Indicator 8: Met need for emergency obstetric care

Direct obstetric complications cause approximately 75% of maternal deaths worldwide.¹ Providing emergency obstetric care services, or treatment for women with the major direct obstetric complications, is a key action that must be taken to prevent these maternal deaths.² The indicator, met need for emergency obstetric care, was developed to help program planners and policymakers track to what extent women with direct obstetric complications used emergency obstetric care services.

Holmer et al (2015)³ extracted met need for emergency obstetric care from 51 countries and estimated that the global met need for emergency obstetric care was 45%. They found that met need for emergency obstetric care varied by income level of countries: low-income countries had the lowest met need (21%), middle-income countries had somewhat higher met need (28-51%), and high-income countries had the highest met need (99%). The authors also found an inverse relationship between met need and the maternal mortality ratio.

Met need for emergency obstetric care is a systems level indicator that aggregates data across all facilities (both government and private sector facilities), and is best assessed on a 3-5 year basis at the start of a strategic planning or programme planning cycle. Countries that collect data for this indicator in all facilities as part of their HMIS (see below) may choose to calculate this indicator more frequently (e.g., annually).

The numerator for this indicator is the number of women with direct obstetric complications recorded in facility registers and who were not referred out, which serves as an estimate of the number of women with direct obstetric complications who received definitive treatment in facilities. Definitive treatment is defined as clinical care provided to a woman with direct obstetric complications that results in complication resolution and discharge from the facility or her death. It is assumed that if a woman’s complication is written down in facility registers and she was not referred out, that she received definitive treatment. Women with complications who are referred out to another facility are excluded from this indicator, even if they received some clinical care before referral.⁶

Data on obstetric complications can be found in registers in the maternity unit, operating theater, female or gynecological unit, and the emergency, inpatient and outpatient departments. Patient records can also be used for further investigation of complications. Some countries record direct obstetric complications in their HMIS and if the data are determined to be good quality (see below under ‘Supplemental studies’), they can be used in place of reviewing individual facility registers.⁷ The direct obstetric complications included in this indicator are: hemorrhage (antepartum and postpartum), severe pre-eclampsia/eclampsia, ruptured uterus, prolonged/obstructed labor, ectopic pregnancy, sepsis/maternal peripartum infection and complications of abortion. Operational definitions of major direct obstetric complications and ICD-11 codes can be used to standardize data collection. The denominator is an estimate of the expected number of women with direct obstetric complications. It is calculated by multiplying the expected number of live births⁸ by 15%. (Estimating need for emergency obstetric care provides more information about the rationale for the 15% estimate.)⁹

This indicator is expressed as a percentage.

The benchmark for this indicator is 100% of women with direct obstetric complications receive definitive treatment in facilities.

Methodological issues around recording complications of abortion can impact the magnitude of met need. Abortion complications are often recorded in a highly variable manner, including inaccuracies regarding severity, e.g. whether the abortion was only incomplete (which could lead to a complicated abortion) or truly complicated (with hemorrhage or sepsis). It is therefore recommended that met need for emergency obstetric care be calculated and reported two ways for comparison: 1) with all incomplete abortions, and 2) with only the most severe complications of abortions. For example, an assessment conducted in 2016 in Ethiopia showed that when non-severe post abortion complications were included, met need increased from 18% to 29% in all facilities. By showing how the two estimates vary we gain a fuller picture of women’s use of services and the demands being made on the health system.

This indicator is used by planners to understand whether emergency obstetric care services are being delivered to and used by women with obstetric complications. If met need for emergency obstetric care is low (i.e., less than 100%), it may indicate that: 1) women with obstetric complications are not accessing EmONC services; and/or 2) health facilities are not providing definitive EmONC services; and/or 3) providers are not identifying obstetric complications; and/or 4) record-keeping is poor (e.g., under-reporting of complications).

Calculating met need at the national level is the most straightforward way to interpret the indicator. If women with complications at all public and private sector facilities are well-recorded in registers, this should provide a strong national estimate of met need for emergency obstetric care. Interpreting met need is more challenging when not all facilities in an area are included in its calculation, such as only counting women with complications definitively treated in public facilities, especially in an area where many private facilities also deliver EmONC.

