A statistical model for defining priority geographic areas for the control of neonatal tetanus

We applied factor analysis to indicators of health and living conditions from Brazilian states in order to define geographic areas at potential risk for neonatal tetanus. Two factors, namely 'health profile in the rural area' and 'proportional mortality on neonatal tetanus' were selected and plotted against each other. A cluster composed of States from the Northeast plus Para and Amapa was found to include most neonatal tetanus risk areas and a low case-reporting rate. Another cluster included States from the Southeast and South and displayed a neonatal tetanus reporting rate that was compatible with that for other indicators. Espirito Santo, however, was found to be a silent productive area. The Federal District appeared alone, showing the best health conditions. Finally, the States of the Middle West and Roraima constituted the last cluster, characterized by intermediate health status and high neonatal tetanus case-reporting rate. Our results were consistent with the overall Brazilian health profile, distinguishing the North and Northeast from the South and Southeast, with the Middle West in an intermediate position.