Distribution of midwives in Mongolia: A secondary data analysis

Introduction: Midwives are at the core of strengthening the health system, especially in the crucial period around pregnancy, childbirth, and the early weeks of life. In 2016, the national-level maternal mortality ratio in Mongolia was 48.6 deaths per 100,000 live births, but this was much higher (up to 212.9 deaths/100,000) in some rural provinces of the country. The wide variation in maternal mortality between urban and rural areas of Mongolia might be related to the distribution of midwives and equity of access to maternity care.

Objectives: In the present study, we aimed to determine the distribution of midwives in each province of Mongolia and to examine inequality in the distribution of midwives nationwide.

Design: A secondary data analysis.

Methods: Data from the Centre of Health Development and the National Statistical Office of Mongolia were obtained and analysed. First, we assessed the distribution of midwives at provincial and regional levels, along with the association between these distributions and the maternal mortality ratio in 2016. We then calculated the inequality of these distributions using the Gini coefficient and examined trends for the years 2010-2016. We compared results for urban, suburban, and rural provinces. Rural areas are sparsely populated and enormous in size, so it may be difficult access to basic healthcare services. It was considered important, therefore, to assess the number of midwives per 1000 km² as well as the commonly used indicator of per 10,000 population.

Results: When the land area in each province was taken into consideration rather than only the population, wider variations between urban, suburban, and rural provinces became apparent. Provinces varied according to the number of midwives per 10,000 population by a factor of three (range 2.0-6.2/midwives); while provinces varied according to the number of midwives per 1000 km² by a factor of approximately 300 (range 0.2-61.2/midwives). The Gini coefficient for the number of midwives per 10,000 population in 2016, R = 0.201, revealed "relative" equality. This was slightly reduced from R = 0.305 in 2010 and indicated a shift toward equality. However, the Gini coefficient for the number of midwives per 1000 km² area indicated "severe" inequality of R = 0.524 in 2016. This was increased from R = 0.466 in 2010, indicating that no improvement has been seen over these years.

Conclusions: Our study suggests that two different measures of midwifery distribution should be used as indicators: number of midwives "per 10,000 population" and "per 1000 km² area". In rural areas such as parts of Mongolia, geographical features and population density are important features of the local context. To deliver healthcare services in a timely manner and within a reasonable distance for pregnant women who need care, the indicator of per 1000 km² area would be more suitable for rural and sparsely populated areas than the indicator of per 10,000 population, which is commonly used for urban and settled areas. Based on our findings, to reduce the wide gap in MMR between rural and urban areas, we recommend at least one midwife per 1000 km² area in rural regions of Mongolia.