Background
Despite failing to reach the millennium development goals (MDGs), much progress has been made in improving the health of mothers and children globally [
1]. Maternal mortality ratio (MMR) fell by 44% and under-five mortality rate declined by 53% between 1990 and 2015. Ethiopia registered a more remarkable progress, reducing MMR by 71.8% and meeting the MDG target for reducing under-five mortality by two-thirds [
2,
3]. However, the levels of maternal mortality ratio (353 per 100, 000 live births), neonatal mortality rate (28 per 1000 live births) and stillbirth rate (29.7 per 1000 births) remain high, making Ethiopia one of the largest contributors to the global burden of maternal deaths, newborn deaths, and stillbirths, ranking fourth, sixth, and fifth, respectively [
4,
5].
In 2015, the United Nations General Assembly adopted the more ambitious sustainable development goals (SDGs), which include targets for ending preventable neonatal deaths and drastically reducing global MMR to less than 70 per 100,000 live births [
1]. In line with this global aspiration, the Government of Ethiopia committed to markedly reduce MMR to 199 per 100,000 live births and neonatal mortality (NMR) rate to 10 per 1000 live births by 2020 [
6].
Meeting these ambitious global and national goals for maternal and newborn health requires improving the quality of maternal and newborn care. In view of the fact that intrapartum and postpartum periods are the time of greatest risk for the mother, fetus and newborn [
7], assuring the quality of care provision during labor, childbirth and immediate postpartum period is of utmost importance. Encouraged by a positive trend in coverage of healthcare services during the MDG period, the Government of Ethiopia has also put unprecedented emphasis on improving quality of care in its current health sector plan [
6]. In addition to improving health systems and health outcomes [
8], improving quality of care can increase demand for maternal health care [
9], which is still a challenge in Ethiopia [
10].
Improving quality of care requires measuring it accurately and addressing identified gaps [
11]. There is a clear need for more and better research evidence on quality of intrapartum care and quality of maternal health workforce especially from low and middle income countries [
12‐
15]. Most previous studies on quality of care or workforce from Ethiopia and other resource-constrained settings are based on self-report, written test, or simulation with anatomical models [
16‐
23]. In addition, most studies assessed emergency obstetric and newborn care [EmONC] capability but not quality of routine childbirth care [
12,
18,
22,
24‐
28].
The literature on healthcare quality measurement and improvement describe multiple dimensions of healthcare quality. The Donabedian model and its derivatives focus on the structure-process-outcome dimensions as the basis for healthcare quality measurement and improvement, where structure encompasses the physical environment that is conducive to providing quality care, process refers to professional competence of providers and effective communication with clients, and outcome includes mortality, morbidity and patient satisfaction [
11,
29‐
35]. On the other hand, a systematic review of performance measurement and improvement frameworks in health, education and social service sectors identified 16 quality concepts and categorized them under five domains: collaboration, learning and innovation, management perspective, service provision, and outcome [
36].
For the purpose of our study, we assessed some elements of quality of intrapartum care described in both models [
29,
36]; namely, aspects of structure, process and outcome in the Donabedian framework; and aspects of learning and innovation, management perspective, service provision, and outcome in the cross-sectoral performance improvement framework. Our study also sought to assess quality of care in workplace settings through direct observation. Specifically, we assessed competence of midwives in provision of routine and emergency care during labor, childbirth, and immediate postpartum period including maternal and newborn outcomes. Secondly, we evaluated availability of essential resources for provision of quality labor, delivery, and immediate postpartum care. Thirdly, we assessed availability of opportunities for continuous quality improvement of labor, delivery and immediate postpartum care.
Discussion
Our findings demonstrate the presence of gaps to provide quality intrapartum care in government health facilities in Amhara Regional State of Ethiopia. There were major deficits in availability of essential physical resources and mechanisms for continuous performance and quality improvement. A significant proportion of midwives were also found incompetent.
