| To Register SMDM Homepage | |
|
|
|
||||
PURPOSE: Cost-effectiveness RCTs, where the principal aim is to determine both the difference in effect and the resource costs of alternative treatment strategies, are increasingly common. Whilst accepted methods exist for meta-analysis of measures of effectiveness, combining cost-effectiveness results is more complex, and cannot be performed without access to the individual trial datasets.
Dyspepsia is a common condition in primary care, for which there are a number of investigative and empirical therapeutic options. A collaborative group has been meeting intermittently since 1996, prospectively registering RCTs with the aim of determining the cost-effectiveness of Helicobacter pylori ‘test and treat’ compared to endoscopy- based management and empirical acid-suppression.
METHODS: Individual patient data meta-analysis pooling effects of endoscopy, acid-suppression and H.pylori ‘test & treat’ on dyspepsia symptoms and costs. Standardized unit costs (UK national standard costs 2001) were applied to resource utilization at patient level and net benefit calculated at patient level using INB =λ Δ E -Δ C. Effect, and net benefit were pooled at study level and a sensitivity analysis on the ceiling ratio (λ).
RESULTS: Five trials that compared H.pylori ‘test and treat’ with endoscopy were pooled. Although a small effect difference was found in favour of endoscopy, Relative Risk of dyspepsia recurrence = 0.95, 95% confidence interval (CI) 0.92 to 0.99), this was not cost-effective. Even at a λ of $1750 per patient ‘cured’ INB remained negative at -$282 (-$324to -$240). Five trials that compared empirical acid suppression with endoscopy were pooled. No significant difference was found, Relative Risk of dyspepsia recurrence = 1.04, 95% confidence interval 0.97 to 1.11), Endoscopy was not cost-effective, INB remained negative at -$156(-249 to -116) at λ =0, and was insensitive to increasing λ.
CONCLUSIONS: Endoscopy is not cost-effective compared with H.pylori test and treat, or empirical acid suppression. IPD meta-analysis can be used to unpick complex RCTs with both cost and effect data. Using incremental net benefit as the variable to pool maintains the cost-effect correlation at patient level. INB is normally distributed, and can be summarised across studies as a continuous variable.
See more of Poster Session - CEA: Methods and Applications; Health Services Research
See more of The 26th Annual Meeting of the Society for Medical Decision Making (October 17-20, 2004)