|
Routine use of granulocyte colony-stimulating factor is not cost-effective and does not increase patient comfort in the treatment of small-cell lung cancer: an analysis using a Markov model |
Chouaid C, Bassinet L, Fuhrman C, Monnet I, Housset 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 Granulocyte colony stimulating factor (G-CSF) in the treatment of small-cell lung cancer (SCLC).
Study population Patients newly treated for SCLC.
Setting The practice setting was a secondary care institution and the economic analysis was performed in France.
Dates to which data relate Effectiveness data for strategies 2 and 3 were based on data published during the period 1990 to 1996. For strategy 1 effectiveness data were collected during 1993. The price years were not stated.
Source of effectiveness data The evidence of final outcomes for strategies 2 and 3 was based on a review of previously completed studies. For strategy 1 effectiveness data were obtained from a single study.
Link between effectiveness and cost data The costing was undertaken retrospectively on the same patient sample as that used in the effectiveness analysis.
Study sample The strategy one study involved 39 patients, who underwent 172 chemotherapy cycles during 1993 at the authors' institution. No power calculations to determine the sample size were reported.
Study design The study was a retrospective case series carried out in a single centre. The length of follow-up was until discharge and there appears to have been no loss to follow-up.
Analysis of effectiveness The following health outcomes were considered: the probability at any given time of treatment of a first episode of FN, a recurrence, and premature cessation of chemotherapy with previous FN and without previous FN.
Effectiveness results The probability of a first episode of FN was 17.5%, the probability of a recurrence was 37.1%, and the probability of premature cessation of chemotherapy with previous FN was 5.2%. The probability of premature cessation of chemotherapy without previous FN was 17.1%.
Clinical conclusions The retrospective review of the medical records of patients with SCLC provided important inputs for the model.
Modelling A Markov model was used to consider the three therapeutic strategies (chemotherapy dose reduction after FN and non-use of G-CSF, and primary or secondary prophylactic use of G-CSF) and two sets of rewards: the expected monetary cost of going through the Markov model and a comfort utility scale.
Outcomes assessed in the review Effectiveness estimates for strategies 2 and 3 were obtained from a review, which assessed the reduction in the incidence and duration of FN.
Study designs and other criteria for inclusion in the review Randomised clinical trials were included in the review.
Sources searched to identify primary studies A search of MEDLINE was undertaken.
Criteria used to ensure the validity of primary studies Methods used to judge relevance and validity, and for extracting data The judgement criteria were not stated. Data from the primary studies had already been summarised in an existing meta-analysis.
Number of primary studies included The analysis of published data identified three randomised clinical trials that evaluated the effectiveness of primary G-CSF therapy.
Methods of combining primary studies The results of the individual primary studies had already been summarised in a meta-analysis.
Investigation of differences between primary studies Results of the review Primary G-CSF therapy reduced the incidence of FN from 68.3% to 38.3%. Indirect analysis, authorised by the framework of two of these studies, showed that secondary G-CSF therapy led to a 40% reduction in the incidence of FN. As the impact of G-CSF on the hospital FN stay cannot be clearly stated, the authors made the baseline assumption of no impact.
Measure of benefits used in the economic analysis Utilities were determined from a standardised interview conducted by one of the authors with 10 consecutively treated SCLC patients who had experienced a single episode of FN. They were shown a hypothetical case scenario in which the probability of FN after a chemotherapy cycle was 18%, the time spent in hospital for FN was 6 days, and the use of G-CSF reduced the risk of FN by 50%. Using a rating scale method, patients were asked to score from 0 to 1 the perceived discomfort of three states: FN without G-CSF, G-CSF without FN and FN despite G-CSF.
Direct costs The economic analysis adopted the health care payer's perspective and took into account direct costs. The direct costs of managing FN included hospitalisation and medical costs, which were added together to calculate an average cost per episode of FN. Hospitalisation costs were based on a per diem cost. Medical costs were determined retrospectively by chart review of all treated SCLC patients hospitalised for FN during the study period. The cost of a G-CSF course was calculated on a basis of 10 days of treatment in an outpatient setting using a daily dose adapted to body weight, and included ancillary costs charged to the patient.
