Changes in health facility readiness for obstetric and neonatal care services in Nepal: an analysis of cross-sectional health facility survey data in 2015 and 2021

Data were drawn from the 2015 and 2021 Nepal Health Facility Surveys (NHFSs), with data being publicly available from https://​dhsprogram.​com/​Data/​. Both are comparable, nationally representative and cross-sectional surveys comprising the following components: inventory assessment; service provider interviews; observations of client-provider sessions for selected services; and exit interviews with patients or those taking care of patients for the selected services upon being discharged from a HF or leaving the service site. This study used data from the inventory assessment and delivery service provider interviews.

This analysis comprised all formal sector HFs of Nepal that reported providing normal low-risk delivery services when data for the 2015 and 2021 NHFSs were collected. The NHFSs obtained their sampling frame from the Ministry of Health and Population of Nepal, which included 4,719 HFs in 2015 and 5,681 HFs in 2021 [12]. The increase in the number of HFs in 2021 was primarily due to the establishment of community health units by the local governments after the country’s health system was federalized in 2017. The NHFSs in 2015 and 2021 surveyed 940 and 1,564 HFs out of which 48.6% (457) and 51.4% (808) HFs respectively provided normal low-risk delivery services. The sample size of the NHFSs in 2015 and 2021 are comparable and allow for representative estimates nationally and by managing authority. The larger sample size in the NHFS 2021 accounts for the seven provinces of a federal Nepal, in contrast to the five administrative regions in 2015.

The HFs included in this analysis comprise all public hospitals offering basic healthcare services and all primary health care centers, and a sample of other basic healthcare centres (health/sub-health posts, community health units, and urban health centers). Similarly, for private HFs, this analysis in 2015 comprises sample of private hospitals with 15 or more in-patient beds, and complete enumeration of all private hospitals with 100 in-patient beds. In 2021, the analysis includes a sample of private hospitals with at least one in-patient bed, but complete enumeration of all private hospitals in the provinces which have fewer private hospitals.

The WHO Service Availability and Readiness Assessment (SARA) framework is used which is composed of three domains: i) staff and guidelines; ii) equipment; and iii) medicines and commodities [16]. To make the analysis of HF readiness for normal low-risk delivery services more granular and contextualised for Nepal, this study sub-divided the domain “staff and guidelines” into two distinct domains, i.e. “trained provider” and “guidelines”. The availability of guidelines merits to be assessed separately due to continuous changes in the evidence base informing health care delivery and related revisions of guidelines. Similarly, the domain “medicines and commodities” was sub-divided into a domain on “essential medicines for mothers”, and a domain on “essential medicines for newborns” to ensure adequate attention to newborn care. Each of these five domains received equal weight. For the BEmONC service functionality analysis, seven signal functions prediscribed by the SARA framework were used [16]. As the SARA framework does not contain a separate measure for IPC, this study applied selected domains of general HF readiness – i.e., i) trained provider, ii) guidelines, and iii) equipment and supplies – and used data on universal precaution measures and IPC available in the surveys. Each of the three domains received equal weight. This study focused on supply-side issues at the level of HFs and the domains recommended by WHO’s SARA manual. In contrast, further critical determinants of HF readiness according to the WHO health system building blocks framework, notably governance, information system and financing, where not explicitly addressed, as these primarily exert their influence at local or national government level, rather than at HF level.

For each HF, three summary indices were calculated and used as the main outcome variables in subsequent statistical analyses: (a) HF readiness index for normal low-risk delivery services; (b) BEmONC services functionality; and (c) IPC index.

Original variables: Tables 1, 2 and 3 show the original indicators and signal functions available in the NHFS 2015 and 2021 dataset for each SARA domain.

Calculation of domain scores and indices: For each SARA domain, a domain score was calculated (see Tables 1, 2 and 3, column ‘Calculation’), ranging from 0 to 100%. At the end of each table, a summary index is described, calculated as the mean of the related domain scores, and ranging from 0 to 100%.

