PubMed1996
Comparison of infant mortality rates (IMRs) among the world's countries requires assessment of completeness and accuracy of data. The United Nations Statistical Office classifies as "C", complete, meaning at least 90% of events are actually recorded, 1994 data supplied by 80 governments, comprising one fourth of the world's population, ie, 1,450,000,000 people, and as incomplete the other three fourths, 4,180,000,000. All the "C" countries officially accept the World Health Organization definition of a live birth (any product of gestation showing any sign of life), but it has been argued that some countries routinely report as stillbirths infants counted as live births in the United States (US), thus understating their IMRs. In 1994, 22 countries had IMRs varying from 4.2 for Japan to 8.0 for the US, a remarkable achievement in the light of IMRs of 124.0 and 60.0 for these two countries in 1930. Compensating for possible underreporting of live births by excluding all deaths in the first hour of life would reduce the US IMR to about 7, still higher than 17 other countries. Between 1930 and 1994 the IMR in the US declined more slowly than several other countries, particularly during the time period 1951 through 1965, when the US rate declined by 16% and the Japanese rate, for instance, declined by 68%. Between 1983 and 1994, decline in Puerto Rico was slower than in Chile, Cuba, and the US. IMRs in all the "C" countries are lower than the US rate was in 1930. IMRs in most of the world, estimated from surveys and special studies, vary from 27 to 190. Correlation studies suggest that a high rate of teenage pregnancies has relatively little effect on IMRs but that high total fertility rates are accompanied by high infant mortality.