Indirect Costs Direct costs were not included.
Sensitivity analysis Sensitivity analyses were performed on the probabilities of a first FN episode, FN recurrence, FN reduction rate, the cost of a FN episode, the cost of a G-CSF course and utility values.
Estimated benefits used in the economic analysis On the comfort utility scale, >never= remained the best strategy, with a comfort value of 378 U, compared with 365 U for >CSF if FN=, and 327 U for >systematic CSF=.
Cost results Using the monetary utility scale, the >never= strategy was the best choice. With this strategy, each patient entering the process would cost $4,875, compared with $5,816 for >CSF if FN=, and $7,690 for >systematic CSF=.
Synthesis of costs and benefits A synthesis was not undertaken by the authors because it was not relevant since, for the 'never' strategy, incremental costs were negative and incremental benefits were positive. Up to a 49% probability of a first FN episode, the 'never' strategy was less costly than either strategy using G-CSF; above this threshold value, systematic CSF was preferable. 'Never' was also the best strategy up to a probability of relapse of 60%, whereas at higher values the least costly strategy was 'CSF if FN'.
The sensitivity analysis of the efficacy of G-CSF showed that, even with a very large reduction in FN, the 'never' strategy remained the best choice. The 'never' strategy remained the least costly as long as CFN was less than $6077.60, or if G-CSF cure cost more than $1,072.80. A three-way sensitivity analysis of the utilities showed that the 'never' strategy should be preferred, except for patients who considered aplasia markedly 'uncomfortable' and G-CSF therapy perfectly 'comfortable'. Routine use of G-CSF during SCLC chemotherapy is not justified by clinical benefits, improved patient comfort, or economic considerations.
CRD COMMENTARY - Selection of comparators The reason for the choice of comparator is clear.
Validity of estimate of measure of benefit The effectiveness evidence was derived from the published literature and a retrospective medical chart review. Insufficient detail was provided about the methodology of the literature review, which might not have been systematic. Also, it is not clear whether the review of the medical charts included a sufficient number of patients. The authors noted that one limitation of their study was that the outcome measure (completion of initial planned chemotherapy) assumed that the three strategies had the same clinical efficacy. This assumption requires further investigation. The authors used time trade-off and standard gamble as the reference methods to assess utilities. They acknowledged the possible limitations of these methods in terms of their validity and reliability, and they reported attempts to overcome some of these limitations in the process of quantifying patient preferences for different health states. However, it is not clear whether a sample of 10 patients was appropriate for these purposes.
Validity of estimate of costs Resource quantities were not reported separately from prices, but adequate details of cost estimation were given. The authors reported cost results from other studies, which compared favourably to their findings. However, it was noted that these results require careful consideration as they depend on the costs identified.
Other issues Various sensitivity analyses were performed to account for the uncertainties in the data. The authors made appropriate comparisons with other studies.
Bibliographic details Chouaid C, Bassinet L, Fuhrman C, Monnet I, Housset B. Routine use of granulocyte colony-stimulating factor is not cost-effective and does not increase patient comfort in the treatment of small-cell lung cancer: an analysis using a Markov model. Journal of Clinical Oncology 1998; 16(8): 2700-2707 Other publications of related interest Messori A, Trippoli S, Tendi E. G-CSF for the prophylaxis of neutropenic fever in patients with small cell lung cancer receiving myelosuppressive antineoplastic chemotherapy: meta-analysis and pharmacoeconomic evaluation. Journal of Clinical Pharmacy and Therapeutics 1996;21:57-63.
Indexing Status Subject indexing assigned by NLM MeSH Antineoplastic Combined Chemotherapy Protocols /therapeutic use; Carcinoma, Small Cell /economics /therapy; Cost-Benefit Analysis; Decision Trees; Granulocyte Colony-Stimulating Factor /adverse effects /economics /therapeutic use; Humans; Lung Neoplasms /economics /therapy; Markov Chains; Neutropenia /etiology AccessionNumber 21998001172 Date bibliographic record published 31/03/2001 Date abstract record published 31/03/2001 |
|
|
|