Data analysis was conducted using IBM SPSS Statistics 25. Sample weights were applied to ensure the actual representativeness of findings at the national level and according to the managing authority. Initially, the weighted mean of the three outcome variables with their domains and sub-components, overall and stratified by the managing authority, was calculated for 2015 and 2021 respectively. Subsequently, a weighted t-test for independent samples was carried out to test for a statistically significant change over time in each of the outcome variables. The weighted mean difference of the change observed between 2015 and 2021 and its 95% confidence interval (CI) were calculated. A level of significance of 0.05 was assumed for all analyses.

The 2015 and 2021 NHFSs obtained ethical approval from the Nepal Health Research Council, while this study was approved both by the Ludwig-Maximilians-Universität (LMU Munich) Ethics Commission, Munich, Germany and by the Nepal Health Research Council in June 2021.

As shown in Table 4, this study analyzed 457 and 804 HFs providing normal low-risk deliveries in 2015 and 2021 respectively. In both years, more than eight out of ten HFs providing normal low-risk delivery services were public basic healthcare centres;

As shown in Fig. 2, nationally, the average HF readiness index for normal low-risk delivery services was below 50% in both years despite a statistically significant increase from 37.9% in 2015 to 43.7% in 2021. There was, however, considerable variation between domain scores. Between 2015 and 2021, the greatest improvement was observed for essential medicines for newborns (statistically significant increase from 42.0% to 53.6%), followed by essential medicines for mothers (from 76.6% to 85.8%), equipment and supplies (from 70.0% to 78.2%), and trained providers (from 62.2% to 71.1%). Only one score showed a statistically significant decrease, i.e. the use of guidelines for essential delivery care decreased from 21.8% to 12.8%. Details are presented in Supplemental Table a.

Table 5 shows inter- and intra-domain variability of HF readiness. Availability of all 13 basic equipments and supplies for normal low-risk delivery services was excessively low and insignificantly increased from 6.5% in 2015 to 8.8% in 2021. Of the 13 items, the availability of nine significantly increased: emergency transport, examination light, delivery pack, neonatal bag and mask, delivery bed, partograph, blood pressure set, latex gloves, and infant weighing scale with levels in 2021 between 82.0% and 98.7%. The availability of the four remaining items – manual vacuum extractor, vacuum aspiration kit, suction apparatus and sterilization equipment – increased but not in a statistically significant manner.

Nationally, one in two HFs (49.8%) had in stock all five essential medicines and commodities for mothers in 2021 which represents a significant increase from 29.7% in 2015. The availability of injectable uterotonics, skin antiseptics, and intravenous solutions with an infusion set was high, being in stock at nearly nine out of ten HFs in 2015, which improved significantly to nearly all HFs in 2021. Injectable antibiotics were available at two-thirds (66.1%) of facilities in 2021, representing a significant increase from 40.9% in 2015. Seven out of ten HFs stocked magnesium sulphate in both years. Compared to the availability of all five essential medicines and commodities for mothers, the availability of all five essential medicines for newborns in both surveys was extremely low and changed only slightly from 0.7% in 2015 to 2.2% in 2021. Stockout of tetracycline eye ointment was paramount, with only one in ten HFs having this stock in 2021. The availability of 4% chlorhexidine antiseptics, amoxicillin syrup, ceftriaxone injection, and injection gentamycin significantly increased between 2015 and 2021, ranging from 38.1% to 80.2% in 2021.

Overall, at the national-level, functionality of BEmONC services was inadequate in HFs providing normal low-risk delivery services and did not show relevant improvements by 2021. As Table 6 shows, BEmONC service functionality was extremely low in both years, 4.2% in 2015 and 2.5% in 2021. Of the seven signal functions, the administration of parenteral oxytocin was high in both years and slightly increased to 88.2% by 2021, while all other signal functions decreased between 2015 and 2021. A significant decrease in functionality was found for assisted vaginal delivery, neonatal resuscitation, removal of retained products of conception, and manual removal of the placenta. The administration of anti-convulsants and assisted vaginal delivery were performed least frequently, below or at ten percent, among the seven signal functions in both years.

