Volume 51 | Number 2 | April 2016

Abstract List

Rie Sakai M.D., M.P.H., Ph.D., Günther Fink Ph.D., Hiraku Kumamaru M.D., ScD., Ichiro Kawachi M.D., Ph.D.


To investigate the effect of pediatrician supply on under‐5 mortality over the period 2000–2010.

Data Sources

Multiple publicly available data sources were used.

Study Design

Japan's 366 “Secondary Tier of Medical Care Units” () were used as study units. To evaluate the association between under‐5 mortality and pediatrician supply, we explored time and area fixed‐effects Poisson regression model. The following factors were introduced into the models as time‐varying controls: (1) number of physicians other than pediatricians per total population except for under‐5‐year‐old population, and (2) income per total population by year and . Extensive sensitivity analyses were conducted to assess robustness of results.

Principal Findings

Pediatrician density was inversely associated with under‐5 mortality. We estimated that a unit increase in pediatrician density was associated with a 7 percent (95 percent : 2–12 percent) reduction in the child mortality rate after adjustment for all other variables. The results were consistent and robust across all specifications tested.


The results suggest that increasing human health resources can have positive effects on child health, even in settings where child mortality of less than 5 per 1,000 has been achieved.