Costo efectividad de las pruebas para la detección de enfermedad mínima residual en niños con leucemia linfoide aguda al final de la inducción

Objective To assess the cost-effectiveness of polymerase chain reaction (PCR) compared to flow cytometry to detect minimal residual disease at the end of induction in acute lymphoid leukemia (ALL) patients under 18 years old. Methods A decision tree was built in TreeAge®, with outcome measured in correctly identified cases (true positives plus true negatives). The data of the operating characteristics of the tests were taken from the literature. As PCR operating characteristics were found only for the T cell subgroup, 2 models were built that only differ in the extrapolation method from the subgroup to the population. The perspective is that of the health system. All monetary amounts are expressed in 2010 Colombian pesos. Univariate and probabilistic sensitivity analyses were performed. Results In the first model, the cost of an additional correctly identified case with PCR compared to flow cytometry was COP $ 519.642, ranging from COP$475.071 to $17.148.505 when the cost of PCR varies between COP$300.000 y $3.000.000. In the second model, the cost of an additional correctly identified case with PCR is COP $24.442.226 and varies between COP $7 million and $245 million. Conclusion For the parameters of the first model, if the average cost of treating a patient misdiagnosed by EMR exceeded COP $ 17,148,505, it would cost effective to use PCR in pediatric patients with ALL in the range studied, if the cost of treating a misdiagnosed patient at this stage were higher than $17.148.505; this figure is much higher for the second model. The variability of the incremental cost effectiveness ratio between the two models demands caution when interpreting the results, as they are very sensitive to the extrapolation method from the T-cells subgroup to the ALL population. From the point of view of cost effectiveness, it might be more appropriate to consider the possibility of using PCR only for the T-cell subgroup, for which the results from the first model can be applied more directly.