Testing for occult cancer in patients with idiopathic deep vein thrombosis--a decision analysis.

OBJECTIVE: To determine the effectiveness and cost-effectiveness of testing for occult cancer in idiopathic deep vein thrombosis (IDVT). DESIGN: Threshold analysis was performed on the risk-adjusted cancer prevalence in a cost-effectiveness model of ideal testing for selecting cancers with potentially desirable utility (candidate cancers). Decision analysis was employed to compare different testing programs for candidate cancers with that of no testing. Life expectancy (LE) of early- and late-detected cancers and costs of testing were the dimensions of utility. Cost-effectiveness was expressed as marginal cost per year of life saved. The perspective of the third payer was adopted, and a discount rate of 3% was applied to both costs and benefits. DATA SOURCES: Risk of cancer in IDVT, testing policies, test characteristics, and LE were gathered from literature. Costs were provided from our hospital rate book and accounting service. RESULTS: Ideal testing would support a gain of LE of 40 days or more for prostate, colon and bladder cancer in males and for colon, breast and endometrium cancer in females aged from 60 to 69 years. Testing females with colonoscopy and mammography in any sequence provides 70 days of life gained. Testing males with colonoscopy provides 27 days of life gained. Lower and older ages reduce testing effectiveness. The qualitative results are stable over plausible ranges of test characteristics, while variations in the value of benefit for early cancer diagnosis may modify the strategy. Incremental cost-effectiveness ranges from $1,789 to $ 6,979 per year of life gained. CONCLUSIONS: According to the effectiveness criterion adopted, the only worthwhile investigation strategy includes colon and breast cancer in females. Testing for colon cancer in males is desirable at a lower criterion of effectiveness. All the strategies are cost effective.