Cost-utility analysis of primary prophylaxis versus secondary prophylaxis with granulocyte colony-stimulating factor in elderly patients with diffuse aggressive lymphoma receiving curative-intent chemotherapy.

Abstract

The 2006 American Society of Clinical Oncology (ASCO) guideline recommended primary prophylaxis (PP) with granulocyte colony-stimulating factor (G-CSF) instead of secondary prophylaxis (SP) for elderly patients with diffuse aggressive lymphoma receiving chemotherapy. We examined the cost-effectiveness of PP when compared with SP. We conducted a cost-utility analysis to compare PP to SP for diffuse aggressive lymphoma. We used a Markov model with an eight-cycle chemotherapy time horizon with a government-payer perspective and Ontario health, economic, and cost data. Data for efficacies of G-CSF, probabilities, and utilities were obtained from published literature. Probabilistic sensitivity analysis (PSA) was conducted. The incremental cost-effectiveness ratio of PP to SP was $700,500 per quality-adjusted life-year (QALY). One-way sensitivity analyses (willingness-to-pay threshold = $100,000/QALY) showed that if PP were to be cost-effective, the cost of hospitalization for febrile neutropenia (FN) had to be more than $31,138 (2.5 × > base case), the cost of G-CSF per cycle less than $960 (base case = $1,960), the risk of first-cycle FN more than 47% (base case = 24%), or the relative risk reduction of FN with G-CSF more than 91% (base case = 41%). Our result was robust to all variables. PSA revealed a 10% probability of PP being cost-effective over SP at a willingness-to-pay threshold of $100,000/QALY. PP is not cost-effective when compared with SP in this population. PP becomes attractive only if the cost of hospitalization for FN is significantly higher or the cost of G-CSF is significantly lower.