Correspondence

Is the Use of t-PA as Compared with Streptokinase Cost Effective?

N Engl J Med 1995; 333:1009-1010October 12, 1995

Article

To the Editor:

The conclusion by Mark et al. in their analysis of the cost effectiveness of thrombolytic therapy for acute myocardial infarction (May 25 issue)1 that the routine use of accelerated tissue plasminogen activator (t-PA) in patients with acute myocardial infarction would provide 3.5 million additional years of life after myocardial infarction appears to be a gross overstatement of the annual survival benefit. The authors' data do not support the contention that each of the 250,000 potentially eligible recipients of this therapy would gain 14 additional years of life from this treatment.

Instead, the data from the Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries (GUSTO) trial indicate that an additional 1 percent of the 250,000 patients — that is, 2500 patients per year — would survive the infarction as a result of receiving the thrombolytic therapy. Given the reasonable though unproved estimate that each such survivor would live an additional 14 years, routine use of accelerated t-PA would provide an annualized survival benefit of 35,000 life-years.

Eric A. Rose, M.D.
Columbia University, New York, NY 10032

1 References

1. 1

   Mark DB, Hlatky MA, Califf RM, et al. Cost effectiveness of thrombolytic therapy with tissue plasminogen activator as compared with streptokinase for acute myocardial infarction. N Engl J Med 1995;332:1418-1424
   Full Text | Web of Science | Medline

To the Editor:

Mark et al. assessed the cost effectiveness of t-PA, as compared with streptokinase, in patients with acute myocardial infarction. The authors used a Cox proportional-hazards model based on 4379 patients from the Duke Cardiovascular Disease Database to extend 1-year survival data an additional 14 years and a Gompertz function to extrapolate the tail of this survival curve. They found an increase in life expectancy of 0.14 year per patient (51 additional days of life). From these results, and extrapolating an incremental, undiscounted cost of $2,845 for each patient receiving t-PA, the authors concluded that the administration of t-PA rather than streptokinase to approximately 250,000 eligible patients with acute myocardial infarction in the United States would cost $500 million each year and provide 3.5 million additional years of life. Now, 250,000 patients times 0.14 additional year of life per patient is only 35,000 and not 3.5 million additional years of life. On the other hand, the real cost would be 250,000 times $2,845 (the incremental cost of t-PA per patient), or $711,250,000.

Lee1 stated that cost-effectiveness analyses require that assumptions about future outcomes and costs be made, and the degree of confidence placed in the conclusions of the study must reflect the credibility of these assumptions. On this basis, spending $32,678 to save one year of life may not be regarded as extraordinary as compared with the cost of other health care interventions. However, it should be noted that the additional year of life was assessed by using one-year follow-up of 10 percent of the United States patients in the GUSTO study and by using a mathematical approach to extrapolate survival, assuming that the hazards of death after one year did not depend on the thrombolytic agent received. The 95 percent confidence interval at one year yields a range of $20,000 to $70,000 for the cost-effectiveness ratio per year of life saved, but no information is available about the errors involved in mean life-expectancy calculations.

We suggest a simple approach to determine the real cost of each life saved, taking into account the incremental, undiscounted cost of $2,845 for each patient receiving t-PA and the 1.1 percent reduction in one-year mortality resulting from the use of t-PA instead of streptokinase. In this case, the administration of t-PA to 250,000 eligible patients with acute myocardial infarction in the United States would cost $712 million each year and save 2750 additional lives (1.1 percent of 250,000), yielding $260,000 for each life saved at one year of follow-up. Using the 95 percent confidence interval for one-year survival, the number of lives saved would range from 4360 to 1150, and the cost would be $164,000 to $620,000 per life saved.

Ricardo Armentano, Ph.D.
René G. Favaloro, M.D.
Instituto de Cardiologia y Cirugia Cardiovascular, 1093 Buenos Aires, Argentina

1 References

1. 1

   Lee TH. Cost effectiveness of tissue plasminogen activator. N Engl J Med 1995;332:1443-1444
   Full Text | Web of Science | Medline

Author/Editor Response

Dr. Mark replies:

To the Editor: In the concluding paragraph of our article, we attempted through extrapolation to provide a perspective on the national implications of our findings in the GUSTO trial. The figures are meant to be illustrative, since they are based on the assumption that there are 250,000 patients eligible to receive thrombolytic therapy in the United States each year. We are not aware of any authoritative statistics to support this estimate, so the resulting extrapolations should also be regarded as approximate.

Both Dr. Rose and Drs. Armentano and Favaloro identified an error in the text that has been corrected in a recent issue of the Journal. 1 Switching from streptokinase to accelerated t-PA would provide not 3.5 million additional years of life but approximately 38,500 (undiscounted) years of life during each year that the policy was in effect.

The costs to the nation of such a switch in policy would probably be between $500 million and $600 million, depending on which of several cost scenarios one selects as most representative. Drs. Armentano and Favaloro calculate this figure to be more than $700 million, using the cost difference of $2,845 from our base-case analysis. This figure is based on the published average wholesale prices of the two drugs, which, according to our survey of GUSTO hospitals, significantly overstate the costs hospitals actually pay for these agents. Thus, our feeling is that the national cost of making t-PA the standard of care for eligible patients would be lower than $700 million, but an exact figure must await better data on the number of patients who require treatment each year.

Drs. Armentano and Favaloro propose “a simple approach to determine the real cost of each life saved.” There are a number of serious problems associated with expressing cost effectiveness in terms of dollars per life saved, as they suggest, instead of in terms of dollars per year of life saved. Perhaps foremost among these is a lack of adequate benchmark values to use in assessing the attractiveness of any particular result. Without a benchmark, it is impossible to judge whether or not spending a particular sum to save a life is consistent with other social decisions about worthwhile investments.

Daniel B. Mark, M.D., M.P.H.
Duke University Medical Center, Durham, NC 27710

1 References

1. 1

   Correction to: Cost effectiveness of thrombolytic therapy with tissue plasminogen activator as compared with streptokinase for acute myocardial infarctionN Engl J Med 1995;333:267-267
   Web of Science

Citing Articles (2)

Citing Articles

1. 1

   M. de Montalembert, M. Belloy, F. Bernaudin, F. Gouraud, R. Capdeville, R. Mardini, N. Philippe, J. P. Jais, J. Bardakdjian, R. Ducrocq, M. Maier-Redelsperger, J. Elion, D. Labie, R. Girot. (1997) Three-Year Follow-Up of Hydroxyurea Treatment in Severely Ill Children with Sickle Cell Disease. Journal of Pediatric Hematology/Oncology 19:4, 313-318
   CrossRef

2. 2

   Martin H. Steinberg. (1996) Sickle Cell Disease: Present and Future Treatment. The American Journal of the Medical Sciences 312:4, 166-174
   CrossRef