulary.
The average costs were calculated for each treatment group and discounting was applied. The costs were in year 2000 prices. The NHS Hospital and Community Inflation Index was used for conversions, as required. The costs were prorated to 14-day intervals to facilitate the use of Kaplan-Meier Sample Average (KMSA) estimators.
Statistical analysis of costs The costs were treated as point estimates. As the follow-up data were incomplete, KMSA estimators were used to estimate the mean costs. Survival follow-up data were collected for between 1 and 3 years, and for the analysis these were truncated to the shortest maximum follow-up period per treatment group. Variance in the costs between physicians was investigated using a bootstrapping procedure.
Indirect Costs The indirect costs were not included.
Sensitivity analysis Sensitivity analyses were used to explore analytical methods and to extrapolate from primary data. One-way sensitivity analyses were undertaken on the discount rate for benefits, by applying utility weights of 1.0 at all treatment phases and by eliminating truncation of survival follow-up across the treatment groups. Four other scenarios were examined:
all patients surviving the first 2 months of the follow-up phase were given a second course of chemotherapy;
all censored patients were assumed not to have survived, so supportive care costs were reassigned backward in time from that date;
the utility values gathered from administering the SF36 during the original clinical trial were applied; and
the analyses were re-run for all 375 patients in the original study.
Estimated benefits used in the economic analysis The discounted LYs were 1.5718 for the intervention group and 0.9409 for the placebo group. The net benefit of 0.6309 favoured the intervention group.
The discounted QALYs were 1.0375 for the intervention group and 0.5570 for the placebo group. The net benefit of 0.4805 also favoured the intervention group.
The benefits were discounted at a rate of 1.5%.
Cost results The total cost per patient was 10,768 for those receiving epoetin-alpha and 6,515 for those receiving placebo.
The costs were discounted at a rate of 6%.
The net cost of treatment with epoetin-alpha was 4,253 (95% confidence interval: -1,252 - 9,078).
Synthesis of costs and benefits The costs and benefits were combined by calculating incremental cost-effectiveness ratios (ICER) and cost-utility ratios (ICUR). The ICER was 6,741/LY and the ICUR was 8,851/QALY. It was also calculated that the probability of a positive net benefit was 99%, and that the probability of achieving an ICUR of less than or equal to 30,000/QALY was 94%.
The first two scenarios in the sensitivity analysis (that all surviving patients had a second course of chemotherapy and that all censored patients were assumed to have died) increased the ICUR 1.5 times above the baseline analysis value. The probabilities of a positive net benefit were scarcely changed, but the probabilities of the ICUR remaining under the threshold value of 30,000/QALY were 75% (second-line chemotherapy) and 90% (censored patients died), respectively. For the final scenario, in which analyses were re-run for all 375 patients in the original study, the ICUR increased by more than 4 times and the probability value was 29%.
Authors' conclusions From the data for patients with Stage IV breast cancer receiving nonplatinum-containing chemotherapy, compared with placebo, there is a high probability of favourable cost-utility outcomes for treating anaemia with epoetin-alpha in this patient group. A caveat was applied to these findings. Further research is required to ascertain if the findings are replicable and to assess their generalisability. The cost of drug and patient care requirements, consequent to increased survival, were identified as the main drivers of higher costs in the intervention group.
CRD COMMENTARY - Selection of comparators The validity of the choice of a placebo as the comparator for epoetin-alpha relies on it being an adjunctive rather than a substitute therapy. By using a placebo, the active value of the treatment can be observed.
Validity of estimate of measure of effectiveness The basis of the analysis was an RCT, which was not powered to evaluate survival as an outcome. The study sample for the cost-utility analysis was not selected a priori, but was a sub-group analysis. The intervention and placebo groups were not shown to be comparable at analysis. Although stratified by baseline haemoglobin level, the groups were not stratified by variables that influence survival. The sensitivity analysis attempted to ameliorate these weaknesses.
Validity of estimate of measure of benefit The estimation of benefits was derived from the effectiveness data, using a combination of the KMSA estimator of net benefit and bootstrapping simulations to examine uncertainty in the data. In supporting their use of utility weightings obtained from oncology nurses, the authors noted the general controversy over whose utility values should be used.
Validity of estimate of costs All the categories of cost relevant to the perspective adopted were included in the analysis. The costs and the quantities were not reported separately. Some resource use quantities were taken from the study data. Other resource use quantities were obtained from a survey of physicians' opinions, which had a 30% response rate. The analysis was based on KMSA estimates for net costs and net benefits. Where resource use was derived from the survey, it was assumed to be identical between the groups. The costs were established from prices in published sources. Neither the costs nor the discount rate were investigated in the sensitivity analyses.
Other issues The authors made only limited comparisons of their findings with other studies. Due to the nature of the sub-group analysis, the results are not suitable for generalising to other patient groups prior to further definitive research. No rationale was given for the choice of the parameters investigated in the sensitivity analyses. The authors stated that adverse events were not included in the study for two reasons. First, the number reported that were suspected to be related to epoetin-alpha was low. Second, there was no significant difference between the groups in the number of adverse events.
Implications of the study The authors were encouraged by their results and the possible significance of treating anaemia in cancer on the prolongation of patient survival. They emphasised the interim nature of these results and the need for further research.
Source of funding Supported by Johnson & Johnson Pharmaceutical R&D, LLC, Raritan (NJ), USA.
Bibliographic details Martin S C, Gagnon D D, Zhang L, Bokemeyer C, Van Marwijk Kooy M, van Hout B. Cost utility analysis of survival with epoetin-alfa versus placebo in stage IV breast cancer. PharmacoEconomics 2003; 21(16): 1153-1169 Other publications of related interest Littlewood TJ, Bajetta E, Nortier JW, Vercammen E, Rapoport B, Epoetin Alfa Study Group. Effects of epoetin alfa on hematologic parameters and quality of life in cancer patients receiving nonplatinum chemotherapy: results of a randomized, double-blind, placebo-controlled trial. Journal of Clinical Oncology 2001;19:2865-74.
Brown R E, Hutton J, Burrell A. Cost effectiveness of treatment options in advanced breast cancer in the UK. Pharmacoeconomics 2001;19:1091-102.
Indexing Status Subject indexing assigned by NLM MeSH Adult; Anemia /drug therapy /etiology; Antineoplastic Agents /adverse effects /therapeutic use; Breast Neoplasms /drug therapy /pathology; Cost-Benefit Analysis; Double-Blind Method; Drug Administration Schedule; Endpoint Determination; Epoetin Alfa; Erythropoietin /administration & Female; Hematinics /administration & Humans; Injections, Subcutaneous; Neoplasm Staging; Quality of Life; Recombinant Proteins; Survival Analysis; Time Factors; Treatment Outcome; dosage /economics /therapeutic use; dosage /economics /therapeutic use AccessionNumber 22004008038 Date bibliographic record published 31/08/2004 Date abstract record published 31/08/2004 |
|
|
|