Subscribe to Research Article | IHME

Publication date: 
April 13, 2010


Researchers have found a way to measure child mortality more accurately and less expensively, through the development of new analytical methods, enabling policymakers to respond more quickly to pressing public health concerns. The study, Measuring under-five mortality: validation of new low-cost methods, shows how these new methods can be used to evaluate mortality trends in specific regions, revealing health disparities. The work was done in collaboration with scientists at the University of Queensland.

Research findings

Researchers compared their new methods for calculating mortality for children under 5 with more traditional mortality rate calculations, and found that the new methods outperformed the standard method by an average of 43.7%. The new methods worked even better for more recent time periods. The researchers found that, in the five years just prior to the survey, the new methods showed a 53.3% improvement, on average, over the standard method.

Analytical approach

In countries lacking death registration data, researchers rely on two primary methods for counting the number of children who have died:

  1. A summary birth history, where survey takers ask mothers how many children they have had and how many children have died.
  2. A complete birth history, where interviewers ask more detailed questions, including when the births and deaths occurred. Complete birth histories generally result in stronger mortality estimates, but they are far more costly to conduct because of the amount of time and staff required.

Researchers analyzed data from 166 health surveys that used summary birth histories and complete birth histories to calculate mortality rates.The researchers then created five new methods for calculating under-5 mortality from summary birth histories and checked those methods against the data produced by analyses based on complete birth histories.

Research objective

Accurate and comparable data about child mortality are essential for decision-making about resource allocation and health planning. IHME is working to develop better analytical methods that will lead to a better understanding of health disparities and more effective prevention programs and intervention delivery.

Recommendations for future work

The researchers propose that the two questions that comprise a summary birth history can easily be added to existing survey programs and national censuses to allow further analysis and better monitoring of progress in global child mortality rates.


These data show the models used to estimate under-5 mortality from summary birth history data, as well as a tutorial so others can apply the methods to their own datasets.

The zip file contains a tutorial that walks a user step by step through the process of applying the model coefficients to microdata. Model coefficients and a sample set of microdata are included in two formats: comma-separated values and Stata-formatted. This will allow you to be familiar with the structure of each dataset and how to apply the methods to obtain estimates of under-5 mortality.
IHME is also currently developing an automated application to apply the new summary birth history methods that will include uncertainty estimation.

Rajaratnam JK, Tran LN, Lopez AD, Murray CJL. Measuring under-five mortality: validation of new low-cost methods. PLoS Medicine. 2010 April 13; 7(4):e1000253.