|
Health economic comparisons of rofecoxib versus conventional nonsteroidal antiinflammatory drugs for osteoarthritis in the United Kingdom |
Moore R A, Phillips C J, Pellissier J M, Kong S X |
|
|
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 health technology under investigation was rofecoxib, a new drug for the treatment of patients suffering from osteoarthritis. It is a selective COX-2 inhibitor and represents one of the two drugs in the "Coxibs" subclass of the M1A class based on the World Health Organisation's classification.
Economic study type Cost-effectiveness analysis.
Study population The study population included patients suffering from osteoarthritis.
Setting The setting was community. The economic study was carried out in the UK.
Dates to which data relate Effectiveness and resource use data were gathered between 1989 and 2000. The price year was 1999.
Source of effectiveness data The effectiveness data were derived from a review of the literature.
Modelling A decision-tree model was used to describe the GI-related health care utilisation of osteoarthritis patients receiving either rofecoxib or conventional NSAIDs for up to a one-year period. The decision-tree structure accounted for transitions among health states and considered each drug equally effective. Three possible scenarios regarding effectiveness evidence were considered:
the first model (Analysis 1) was based on data about observed perforations, ulcers, and bleeds (PUBs);
the second and third model were based on endoscopic ulcers with adjustments for silent ulcers of 85% (Analysis 2) and 40% (Analysis 3).
Outcomes assessed in the review The outcomes assessed in the review and used as inputs in the three models were the probabilities of NSAID-related GI adverse event, the rate of real and suspected PUB given GI adverse events, the effectiveness of treatment given minor GI adverse event, and the death rate after PUB. Resource use probabilities were also estimated in the review: gastroprotective agents (GPAs) use, hospitalisation given a PUB, inpatient investigation of suspected PUB, and surgery given hospitalisation.
Study designs and other criteria for inclusion in the review Inclusion criteria for study designs were not stated. Several study designs were included in the review (randomised controlled trials, prospective observational studies, etc.). Effectiveness data were mainly derived from 8 double-blind, randomised, phase IIb/III trials.
Sources searched to identify primary studies 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 Approximately 37 primary studies were considered for the estimates of probability values used in the decision model.
Methods of combining primary studies Ranges of estimated values were presented only for the resource use data. For the remaining data, probability values from primary studies were not combined and the authors reported only the point estimation.
Investigation of differences between primary studies Results of the review The results of the review were as follows:
Among the values estimated in the review, the probabilities of NSAID-related GI adverse events for rofecoxib were 0.299 in model 1, 0.248 in model 2, and 0.248 in model 3.
For conventional NSAIDS the probabilities of NSAID-related GI adverse events were 0.295 in model 1, 0.366 in model 2, and 0.366 in model 3.
For rofecoxib, the rates of PUB, given GI adverse events were 0.050 in model 1, 0.039 in model 2, and 0.154 in model 3.
For conventional NSAIDs the rates of PUB, given GI adverse events were 0.091 in model 1, 0.117 in model 2, and 0.47 in model 3.
The probability of hospitalisation, given a PUB, was 0.207 (range: 0.056 - 0.67).
The GPA use was 0.206 (range: 0.171 - 0.418).
Measure of benefits used in the economic analysis The benefit measures used in the economic analysis were the number of PUB events avoided and the number of years of life saved. Life-years were calculated from age and life expectancy data and the number of deaths avoided in a cohort of 10,000 patients, and were not discounted.
Direct costs Costs were not discounted since they were incurred within one year. Quantities and costs were not reported separately and their estimation was derived using modelling. The source of the cost estimation was the literature, and data were also confirmed by discussion with financial staff at three large acute NHS Trusts from three different regions in England. The resource/cost boundary was that of the hospital. The cost items included in the study were medications (drugs and dispensing), GP consultations, outpatient gastroenterologist consultations, investigations, inpatient days, and intensive care days. Resources were gathered between 1989 and 2000. The price year was 1999.
Statistical analysis of costs No statistical analysis was reported.
Indirect Costs Indirect costs were not included.
Sensitivity analysis One-way sensitivity analyses were performed in order to investigate the robustness of the results to variability in data, given that many estimates were derived from the literature. Several parameters were varied: the cost of NSAID comparators, the probability of hospitalisation given a PUB, the probability of surgery given a hospitalisation, the rate of prophylactic GPA use, the probability of inpatient investigation for a suspected PUB, etc.