Global maternal and newborn health care standards state that competent staff must be available at all times to provide quality care to every woman and every newborn [
38]. While it is encouraging that most midwives in our study are competent in providing intrapartum care, the significant proportion of midwives who displayed unsatisfactory performance in routine child birth care (1 in 6), and basic emergency obstetric and newborn care (1 in 11) makes it difficult to guarantee that every mother and every newborn will receive high quality care. It is also noteworthy that more substantial gaps were observed in rapid initial evaluation, history taking, partograph use, infection prevention, assisting normal birth, immediate postpartum care, and newborn resuscitation (Figs.
1 and
2). The World Health Organization guide for essential practice in pregnancy, childbirth, postpartum and newborn care recommends the first five care practices for every woman during childbirth and newborn resuscitation for a baby who is not breathing or is gasping [
39]. Systematic review of evidence-based guidelines also recommend partograph use for monitoring labor [
40]. While acknowledging health systems weaknesses may limit partograph use and effectiveness, a realist review of the partograph has also suggested that it may improve outcomes in low resource settings [
41]. In our study, aside from a quarter of midwives who demonstrated unsatisfactory performance in partograph use, additional ten midwives excused themselves from completing a partograph wrongly thinking there was no need to use a partograph if a woman was in second stage of labor at admission. Our study findings also indicated that midwives working in health centers had larger gaps in their capacity than those from hospitals in almost all domains. However, the difference in the composite score was not statistically significant possibly due to small sample size (Table
3). Although direct comparison is difficult due to differences in methodology, past studies from Ethiopia and other resource-constrained settings have also pointed to shortfalls in competence of midwives to provide intrapartum care [
16‐
22,
24,
42,
43].
While weaknesses in quality of the health workforce are acknowledged to be pervasive, there are also calls for better measurement and improvement of health workforce performance (especially in low and middle in-come countries) to achieve global health development goals [
13‐
15,
44‐
46]. We believe our use of direct observation to measure performance of midwives in workplace settings responds to the call for better measurement of quality of intrapartum care. The gaps uncovered also warrant strengthening pre-service midwifery education with focus on curriculum review, faculty development, use of simulation methods, and strengthening accreditation and regulation processes, among other things [
46].
All midwives have a responsibility to undertake continuing professional development activities [
47] and ensuring a high performing midwifery workforce also requires creating a work environment that fosters continuous quality improvement in every facility [
38]. Provided effective implementation, in-service training or continuing professional development, supervision and coaching, audit, feedback, and job aids coupled with an enabling environment can improve provider performance [
48‐
54]. However, our results did not show every midwife had sufficient opportunities for in-service training in general and those pertaining to intrapartum care (BEmONC, ENC or HBB, IP, and PMTCT) in particular. While it is surprising that majority of respondents did not receive training on these high priority topics, it demonstrates access to in-service training on intrapartum care remains limited in Ethiopia [
18,
42]. One explanation could be that pre-service education systems are producing midwives more rapidly than the capacity of in-service training systems to cope. Another possible explanation is gaps in targeting relevant in-service training to those who need it the most.
Other opportunities for practice-based learning and improvement (like supportive supervision, structured case discussion, clinical audit or maternal death review, job aids, and performance-based reward or incentive) were also found inadequate. Generally speaking, a higher proportion of midwives working in hospitals reported learning and quality improvement opportunities with the exception of supportive supervision visit, which was reported significantly more frequently from health centers. Our findings are consistent with program and study reports that highlighted health systems weaknesses in implementing audit and supportive supervision. Maternal death surveillance and response systems in Ethiopia [
6,
55] and globally [
56] suffer from inadequate leadership commitment at sub-national level, poor documentation and under-reporting of maternal deaths, fear of blame, and lack of trained staff, among other things. A study of barriers to quality EmONC from Ethiopia has also identified gaps in supervision including, but not limited to, being sporadic, unsupportive, and donor-driven [
42]. All these findings indicate the need for strengthening health worker performance and quality improvement strategies in health facilities.
