The costs and benefits of community thrombolysis for acute myocardial infarction : a decision-analytic model.

BACKGROUND

There is evidence that the earlier a patient reaches hospital and receives thrombolysis, the better the outcome. The GREAT (Grampian Region Early Anistreplase Trial) directly addressed the issue of early thrombolysis by evaluating, in a randomised controlled trial, the efficacy of thrombolysis in the community compared with that administered in hospital.

OBJECTIVE

This paper aimed to model the cost and benefits of community compared with hospital thrombolysis from the UK NHS perspective, using efficacy data from the GREAT.

METHODS

A decision-analytic approach was used to model these two alternatives. Resource use and cost estimates were estimated for a single tertiary centre. Estimates of effectiveness in life-years were obtained from the 4-year follow-up for patients recruited to the GREAT, using declining exponential approximation of life expectancy. Costs are in pounds sterling, 2000/1 values.

RESULTS

Community thrombolysis had an average life expectancy of 12.48 years and hospital thrombolysis had an average life expectancy of 12.39 years. Costs were 361 pounds sterling for community thrombolysis and 300 pounds sterling for hospital thrombolysis. Community thrombolysis led to an additional 0.09 years of life-expectancy gained compared with hospital thrombolysis at an additional cost of 61 pounds sterling per patient. Therefore, the incremental cost per life-year gained for the community thrombolysis service over the hospital thrombolysis service was 667 pounds sterling. Sensitivity analysis showed that estimates of cost per life-year gained were most sensitive to the estimates of survival.

CONCLUSION

This model suggests that, from the UK NHS perspective, implementing community thrombolysis may lead to extra survival but at extra cost over hospital thrombolysis. Although the incremental cost per life-year is modest, judgements still have to be made, however, as to whether the extra benefits estimated are worth the additional resources required. This requires consideration of the local context in which the service may be introduced.