In the base case, TTG testing detected 28 of the 30 cases of coeliac disease present in the cohort of 1000 's'-IBS patients (assuming a prevalence of coeliac disease of 3% in 's'-IBS). Approximately one additional coeliac disease case was discovered with the serological panel compared with TTG testing alone ( Table 2 ). Without coeliac disease testing, the average 35-year-old patient with 's'-IBS experienced a remaining quality-adjusted discounted life expectancy of 18.12 QALYs. Coeliac disease testing with TTG or the panel both increased QALYs per subject compared with no coeliac disease testing, with only a relatively small gain by the serological panel over TTG testing alone ( Table 2 ).
The discounted lifetime cost of caring for IBS in the absence of coeliac disease testing was $11 835 per patient. Testing for coeliac disease with TTG alone increased the cost per patient by approximately 1%, and using the serological panel increased the cost per patient by approximately 2% ( Table 2 ). Thus, the aggregate cost of serological testing in all patients, followed by endoscopy and biopsy in those with positive serology, and including the cost of complications, was very small compared with the lifetime cost of IBS care.
Compared with no coeliac disease testing, TTG testing cost $4600 per case detected, and testing with the serological panel cost $8800 per case detected ( Table 2 ). The incremental cost of coeliac disease testing using the panel compared with TTG was $135 000 per incremental case detected, which can be understood by appreciating that the additional cost of $124 per subject for the panel strategy compared with the TTG strategy is incurred by all 1000 's'-IBS patients in the cohort in order to detect approximately one additional case of coeliac disease.
Compared with no coeliac disease testing, TTG testing cost $7400 per QALY gained, and testing with the serological panel cost $14 000 per QALY gained ( Table 2 ). As with the incremental cost per case detected, the incremental cost per QALY gained by testing with the panel compared with TTG alone was significant ($287 000/QALY gained). That is, TTG testing alone achieved most of the gains in QALYs afforded by using the panel, and testing with the panel compared with TTG alone incurred a substantial incremental cost for only a small increase in quality-adjusted life expectancy.
Up-front EGD with biopsy in all patients along with confirmatory antibody panel testing in those with abnormal histology detected all cases of coeliac disease, yielded an average quality-adjusted discounted life expectancy of 18.1378 QALYs, and resulted in a total discounted cost per subject of $937 higher than no testing. Thus, compared with no testing, the cost per case detected was $31 200, and the cost per QALY gained was $54 800. However, the incremental cost per case detected was $450 000 compared with TTG testing and $776 000 compared with panel testing. Both serology-based strategies gained more QALYs than up-front EGD with biopsy ( Table 2 ), due to differences in EGD-related complications. Since the serology-based strategies were also less costly, they dominated up-front EGD with biopsy with regard to cost per QALY gained.
In patients with concomitant upper gastrointestinal symptoms who were already scheduled to undergo EGD, adding a small bowel biopsy and confirmatory antibody panel testing if histology was abnormal detected all coeliac disease cases, gained 0.019 QALYs per subject, and cost an additional $392 per subject compared with undergoing EGD without biopsy. Thus, in those already undergoing diagnostic EGD, the cost of adding a biopsy and serological panel testing after abnormal histology was $13 000 per case detected and $20 700 per QALY gained.
Given the prohibitively high incremental costs per incremental case detected by up-front EGD with biopsy compared with serological testing, and the dominance of serological over up-front endoscopic testing in terms of cost per QALY gained, subsequent analyses focused on serological testing.
The clinical and economic outcomes of serological testing were most affected by the prevalence of coeliac disease in the population with 's'-IBS and the increase in health state utility afforded by diagnosing and treating coeliac disease with a gluten-free diet ( Table 3 ). The test performance characteristics and the risks of EGD affected the incremental comparisons between the panel and TTG alone ( Table 3 ).
In a two-way sensitivity analysis, we examined the combinations of coeliac disease prevalence in 's'-IBS and utility gain with coeliac disease diagnosis under which TTG testing might be a cost-effective intervention. Given a utility gain of 0.025 with coeliac disease diagnosis (comparable with our base case assumption), TTG testing cost under $50 000 per QALY gained even at a coeliac disease prevalence in 's'-IBS as low as 0.5% (Figure 2). With a much more modest utility gain of 0.005 with coeliac disease diagnosis, TTG testing cost under $50 000 per QALY gained at a coeliac disease prevalence in 's'-IBS as low as 2%, and under $100 000 per QALY gained at a coeliac disease prevalence in 's'-IBS as low as 1.1% (Figure 2). At coeliac disease prevalence rates in 's'-IBS lower than 0.5-1%, the cost per QALY gained by TTG testing rose sharply for utility gains with coeliac disease diagnosis in the range of 0.025-0.005. Alternatively, if the utility gain with coeliac disease diagnosis was very low, for example, 0.001, then the cost of TTG testing in 's'-IBS generally exceeded $100 000 per QALY gained in the range of coeliac disease prevalence in 's'-IBS that is clinically plausible (Figure 2).
Two-way sensitivity analysis of the cost per quality-adjusted life-year (QALY) gained by testing 35-year-old patients with suspected irritable bowel syndrome (IBS) with tissue transglutaminase antibody (TTG) as a function of the prevalence of coeliac disease, and the utility gain associated with diagnosing coeliac disease. With a modest utility gain of 0.005 with coeliac disease diagnosis, TTG testing cost under $50 000 per QALY gained at a coeliac disease prevalence in suspected IBS as low as 2%.
The results of the Monte Carlo simulation ( Table 4 ) suggest that the base case results are robust. That is, the results do not change significantly even when the uncertainty inherent in all the model inputs is accounted for. Compared with no testing, TTG testing achieved an incremental cost-effectiveness ratio of <$50 000 per QALY gained in 98% of iterations. Compared with no testing, using the panel achieved an incremental cost-effectiveness ratio of <$50 000 per QALY gained in 88% of iterations, and <$100 000 per QALY gained in 98% of iterations. The incremental cost-effectiveness ratio of the panel vs. TTG testing was <$50 000 per QALY gained in 46% of iterations, and <$100 000 per QALY gained in 63% of iterations.
Aliment Pharmacol Ther. 2004;19(11) © 2004 Blackwell Publishing
Cite this: Serological Testing for Coeliac Disease in Patients With Symptoms of Irritable Bowel Syndrome: A Cost-Effectiveness Analysis - Medscape - Jun 01, 2004.