INTRODUCTION

The third-party payer whom provide health services around the world need mathematical and methodological tools to estimate the future expenditures related health services in different populations. We use breast cancer data to generate a predictive mathematical model as help to estimate the health services’ cost in the context of a developing country as Mexico.

OBJECTIVES

: To evaluate the concordance between observed and calculates expenditures in the treatment of a sample of breast cancer patients attended in a Mexican public health institution.

METHODS

: Were reviewed the clinical records from 407 breast cancer women to calculate through micro-costing, the total expenditure in health during their treatment. After that was generated the Montecarlo mathematical predictive model to try to estimate the health services expenditure in a 100 women hypothetical sample, simulating the clinical characteristics, as primary tumor size, clinical stage, histologic features, lymphatic ganglia, surgery, radiotherapy and chemotherapy kind, metastases, metastasis anatomic localization, and recurrence treatment. Chi squared and student T test for independent samples were used to estimate p value between studied patients and hypothetical population.

RESULTS

: Were not found statistical differences between clinical stage ((p=0.865), tumor receptors (p=0.893), and kind of metastases between patients sample and hypothetical sample (p=0.699). The mean cost for the breast cancer women sample was $251,454.43 MXN, and for hypothetical sample $329,503.85 MXN (p<0.05), however we did not find differences in surgical procedures cost (p=0.441)

CONCLUSIONS

: The Montecarlo predictive model could be a useful tool to estimate the health services expenditure and to calculate a budget close to real cost