|
Economic analyses of phase III cooperative cancer group clinical trials: are they feasible? |
Bennett C L, Golub R, Waters T M, Tallman M S, Rowe J M |
|
|
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-macrophage colony stimulating factor (GM-CSF).
Economic study type Cost-effectiveness analysis.
Study population The study population was a hypothetical cohort of adult patients (aged 55-70) with acute myelogenous leukemia.
Setting Hospital. The economic study was carried out in the USA.
Dates to which data relate The source of effectiveness data was a study published in 1995. Neither the dates for the collection of resource use data nor the price year were clearly stated.
Source of effectiveness data Effectiveness data were derived from a single study.
Link between effectiveness and cost data Effectiveness data were derived from a randomized controlled study. Cost data were based retrospectively on a small sample (7/117) of the patients involved in the randomized controlled study.
Study sample A total of 117 adult patients between >56 and 70 years of age with AML were eligible and originally included and randomised: GM-CSF group consisted of 60 patients and the placebo group contained 57 patients. Patients also had to have adequate hepatic, renal and cardiac function, no previous cytotoxic or radiation therapy, and morphological proof of de novo AML without an antecedent myelodysplasia. Analysis was restricted to patients who actually received GM-CSF or placebo; 18 patients did not receive study drug because of early death (14 patients), failure to achieve aplasia (3 patients), or physician decision (1 patient).
Study design The study was a randomized multicentre clinical trial (number of centres not specified). The loss to follow-up was not clearly reported.
Analysis of effectiveness The analysis of the clinical study was based on treatment completers only. The outcomes used in the analysis were median number of days of hospitalization from day 11 to discharge, days receiving GM-CSF or placebo, days with platelets less than 20,000/mm3, days requiring red blood cell transfusions and frequency of infections. The two arms were similar with respect to age, Eastern Cooperative Oncology Group performance status, percent of bone marrow blasts, circulating white blood cell count, circulating platelet count and FAB classification.
Effectiveness results The median number of days of hospitalization from day 11 to discharge, days receiving GM-CSF or placebo, days with platelets less than 20,000/mm3 , and days requiring red blood cell transfusions "did not differ significantly" between the two arms. However, the frequency of grade 4-5 infections (9.6% vs. 36.2%, p=0.002) and grade 3-5 infections (51.9% vs. 74.5%, p=0.0024) was lower among the patients in the GM-CSF group.
Modelling A decision analysis model was developed to estimate the economic cost and benefits of GM-CSF. Clinical strategies were modelled for patients aged >56-70 with AML, undergoing induction chemotherapy. The model described the clinical scenarios and potential outcomes (grade 3/4/5 infection develops or not).
Measure of benefits used in the economic analysis While the study did find significant treatment effects with respect to complete remission rates, survival rates and severe infection rates in favour of GM-CSF, the authors conservatively assumed that clinical outcomes during induction were similar between the GM-CSF and placebo groups.
Direct costs While detailed financial accounts for costs and charges for seven patients from the randomized controlled study were used to derive unit costs (based on internal cost to charge ratios) the total costs were derived using a model which incorporated resource use and infection rate data for patients who actually received either of the study "drugs". The price year was not clearly reported. The major quantities of resource use were analysed separately from the costs. The costs incurred after discharge were omitted from the analysis.
Sensitivity analysis One-way simple and threshold sensitivity analyses addressed the overall cost estimates assuming ranges of daily costs for patients who developed a severe infection, price of GM-CSF, duration of hospitalization, and frequency of development of infection (as reported in the literature).
Estimated benefits used in the economic analysis Cost results Overall, mean GM-CSF cost savings during the hospitalization for induction chemotherapy were $12,513 ($65,770 for GM-CSF versus $78,283 for placebo patients), with savings primarily resulting from lower pharmacy costs ($2,250, exclusive of the cost of GM-CSF), intensive care room costs ($1,720), laboratory costs ($930), and radiology costs ($1,350). The most important parameter in the sensitivity analyses was frequency of infections. If the frequency of severe infection (grade 4/5) is 18% and unaffected by GM-CSF, then CSF would save less than $2,000. If the rate of severe infections is 28% with CSF and 32% in the placebo group then CSF use would save less than $3,900/patient. Threshold analyses indicated that GM-CSF use would be cost-neutral at a daily cost of $906, about 4.5 times as great as the actual daily costs.
Synthesis of costs and benefits Authors' conclusions The economic model indicated that the group treated with GM-CSF was estimated to have lower costs of care, associated with lower frequencies of serious infections and lower overall infection-related costs. Therefore, GM-CSF use appears to be cost-effective over a wide range of plausible scenarios.
CRD COMMENTARY - Selection of comparators No comparator, other than placebo, was used. GM-CSF had recently been approved by the FDA and it is not clear whether another health technology could, or should, have been used as a comparator
Validity of estimate of measure of benefit The authors assumed that there was no difference in clinical outcomes between the GM-CSF and placebo groups although the empirical data did support significant treatment effects in favour of GM-CSF. It is noteworthy that the analysis was based on the treatment completers principle.
Validity of estimate of costs Unit cost data were based only on a small sample of the overall patient population (7 from 117). Not all resource quantities were reported separately from prices. Resources used after discharge were not included in the economic analysis.
Other issues Although the issue of generalisability to other countries was not addressed, the authors noted that the generalisability of the results can be questioned on the grounds of the difference between real-life practice and the conditions of the clinical trial (from which the data were drawn), and the limitations of the source of unit costs (only one centre). The latter, however, were reported to be similar to the corresponding figures for centres in California, Nebraska and elsewhere.
Source of funding Dr Bennett is a recipient of a Senior Career Development Award from the Veterans Affairs. Additional support was obtained from an unrestricted grant from Immunex Corporation.
Bibliographic details Bennett C L, Golub R, Waters T M, Tallman M S, Rowe J M. Economic analyses of phase III cooperative cancer group clinical trials: are they feasible? Cancer Investigation 1997; 15(3): 227-236 Indexing Status Subject indexing assigned by NLM MeSH Adult; Aged; Clinical Trials, Phase III as Topic /economics /methods; Cost-Benefit Analysis; Costs and Cost Analysis; Feasibility Studies; Granulocyte-Macrophage Colony-Stimulating Factor /economics /therapeutic use; Humans; Insurance, Health, Reimbursement; Leukemia, Myeloid, Acute /economics /therapy; Middle Aged; Models, Statistical; Multicenter Studies as Topic /economics /methods; Probability; Quality of Life; Retrospective Studies; Sensitivity and Specificity; United States AccessionNumber 21997000755 Date bibliographic record published 31/10/1998 Date abstract record published 31/10/1998 |
|
|
|