Estimated benefits used in the economic analysis The number of PUB events avoided with rofecoxib (with respect to NSAIDs) was 107 in Analysis 1, 313 in Analysis 2, and 1,251 in Analysis 3. The average life expectancy used in the three models was 19.1 years, but life-years gained with rofecoxib and conventional NSAIDs were not explicitly reported.
Cost results The expected total costs per day were 0.91 and 0.59 for rofecoxib and conventional NSAIDs, respectively, in Analysis 1, 1.03 and 1.13 in Analysis 2, and 1.36 and 2.52 in Analysis 3.
Iatrogenic costs (i.e. the costs caused by the treatment) and expected costs averted due to GI problems avoided with rofecoxib were also calculated. Iatrogenic costs were always lower for rofecoxib than for conventional NSAIDs: 1.15 versus 1.97 in Analysis 1, 1.30 versus 3.75 in Analysis 2, and 1.72 versus 8.41 in Analysis 3.
Expected costs adverted due to GI problems avoided with rofecoxib were 0.17 in Analysis 1, 0.59 in Analysis 2, and 1.65 in Analysis 3.
Synthesis of costs and benefits Costs and benefits were combined by calculating the cost per PUB avoided and per year of life saved. In Analysis 1, the cost per PUB event avoided was 10,759 and the cost per year of life saved was 15,647. In Analyses 2 and 3, the rofecoxib strategy realised cost-savings compared to the conventional NSAIDs options with respect to both PUB events avoided and years of life saved. The sensitivity analyses indicated that in Analysis 1 the results were most sensitive to the rate of prophylactic GPA use. In Analyses 2 and 3, the variable with the greatest impact on the results was the probability of inpatient treatment for a PUB.
Authors' conclusions The analysis showed that, with the introduction of rofecoxib, a substantial reduction in the risk of GI complications for patients on conventional NSAIDs therapy is possible at only a modest additional cost to the NHS budget.
CRD COMMENTARY - Selection of comparators The rationale for the selection of the comparators was clear: the new drug was compared with NSAIDs, which are widely used for the treatment of osteoarthritis.
Validity of estimate of measure of effectiveness The effectiveness outcomes used as input parameters for the model were derived from a review of the literature that was non-systematic, leading to a potential for bias. However, the majority of the studies included were RCTs.
Validity of estimate of measure of benefit The benefit measure was derived from modelling. However, given that the treatment has a major effect on the quality of life, health state-related measures could have been more appropriate to measure the benefit of the strategies. Furthermore, the conduct and methods of the review were not satisfactorily reported.
Validity of estimate of costs All the costs relevant to the perspective adopted were accounted for in the model. However, the calculations of thresholds at which the drugs were considered cost-effective, were not clearly reported.
Other issues The authors did not address the issue of generalisability of the results to other settings, although sensitivity analyses were carried out. A possible limitation of the study may be that the baseline characteristics of the patients enrolled in the trials used as the main source of effectiveness data may not be representative of osteoarthritis patients in general.
Implications of the study The main implication of the study was that rofecoxib could represent an important advantage for the management of osteoarthritis, especially from the patient perspective, given that rofecoxib is associated with a significant reduction of GI complications and with the same effectiveness compared to NSAIDs.
Source of funding Supported by an educational grant from Merck & Co Inc., Whitehouse Station, New Jersey, USA.
Bibliographic details Moore R A, Phillips C J, Pellissier J M, Kong S X. Health economic comparisons of rofecoxib versus conventional nonsteroidal antiinflammatory drugs for osteoarthritis in the United Kingdom. Journal of Medical Economics 2001; 4: 1-17 Indexing Status Subject indexing assigned by CRD MeSH Adult; Aged; Anti-Inflammatory Agents, Non-Steroidal /adverse effects /drug effects /administration & Cost-Benefit Analysis; Drug Costs; Female; Great Britain; Humans; Male; Osteoarthritis /drug therapy /prevention & Stomach Ulcer /complications; control; dosage /economics AccessionNumber 22001008097 Date bibliographic record published 28/02/2002 Date abstract record published 28/02/2002 |
|
|
|