Percent of deliveries attended by skilled health personnel

Number of births attended by skilled health personnel in a defined time period; number of live births in the same geographic area and reference period.

The number of live births is a proxy for the numbers of women who need delivery care. Evaluators should count all births but usually only use live births in calculating this indicator, because of the difficulty in obtaining information about non-live births (Graham and Filippi, 1994).

Where data on the numbers of live births are unavailable, evaluators can calculate total estimated live births using census data for the total population and crude birth rates in a specified area. Total expected births = population x crude birth rate.

If targeting and/or linking to inequity, classify delivery sites by location (poor/not poor) and disaggregate by area served.

Routine health service data; population-based surveys.  Routine health service data typically lack information on pregnancies or births that take place outside the public health sector, for example in homes or private facilities.

The main purpose of an indicator of the skilled attendant at delivery is to provide information on women's use of delivery care services.  It helps program management at district, national and international levels by indicating whether safe motherhood programs are on target with making professional assistance at delivery available and used.  In addition, the proportion of births attended by skilled health personnel is a measure of the health system's functioning and potential to provided adequate coverage for deliveries. 

Many argue that increasing the proportion of deliveries with a skilled attendant is the single most critical intervention for reducing maternal mortality. Moreover, the proportion of births with a skilled attendant is a benchmark indicator for monitoring progress towards the Millenium Development Goals (WHO, 1999b).

The evidence that delivery with a skilled attendant reduces maternal mortality comes from a number of clinical, historical, and epidemiological sources that indicate an association but not a causal link. In general, births with a skilled attendant are associated with lower rates of maternal mortality. However, confounding factors, such as the strong correlation between skilled attendant and institutional delivery, make assessing the impact of skilled attendant alone difficult to determine.

Evaluators can disaggregate skilled attendant at delivery by place to further document the degree of care received at the time of delivery. This measure of care or "skilled attendance" will vary by setting and attendant. A skilled attendant conducting a delivery in hospital, for example, provides a higher level of "skilled attendance" than does a skilled attendant conducting a delivery at home.

The percentage of births with a skilled attendant is one of four mutually supportive indicators in the minimal list measuring maternal health services coverage. The other three indicators are:

• Percent of women attended at least once during pregnancy for reasons related to the pregnancy;
• Availability of basic essential obstetric care; and
• Availability of comprehensive essential obstetric care.

In combination, these indicators measure progress towards the goals of providing antenatal care, trained attendants during childbirth, and access to essential obstetric care for all pregnant women.

Annual monitoring is only feasible when the data are derived from routine data sources. For international comparisons, periods of three to five years are probably sufficient. Frequent surveys are generally undesirable because the survey periods may overlap, and sampling error makes it difficult to assess whether small changes are real or due to chance variation.

Evaluators should not infer that similar rates of skilled attendant deliveries between countries reflect similar levels of care; major differences are likely to exist between countries in how providers are trained, in what providers are allowed to practice and do practice, and in what resources, equipment, and supplies are at their disposal.

Differences in what definitions are used and in how skilled attendants are reported may also account for discrepancies between countries. Most surveys such as the DHS rely on women's self-report but how women interpret the question "who assisted with the delivery?" and whether they accurately identify the health staff attending is unknown.

This indicator uses a birth-based analysis (similar to the ANC indicators), and the sample will over-represent women with multiple births in the survey period. Women with more than one birth are also more likely to have other risk factors, such as high parity and lower rates of health services use. Delivery coverage may therefore be underestimated, although this underestimate is likely to be small.

Since the denominator for this calculation includes only women with live births and excludes women with fetal deaths and stillbirths, the only valid association will be with neonatal mortality and not with perinatal mortality. (See the Newborn Health section.)

This indicator does not take into account the type and quality of care from a skilled health provider.

(Excerpted from: Becker L, Wolf J, Levine R (2006) Measuring commitment to health. Center for Global Development.)

Little or no conclusive evidence exists on differences between the maternal mortality and morbidity
of the rich and that of the poor. There is, however, clear evidence on the difference in the use of obstetric care based on socio-economic class. In a study of 45 developing countries and transition
economies, World Bank researchers found that in every country both the wealthiest quintile and the
population as a whole was significantly more likely than the poorest quintile to have medically-trained personnel present at birth (Gwatkin, et al. 2006).  In many countries the direct correlation between wealth and use of obstetric care is consistent across all five wealth quintiles (Kunst & Houweling, 2001).

It is far more cost-effective to increase birth attendance in areas with low current rates than to  do so in areas with relatively high rates (Graham, et al. 2001).  As a result, cost-conscious policies likely will have  some measure of pro-poor focus built into them. Given this fact and the much lower rates of skilled birth attendance among the poor, it seems likely that any increased focus on increasing skilled birth attendance would have some impact on reducing the gap between the rich and poor in terms of obstetric care and, by extension, maternal mortality.

Although the academic literature does not appear to have addressed this issue, efforts to increase
the prevalence of skilled attendance at birth should also contribute to poverty reduction because of the significant burden that maternal mortality and morbidity can impose on families in developing countries. Women of child-bearing age contribute to the household financially through their labor productivity and by caring for the entire family. The loss of this resource due to death or morbidity- related disability contributes to household poverty and reduces child survival rates. Conversely, actions that prevent maternal mortality and morbidity should decrease financial risks for poor households.

Becker L, Wolf J, Levine R. 2006. Measuring commitment to health. Center for Global Development.

Gwatkin D, et al. 2004. “Socio-Economic Differences in Health, Nutrition, and Population - 45 Countries”. The World Bank. March 20 2006.   https://siteresources.worldbank.org/INTPAH/Resources/IndicatorsOverview.pdf

Graham W, Bell J, and Bullough C. 2001. “Can Skilled Attendance at Delivery Reduce Maternal Mortality in Developing Countries?” Safe Motherhood Strategies: A Review of the Evidence. Eds. Vincent De Brouwere and Wim Van Lerberghe. Vol. 17. Studies in Health Services Organisation & Policy. Antwerp: ITGPress. 97-129.

Kunst A & Houweling T. 2001. “A Global Picture of Poor-Rich Differences in the Utilisation of Delivery Care.” Safe Motherhood Strategies: A Review of the Evidence. Eds. Vincent De Brouwere and Wim Van Lerberghe. Vol. 17. Studies in Health Services Organisation & Policy. Antwerp: ITGPress. 293-311.

Reproductive health indicators: guidelines for their generation, interpretation and analysis for global monitoring.  WHO, 2006.