An Assessment Of Correlation Between Mortality And Distance From Patient’s Residence To Interventional Center In Percutaneous Coronary Intervention In Brazil

OBJECTIVES: To assess the relationship between myocardial infarction in-hospital mortality after percutaneous coronary intervention (PCI) and transportation time from patient's residence to interventional center in the Brazilian public healthcare system (SUS). METHODS: Brazilian public healthcare claim databases (DataSUS) were used to assess the relationship between myocardial infarction in-hospital mortality and the transportation time from patients' residence zip code to interventional center where PCI was performed. DataSUS data between June 2014 and June 2016 was extracted. Transportation time was obtained through Google Maps Distance Matrix API. Patients' sample was defined selecting only patients with records of PCI through balloon inflation or stent placement. In order to reduce sampling noise, patients with inconsistent outcome data, zip code or interventional center identification were excluded from sample. Additionally, any transportation time higher than three hours were excluded in order to reduce risk of bias resulting from potential travel. Transportation times were stratified in two groups: less than one hour and more than one hour. Statistical significance was tested through Fisher's exact test. Brazilian states were clustered according to their annual per capita gross domestic profit (GDP): one group higher than the country's median (group A) and another one below the country's median (group B). RESULTS: Out of 56.883 hospitalizations, 54.141 were included in the analysis. When analyzing the country as a whole, no difference in in-hospital mortality was observed between patients living closer or farther to the hospital that provided the PCI. However, patients living closer to the hospital in group B states had a lower mortality compared to those living farther (OR=0.83. p=0.03), what was not observed in group A states. CONCLUSIONS: Distance between patients' residence and hospital seems to have higher effect over mortality in lower GDP per capita states. Further investigation on reasons for differences between group A and B results is needed.