thiopurines and surgery (if required) over 10 years up to 2004. Eight health states were defined by intensity of therapy. We determined Markov transition probabilities between these states in 3-month cycles, healthcare costs (HC) and quality-adjusted life-years (QALYs). This cohort was then modeled to allow drug-refractory or pre-surgery patients to receive infliximab: either episodically (ET) in one-cycle treatments which were repeated where needed over 10 years, or as a maintenance treatment (MT) in responders for multiple cycles.
The transition probabilities of ST were applied to patients getting infliximab. The postinfliximab probabilities for enabling ET patients to transit to other health states as well as the fixed probability value of 0.989 that allowed continued infliximab administration in MT patients (while correcting for decay of response) were computed from published data. HC and QALYs in ET and MT were estimated for a time horizon of 10 years (discounted at 3%) and compared with those of ST patients. RESULTS: The mean cost (and QALYs gained)
per patient over 10 years was as follows: €12,967 (7.3339) for ST, €12,647 (7.6349) for ET, and €24,279 (7.6362) for MT. ST was associated with higher costs and less favorable outcomes and was thus dominated by ET. The incremental cost-effectiveness ratio (ICER) of MT over ST was €37,417/QALY gained and over ET was €8,999,876/QALY gained.
CONCLUSIONS: Our model suggests that when compared with ST, ET with infliximab increases QALYs and appears to be a cost-saving intervention in ulcerative colitis, while MT which had the highest QALYs is cost-effective but expensive. At current drug costs, MT does not provide good value for money when compared with ET.