Global standards for improving quality of maternal and newborn care also require health facilities to ensure availability of basic infrastructure and adequate stock of essential equipment, drugs and supplies for intrapartum care [
38]. However, the major gaps in availability of essential resources for provision of labor, delivery and immediate postpartum care in our study (Tables
4 and
5) is concerning as it would affect the ability and motivation [
57] of midwives to provide quality care to mothers and newborns. A higher proportion of health centers than hospitals had resource gaps. This assumes greater significance when one takes into account the fact that health centers are the primary and most accessible birthing facilities for most women in Ethiopia. In addition to reducing effectiveness of maternal and newborn healthcare, weak infrastructure can undermine the demand to deliver in health centers [
58].
Maternal and newborn care surveys from Ethiopia, Tanzania, Uganda, Kenya, Namibia, and Bangladesh have all reported gaps in availability of essential commodities. A basic emergency obstetric and newborn care survey of health centers from Addis Ababa, Ethiopia, found that only 50% had parenteral antibiotics and diazepam; none had magnesium sulfate; and only 90% had a functional vacuum extractor [
18]. Inadequate equipment and supplies, and lack of knowledge and skills in performing EmONC were the two main challenges identified in a study of maternity care services in Moshi urban district of northern Tanzania [
24]. Another study in Tanzania involving qualitative interviews with nurse-midwives in basic and comprehensive EmONC facilities also revealed that nurse-midwives lacked essential supplies to do their job [
25]
. A health facility-based survey from Karamoja region of Uganda reported lack of equipment and supplies as the most frequent reason for not performing EmONC signal functions and found that 50% of health centers lacked basic equipment for normal delivery and some lacked equipment for neonatal resuscitation as well as consumable supplies and drugs [
26]. Emergency obstetric care readiness assessment in rural northwest Bangladesh found that availability of EmONC specific medicines and commodities was 62% in public facilities while coverage for equipment and supplies was 90%. Half of the respondents also mentioned not having essential medicines and commodities in stock as main constraint to EmONC provision [
27]. Evaluation of clinical quality of maternal and newborn care in Kenya and Namibia found gaps in essential drugs and commodities including oxytocin, magnesium sulfate, antibiotics, and incubator [
28].
Our study findings add to a growing body of literature reporting health system weaknesses to ensure quality of maternal and newborn healthcare. The 2016 Lancet maternal health series has shown access to good quality and evidence-based care remains inadequate especially in low income countries owing to gaps in provider skill and number, facility capability, basic infrastructure for intrapartum care, availability and implementation of evidence-based guidelines, and access to care, among other things [
40]. Recent multi-country analyses of health systems bottlenecks in high burden countries have also acknowledged providing quality labor and childbirth care, basic newborn care, and neonatal resuscitation is a challenge, with the most significant weaknesses reported from African countries. Health financing, health workforce, service delivery, and essential commodities related challenges were identified as the major bottlenecks [
59,
60]. A systematic review of providers’ perspectives on barriers to quality midwifery care in low and middle income countries have also found professional barrier, which includes, but is not limited to, gaps in education and training, and lack of equipment and supplies, was the most frequently mentioned impediment [
61].
Strengths and limitations
We believe the assessment of quality of care provision during the most critical periods for the mother and the newborn (labor, childbirth, and the immediate postpartum period) makes our study timely and relevant for the global and national maternal and newborn health community. Our attempt to measure the structure (availability of resources for intrapartum care), process (competence of midwives in routine and emergency obstetric and newborn care), and outcome (maternal and newborn morbidity and mortality) dimensions of quality of care as well as strategies for continuous performance and quality improvement is also noteworthy. Moreover, the use of multiple methods including direct observation to measure performance and availability of essential resources lends credibility to our findings. The assessment of quality of both routine childbirth care and emergency care is also important. However, the exclusion of facilities with low volume of delivery services (less than one delivery per day), replacement of some health centers with hospitals (due to challenges with finding expected number of midwives and laboring mothers), and missing data (especially during inventory of commodities) may be considered limitations. Even if we provided brief descriptors of performance in the data collection tool, trained data collectors and conducted pretesting, the subjective judgement involved in performance evaluation can be a source of measurement error but we could not estimate inter-rater or intra-rater reliability. However, internal consistency of the items was found to be very high (Chronbach’s Alpha of 0.94) suggesting the reliability of our results.