Figure 3 shows, on average, that the IPC index among HFs providing normal low-risk delivery services is very low in Nepal, despite a statistically significant increase from 6.0% in 2015 to 8.0% by 2021. Inter-domain variations were large. The availability of at least one provider trained regarding IPC had the highest score in both years and significantly increased from 79.7% in 2015 to 86.4% by 2021. The availability and functionality of IPC equipment and supplies had the largest increment in score (20.5% points), reaching 72.6% by 2021. The availability of guidelines on IPC was below ten percent in both years, which significantly lowered the overall IPC index. Details are presented in Supplemental Table b.

The availability of all ten items under IPC equipment and supplies increased significantly from 0.7% in 2015 to 6.3% by 2021 (mean difference: 5.5, CI: [3.7; 7.4], p < 0.0001), however, these levels are far too low to ensure appropriate IPC measures. The availability of all four items of personal protective equipment – latex gloves, surgical masks, eye protection, and gowns or aprons – significantly increased from a mere 5.9% in 2015 to 51.1% in 2021 (mean difference: 45.3, CI: [41.2; 49.3], p < 0.0001), however, this level is still considered inadequate, particularly for eye protection supplies.

The HF readiness index for normal low-risk delivery services showed statistically significant increases from 37.6% to 43.6% among public HFs, and from 39.9% to 45.3% among private HFs between 2015 and 2021 respectively, thus showing small differences by managing authority. Differences regarding the changes in HF readiness domains between public and private HFs are reported in Supplemental Table a.

In 2021, the disparity in BEmONC service functionality of HFs providing normal low-risk delivery services by managing authority was large: public 1.8% and private 10.5%, which was a decline from 3.0% in public and from 15.7% in private HFs from 2015 respectively. The statistically significant changes in two signal functions among the public HFs are reported in Supplemental Table c.

The IPC index was found to be low for both managing authorities, with a slightly better score among private HFs (6.9% in 2015, 8.4% in 2021) than among public HFs (5.9% in 2015, 7.9% in 2021). The statistically significant changes in domain level scores between public and private HFs are reported in Supplemental Table b.

In the context of the constitutional right of the Nepali people to free basic healthcare services, the mandate for local governments to manage their basic healthcare services, the inclusion of normal-low risk delivery services into the basic healthcare package, and a strategic focus on improving the QoC, the present study analyzed nationally representative data from HFs providing low-risk delivery services with regard to their readiness, functionality of basic emergency obstetric and newborn care services, and infection prevention and control.

The readiness of HFs in Nepal for normal, low-risk delivery services is suboptimal. While there have been improvements in the HF readiness index from 37.9% in 2015 to 43.7% in 2021, these changes are small and equate to a 0.9% annual improvement rate. This indicates that the majority of HFs in Nepal have low scores for HF readiness. A number of studies published using the NHFS 2015 data identifies inadequate number of service providers, irregular services, supply chain issues being key challenges for continuum and quality of delivery care across both public and private HFs [24, 25]. The low readiness of HFs to provide high-quality, normal, low-risk delivery services is not surprising because the QoC for other basic healthcare services is also sub-optimal. For instance, less than two percent of HFs meet minimum standards of care for antenatal, family planning, and sick child services [12, 26]. Similar results have been observed in other low- and middle-income countries (LMICs). For example, in Nigeria, between 2005 and 2009, there was no statistically significant progress in the availability of trained providers for labor and delivery care (from 50.8% to 57.1%). Indeed, the readiness of HFs to provide maternal health services was low in both years [27]. One major challenge for LMICs is the lack of commonly agreed upon QoC indicators, which hinders measuring progress nationally and makes it difficult to compare indicators across countries. The use of the Minimum Service Standard (MSS) tool in public HFs is a relatively new concept but its use has now standardized monitoring of HF readiness in Nepal. The reporting of MSS scores as part of the routine Health Management Information System database has helped local governments and HFs identify gaps in HF readiness, link them with service utilization data, and take corrective actions. This tool has led to positive changes and that policy implications are to strengthen this standardized monitoring.

