CMRJack is a program that computes various measures of child mortality from birth histories. It also computes standard errors for all estimates.

CMRJack can estimate child mortality from full birth histories, full pregnancy histories, or summary birth histories. It also calculates stillbirth rates from pregnancy histories.

The program accepts microdata of the type produced by the Demographic and Health Surveys (DHS), the World Fertility Surveys (WFS), the Multiple Indicator Cluster Surveys (MICS), and the Reproductive and Health Surveys (RHS). Other surveys that use slightly different data formats from the DHS and MICS ones may be easily adapted.

CMRJack is a standalone Windows program. It does not require other software, but may be run from SPSS, R, or Python.

Since 2012, CMRJack has been used to generate the baseline estimates that enters into the child mortality estimation of the UN Interagency Group for Child Mortality Estimation.

CMRJack is free, and can be downloaded from this website.

Full birth histories

From full birth histories (“direct estimates”), CMRJack estimates:

 

  • Neonatal mortality (as well as early and late neonatal)
  • Post-neonatal mortality
  • Infant mortality (IMR, 1q0)
  • Mortality by single year for age 1,2,3,4 (1q1, 1q2, 1q3, 1q4)
  • Under-five mortality (5q0, U5MR)
  • Child mortality (4q1)
  • Sex ratios of neonatal, infant and U5MR
  • Ratios of Neonatal to 1q0, 5q0
  • The ratio of Early neonatal to Neonatal mortality
  • The ratio of IMR to U5MR
  • Bivariate inequity measures
  • Time periods of estimation can be 1,2,3,4,5 or 10 years or can be calculated automatically based on the standard errors of the estimates.
  • Time periods may either be based in calendar years, or in time before the survey.

Summary birth historie

From summary birth histories (“indirect estimates”) CMRJack estimates:

  • Under-five mortality rate (U5MR, 5q0)
  • Infant mortality rate (IMR, 1q0)
  • Estimation may be based on classification by the age of women or time since first birth
  • Sex ratios
  • Bivariate inequity measures
  • Various versions of model life tables may be used

Common features

All estimates may be broken down by a factor variable and may be calculated on the real, log, or logit scale.

Bivariate inequity measures include the range, the range ratio, ratio of highest to lowest category, relative risk, and the relative risk percentage.

In the case of bivariate inequity measures based on a wealth index, CMRJack can reestimate the ntiles so that each ntile contains the same number of births.

CMRJack reads SPSS, comma-separated, and Excel microdata files.