Research

Health worker densities and immunization coverage in Turkey: a panel data analysis

Andrew D Mitchell¹ , Thomas J Bossert¹ , Winnie Yip² and Salih Mollahaliloglu^1,3

¹  Harvard School of Public Health, Boston, Massachusetts, USA

²  University of Oxford, Oxford, United Kingdom of Great Britain and Northern Ireland

³  School of Public Health, Ministry of Health, Ankara, Turkey

author email corresponding author email

Human Resources for Health 2008, 6:29doi:10.1186/1478-4491-6-29

Published:	22 December 2008

Abstract

Background

Increased immunization coverage is an important step towards fulfilling the Millennium Development Goal of reducing childhood mortality. Recent cross-sectional and cross-national research has indicated that physician, nurse and midwife densities may positively influence immunization coverage. However, little is known about relationships between densities of human resources for health (HRH) and vaccination coverage within developing countries and over time. The present study examines HRH densities and coverage of the Expanded Programme on Immunization (EPI) in Turkey during the period 2000 to 2006.

Methods

The study is based on provincial-level data on HRH densities, vaccination coverage and provincial socioeconomic and demographic characteristics published by the Turkish government. Panel data regression methodologies (random and fixed effects models) are used to analyse the data.

Results

Three main findings emerge: (1) combined physician, nurse/midwife and health officer density is significantly associated with vaccination rates – independent of provincial female illiteracy, GDP per capita and land area – although the association was initially positive and turned negative over time; (2) HRH-vaccination rate relationships differ by cadre of health worker, with physician and health officers exhibiting significant relationships that mirror those for aggregate density, while nurse/midwife densities are not consistently significant; (3) HRH densities bear stronger relationships with vaccination coverage among more rural provinces, compared to those with higher population densities.

Conclusion

We find evidence of relationships between HRH densities and vaccination rates even at Turkey's relatively elevated levels of each. At the same time, variations in results between different empirical models suggest that this relationship is complex, affected by other factors that occurred during the study period, and warrants further investigation to verify our findings. We hypothesize that the introduction of certain health-sector policies governing terms of HRH employment affected incentives to provide vaccinations and therefore relationships between HRH densities and vaccination rates. National-level changes experienced during the study period – such as a severe financial crisis – may also have affected and/or been associated with the HRH-vaccination rate link. While our findings therefore suggest that the size of a health workforce may be associated with service provision at a relatively elevated level of development, they also indicate that focusing on per capita levels of HRH may be of limited value in understanding performance in service provision. In both Turkey and elsewhere, further investigation is needed to corroborate our results as well as gain deeper understanding into relationships between health worker densities and service provision.