In this study, wide variation across the five domains of HF readiness was found between the survey years. Four out of five domains, namely equipment and supplies, essential medicines for mothers, trained providers, and essential medicines for newborns, significantly improved in 2021 with large variations within the components. This could be due to the focus on improving QoC during the implementation of the Nepal Health Sector Strategy (NHSS) 2016–2022. In contrast, the fifth domain, the availability of guidelines on delivery care, performed the least and significantly decreased from 21.8% in 2015 to 12.8% in 2021. This is low compared to other LMICs such as Haiti and Tanzania where delivery care guidelines were available in 24.1% and 29.8% of the HFs in 2017 and 2019, respectively [26, 28]. The low and diminished availability of guidelines in 2021 in Nepal could be due to COVID-19 restrictions, resulting in a largely virtual mode of operations and halting the printing of guidelines or their dissemination to HFs. Nevertheless, during the COVID-19 pandemic, digital guidelines, particularly on COVID-19 topics, were provided to HFs. Program guidelines comprise evidence-informed recommendations intended to standardize and optimize care for patients, and their use during clinical decision-making is intended to improve healthcare outcomes [29, 30]. Even though both paper and digital guidelines were included in this analysis, the low results in 2021 suggest little or no effort made to improve guideline availability since 2015. Anecdotal evidence indicates improper storage of digital guidelines in electronic devices, that cannot be accessed by service providers when needed. Improving the skills of providers in accessing these guidelines on computers and mobile devices can support increased use of the digital guidelines. The COVID-19 pandemic disrupted regular trainings [31, 32], but Nepal was able to significantly increase the availability of trained delivery service providers by 2021. Reaching the SDG target of attaining 90% skilled birth attendance was a focus of NHSS (2016-2022), and several training packages were available to train the delivery care service providers on enhancing their knowledge and skills, which were continued virtually during COVID-19. However, gaps still exist, with nearly one in three HFs lacking a provider trained in delivery care. Human resource availability and their skills remain a significant health system challenge that is not unique to Nepal but relevant for many other LMICs [33].

Supply chain issues have emerged, particularly with essential medicines for newborns and basic equipment and supplies. The frequent lack of equipment could be a reason for the low rates of assisted deliveries in HFs. Additionally, the low availability of sterilization equipment increases the risk of nosocomial infections for both the mother and the newborn, and the low availability of tetracycline ointment for the treatment of newborn eye infections illustrates poor readiness for newborn care. This is alarming since the NMR has not been reduced below 21 deaths per 1,000 live births since 2016 [6]. Although the stock of essential medicines for mothers was better than for newborns in both managing authorities, there is still a significant lack of supply, mainly for injectable magnesium sulphate and injectable antibiotics. This is alarming because in 2021, 12.0% and 11.2% of maternal deaths during the post-partum period were due to hypertensive disorders and infections respectively; and 10.2% of the maternal deaths during delivery had fits, seizures, and convulsions [2]. Data from 28 hospitals over the four-year period from 2015 to 2018 reports eclampsia to be the leading cause of maternal deaths (19.0%). The availability of magnesium sulphate in LMICs is consistently low. A study published in 17 LMICs in 2018 revealed that magnesium sulphate was amongst the least available essential medicines and commodities for mothers (median: 63%, range: 10.0–97.0%) [34]. It is important to ensure a reliable supply of magnesium sulfate and antibiotics and enforce its correct use in training. The supply chain system of any country is affected by the service delivery structure. Prior to Nepal being declared a federal republic in 2017, medicines were procured by the central government system. Now, all three tiers of government procure medicines, but there is a lack of clarity on which level of government should be responsible for which task. As a result, there is often over or understocking of medicines at HFs. The overall capacity of local governments to procure medicines through systematic data-driven forecasting and following public procurement guidelines is limited. While both push and pull supply-chain systems are in place, establishing a strong supply chain system remains an ongoing challenge. Data-driven procurement, efficient transportation management, and improving governance at all levels should be prioritized.

