Key points are not available for this paper at this time.
A comparison was made between maternal mortality rates published by the World Bank in its World Development Report (WDR) 1993 with 1988 data and the UNs Human Development Report (HDR) 1993. Both data sources claimed the data was based on World Health Organization (WHO) data but the two sets of figures were different. When rate differences of 50 points were taken to be exactly the same HDR gave higher values for 26 countries lower values for 12 countries and the same values for 17 countries compared to WDR. Even the averages were different. HDR gave an average maternal mortality rate of 393/100000 births for 55 countries and WDR gave a figure of 346/100000. When the data were weighted for the estimated number of married women in each country the figures were 260 for HDR and 231 for WDR. The correlation coefficient between the two values was 0.70 or an R2 = 0.50. There was a less significant relationship between the two values for 23 countries with high maternal mortality rates (>450 in HDR); R2 = 0.18. The data in each volume were reported exactly the same or similarly as in the case of time period. WDR reported that data ranged from 1983 to 1991 by country and HDR stated that data were of uncertain unreliability. A comparison of the WHOs Maternal Mortality: A Global Factbook 1991 and WHOs definition according to the 9th and 10th Revisions of the International Classification of Diseases (ICD) indicated some variance with the WDR definition. The WHO definition was given in full with the notation that the ICD 10th revision included pregnancy-related death within 42 days of the end of pregnancy and regardless of the cause of death. WHO did not produce its own figures but based its figures on prior studies and statistics. Reliability of maternal mortality rates is dependent on reliable maternal mortality and birth data. Reporting anomalies are footnoted where extremely divergent from actual mortality. A comparison of WHO published figures and HDR and WDR figures indicated that for Benin WHO and WDR agreed on a rate of 161 while the HDR figure of 800 was close to an 809 figure based on a hospital study which not representative. There were differences in the two data sets in the countries selected for inclusion and the reasons were not apparent. The suggestions were to liberally use technical notes to coordinate and agree on international publication of figures and to provide the best estimate from a wide range of estimates and qualify with a footnote.
W. Parker Mauldin (Wed,) studied this question.