A cohort analysis of maternal mortality in the United States, 1917-1969
Age-specific period and cohort rates of maternal mortality are examined for the calendar years of 1917-1969 and the birth cohorts of 1898-1927. The information needed to calculate these rates is taken from annual U. S. Vital Statistics volumes.
Four questions are addressed: (1) What are the cohort patterns of age-specific maternal mortality? (2) How have age-differentials in the risk of maternal mortality changed over time? (3) How much do rates standardized for age and parity distributions of births differ from crude maternal mortality rates? (4) Can cohort and period effects on maternal mortality be measured with a nonlinear regression model similar to the one used in Greenberg, B. G., J. J. Wright, C. G. Sheps. "A Technique for Analyzing Some Factors Affecting the Incidence of Syphilis." Journal of the American Statistical Association 45: 373-399 (1950).
The basic findings with respect to the four questions listed above are: (1) cohort patterns of age-specific maternal mortality are mainly a reflection of the rapid decline in period rates of maternal mortality, (2) age differentials in the risk of maternal mortality have increased as the average rates declined, (3) rates standardized for age and parity differ from crude rates by an average of 2 to 4 percent, (4) the mathematical model presented by Greenberg et al. (1950) can be used to measure period and cohort effects on variables, such as maternal mortality, that have a clearly defined relationship to age. Three restrictions on the model's use must be observed however. First, each of the effects (age, period and cohort) must be large enough to be estimated. Second, since the presence of incomplete cohort information produces a confounding of age, period and cohort effects, the model can only be used when data on the complete experience of several cohorts are available. Third, since the model requires a priori knowledge of the functional form of the relationship between the dependent variable and one of the independent variables, its application is limited to topics for which it is possible to specify the model before beginning the analyses.
Thesis76b.B479.pdf
4.8 MB
Adobe PDF
626948dbec7fa74457d43e4882f80921