The underregistration of infant mortality: a new method for its estimation.

A new method for estimating the infant mortality rate is proposed. This method is recommended where a substantial underregistration of infant deaths exist. The infant mortality rate/1-4 years mortality rate ratio varies for the different levels of the 1-4 years mortality rate but the biggest oscillations in the Latin American countries are due to underregistration of infant deaths (less than 1 year of age). In order to determine the value of this "ratio" for the different levels of the 1-4 years mortality rate countries with a tradition in good statistics were chosen. The mean of these ratios: IMR/1-4 years MR and their standard deviations were found for the different levels of the 1-4 years mortality rate. Each mean of ratios with a standard deviation minus and plus were used to calculate the estimation zone of the infant mortality rate. For calculating the latter it is only necessary to multiply the 1-4 years mortality rate registered by the mean of ratios less 1 standard deviation (lower limit) by the mean (median or average) and by the mean plus 1 standard deviation (upper limit). In the same way the infant mortality rate was estimated for each of the Latin American countries in 1975 or the last year published in the international documents. Some of the conditions which can modify the "mean multiplier of estimation" were analyzed bearing in mind the form and degree in which these conditions can influence mortality in the first 5 years of life according to incidence prevalence and lethality of various diseases. In accordance to this proposals were put forward to use the "multiplier" plus or minus 1 standard deviation in relation to prevailing conditions. Some indirect methods are being applied and will be applied to the data of Latin American censuses and surveys and they are already testing the reliability of the method proposed. If this continues to be the case health workers will own a practical and useful tool to improve health diagnosis specifically a reliable and simple method of calculating the actual infant mortality rate for planning and evaluating their activities more effectively. (authors modified)