The numerator for this indicator can be disaggregated by type of facility, facility EmONC classification, or managing authority (e.g., public, private for-profit). This provides an estimate of the proportion of met need for emergency obstetric care that is attributable to facilities with a given characteristic (e.g., the contribution of the private or public sector to met need for emergency obstetric care). The results of such analysis can show which facilities are providing most of the definitive care, which facilities are providing more care than expected, and which facilities are underperforming. For example, in Kigoma, Tanzania, an assessment of all health facilities in the region showed that the overall met need for emergency obstetric care was 61%.¹⁰ But when the numerator is disaggregated, we see the contribution that each type of facility made to the overall met need (Table 1).

Met need for emergency obstetric care is useful for tracking progress over time. In places with focused EmONC implementation support, data quality often improves. So increases in met need may reflect improved recording of complications and not just higher use of emergency obstetric care services. Special studies can be conducted to better understand what is driving the increase (see below).

Met need for emergency obstetric care does not tell us anything about the quality of clinical management, but other indicators in the EmONC Framework can provide information about quality of care (see, Exploring quality of care with the EmONC Indicators). For example, it may be useful to compare met need for emergency obstetric care for facilities that perform all of the EmONC signal functions and are well-equipped and staffed with those that are not.

Assess the quality of record keeping at facilities

Several types of studies could be used to explore the quality of record keeping at a facility:¹²

• Examine how records are kept. Does someone enter complications into the register 24 hours per day, or does the senior nurse document them only once a day from verbal reports by other staff? The latter practice could lead to serious underreporting.
• Compare the complications recorded in the maternity register with patient charts, operating theater registers or emergency admissions registers. What proportion of serious complications is not reported in the register that is usually used for calculating met need? Which complications appear to be most underreported? How do your findings change when you correct for this underreporting? How often does a diagnosis of complication change between the admissions register and the operating theater register?
• Examine how abortion complications are recorded by discussing the records and case notes with staff. Are minor complications, or even all incomplete abortions, counted as ‘complications’?
• Knowing more about how well and how completely registers are kept up can identify problems. Investigate whether staff training or supervision of record keeping reduces underreporting over time, and then disseminate your results.

In countries that report direct obstetric complications to the HMIS, periodic assessments can be conducted in a sample of facilities (at different levels of the health system) to determine whether the number of complications recorded in registers, patient records and the HMIS are consistent. If underreporting or overreporting in the HMIS is identified, data quality improvement interventions can be implemented.

Calculate the direct obstetric complications referral rate

Some obstetric complications cannot be definitively managed at facilities where they are first received. Women with these obstetric complications need to be stabilized and then referred to other facilities providing the required level of EmONC. The direct obstetric complications referral rate is the proportion of women with direct obstetric complications who are referred out to another health facility in order to receive definitive treatment. The direct obstetric complications included in this indicator are the same as those included in met need for emergency obstetric care (see ‘Data collection and calculation’ section above).

To calculate the referral rate, information is needed on the number of women who present at facilities with direct obstetric complications (the denominator) and the number of women who are referred out as a result of those complications (the numerator). These data are ideally found in the HMIS and originate in maternity ward, referral, outpatient and inpatient women’s / gynecological unit registers. All of these registers should be checked, especially to ensure the capture of complications of abortion and ectopic pregnancies. This indicator is expressed as a percentage. There is currently no benchmark for this indicator, though it is useful for tracking change in referrals over time.

The direct obstetric complications referral rate indicator can help program planners and policymakers track to what extent referral is used to deliver definitive EmONC services to women with obstetric complications. This indicator should be disaggregated by sub-national area and then analyzed by type of facility and facility EmONC classification to highlight facilities’ ability to provide the level of care that they are expected to provide. For example, Comprehensive EmONC facilities should provide definitive care for almost all women with obstetric complications, except for the few who require Intensive EmONC. Thus, Comprehensive EmONC facilities with very high referral rates are not functioning as they should. If data can be disaggregated to the facility level, district teams could target investment to facilities with referral rates that seem too high or too low, based on the level of facility and local experience to: 1) ensure they can manage the complications they are expected to; 2) ensure referral practices adhere to standards and enable best outcomes; and 3) establish/strengthen communication between the sending and receiving facilities.

It may be helpful to analyze the direct obstetric complications referral rate in conjunction with EmONC indicator 5, which assesses facility emergency referral readiness. Frequent referral with inadequate referral readiness poses a risk to patient safety. It may also be helpful to analyze this indicator with EmONC indicator 3, which assesses the availability of commodities, and EmONC indicator 4, which assesses the adequacy of human resources since some referrals may be due to lack of facility readiness.