[Using different indicators of preventable mortality as an approach to measuring health inequalities in Chilean...

To analyze differences in avoidable mortality among communes in Chile, using different indicators as an operational approach to estimating health inequalities.

Small area variation analysis in a sample of 117 of all 335 Chilean communes that existed in 1992. By using secondary data, we developed and compared some avoidable-mortality indicators, such as potential years of life lost (PYLL), avoidable mortality (AM) (based on background and criteria drawn from the literature), health care avoidable mortality (HCAMR), and life expectancy. A socioeconomic development index (SEDI) was also developed. The scope of the variation was estimated through the weighted variation coefficient, the Gini coefficient, the ratio between the values for the quintiles at both extremes of the SEDI distribution, and the ratio of the lowest SEDI quintile to the group of municipalities having a SEDI greater than 0.90 (optimal empirical reference value). The socioeconomic pattern of variations was examined through concentration curves and by comparing communal quintiles based on their SEDI.

The various avoidable-mortality indicators used showed an inverse and statistically significant correlation with socioeconomic development, as well as with the profile of the various SEDI quintiles and with the majority of specific causes of avoidable mortality. The distribution profile of AM indicators among SEDI communal quintiles reflects the same tendency, along with most of the mortality from specific avoidable causes. The use of three reference values (the mean, the quintile with the greatest SEDI, and the optimal empirical reference value) makes it possible to measure gaps that could be avoided. The ratio of the lowest SEDI quintile to the empirical optimal reference value was 2.1 for AM, 2.0 for PYLL, 1.7 for infant mortality, and 1.5 for HCAMR.

These results, which are consistent with those found in previous published sources, estimate the magnitude and pattern of variations among communes. The results also provide information, based on data for 1992, with which to start monitoring health inequalities among small geographic areas, which were communes in this particular case. Although interventions for promoting equity tend to focus exclusively on communes having lower socioeconomic development and higher rates of avoidable mortality, reducing the latter implies a two-pronged approach: prioritizing interventions targeting underprivileged communes so as to foster equity, while attempting to cover the majority of communes in an effort to prevent avoidable mortality.