A model for analyzing the cost of the main clinical events after lung transplantation

Background: The aim of this project was to model clinically important event s experienced by lung transplant patients (from the day after transplant to 5 years or death) and costs associated with these events, and to assess th e economic impact of different immunosuppression therapies. Methods: The population comprised 356 lung transplant patients (223 heart-l ung, 102 single lung and 31 double lung) transplanted between April 1984 an d December 1997. AU patients received a cyclosporine-based triple-immunosup pression protocol. We designed a Markov model that included 3 time periods (0 to 6, 7 to 12, and 13 to 60 months), 5 clinical states (well, acute reje ction, cytomegalovirus infection, non-cytomegalavirus infection and bronchi olitis obliterans syndrome), and death. For the well state, cost elements w ere immunosuppression, prophylaxis, and routine clinic visits. For all othe r states, cost elements were diagnosis, treatment, and bed days/visits. We excluded costs of the procedure. Results: The monthly costs associated with the well state decreased over ti me, from pound1,778 ($2,658) in the first 6 months to pound 503 ($752) in m onths 7 to 12 and pound 350 ($523) after the first 12 months. The cost per event of the acute states remained reasonably constant over the 3 periods: pound1,850 ($2,766) for rejection, pound3,380 ($5,053) for cytomegalovirus, and pound2,790 ($4,171) for other infections. The average cost per patient , discounted at 6%, over 5 years was pound 35,429 ($52,966) (95% range, pou nd1,435 [$2,145] to pound 67,079 [$100,283]). This estimate is most sensiti ve to changes in immunosuppression. Substituting tacrolimus for cyclosporin e increased 5-year costs by 5%; substituting mycophenolate mofetil for azat hioprine increased 5-year costs by 26%. Conclusions: This model is valuable in estimating the effect of new immunos uppression agents on the costs of follow-up care.