Fifteen percent of pregnant women develop a potentially life-threatening complication that requires skilled medical care, and some may even require major obstetrical intervention for survival [35]. According to verbal autopsy reports on maternal deaths conducted in Nepal in 2021, the majority of maternal deaths occur during the postpartum period due to obstetric hemorrhage and take place at HFs [2]. This underscores the importance of functional BEmONC services. Sadly, in 2021, only 2.5% of HFs had functional BEmONC services, indicating a decline from 4.5% reported in 2015, which may be attributed to health system challenges arising from the COVID-19 pandemic such as HF closures or disfunctional supply-chain systems. In 2022, eight out of ten deliveries at HFs were attended by skilled birth attendants at HFs [1]. This means that the majority of HFs can provide assisted vaginal delivery services to pregnant women in need. However, in 2021, only 16.2% of HFs offering normal low-risk deliveries reported having assisted vaginal delivery services available. Private HFs consistently had higher BEmONC service functionality in both surveys than public HFs; this could be because all private HFs analyzed in this study were hospitals, unlike only about 5% of the public HFs analyzed. Studies on BEmONC in Africa and Pakistan have identified a shortage of trained providers, high absenteeism rates, poor ambulance services, and supply chain issues as barriers to quality BEmONC services [28, 36]. These barriers are also relevant in the Nepalese context. Other obstacles to quality BEmONC in LMICs include poor provider remuneration and demoralization, high turnover, increased workload and burnout rates, poor coordination, inefficient referral mechanisms, inadequate allocation of limited resources, lack of training and monitoring, and inequality in the distribution of BEmONC sites [37]. In Nepal, under the Aama Program, HFs receive unit costs for attending normal-low-risk, complicated, and cesarean deliveries, a small amount of which is provided to the birth attending team, which can motivate the providers [38]. However, this study did not analyze the barriers and the motivation factors in depth.

With improved access to HF-based delivery services, eight out of ten women giving birth in HFs [6] and over half of the maternal deaths occurring in HFs, the low functionality of the BEmONC service and lack of progress are alarming in fulfilling the SDG commitment of reducing maternal deaths. The low volume of deliveries in basic healthcare centers, due to improper location and inadequate HF readiness, results in overcrowding in hospitals [0 24]. Considerable home deliveries, lack of awareness of the Aama Program, and the impact of the COVID-19 pandemic for 2021 could explain the low BEmONC service functionality in Nepal. Furthermore, these findings reinforce the thrust of the Nepal Safe Motherhood and Newborn Health Road Map (2030) and NHSSP 2023–2030 and suggest re-thinking the strategic locations of the HFs providing BEmONC services, and creating demand for those HFs, as well as improving services, including staffing in hospitals.

Universal precautions for IPC are important not only in the face of the COVID-19 pandemic and other disease outbreaks but also in routine healthcare provision. Birth itself and vaginal examinations make pregnant women more susceptible to infections. Despite significant improvements in the IPC index between 2015 and 2021 found in this study, the 2021 IPC levels of 7.9% for public HFs and 8.4% for private HFs are excessively low. Although the availability of trained providers on IPC and personal protective equipment significantly improved in 2021 due to COVID-19 response efforts, the availability of IPC guidelines is also low, and the overall IPC status is inadequate. The low readiness for IPC found in this study is comparable to HFs in Bangladesh, where only 16.4% of the standard precaution elements were available for delivery and newborn care services in 2017 [29], and another analysis of 17 LMICs where the median national availability of sterilization equipment was 51.0% in 2018 [34]. Items such as waste receptacles and needle destroyers were often not available in HFs, indicating a need to strengthen supply chain systems and put into place quality improvement teams [13, 39]. In the new federal system in Nepal, resources can be leveraged from local governments for local purchase of these easily available items.

This study has both strengths and limitations. It is the first study to draw on data from two nationally representative NHFSs implemented under the global Demographic and Health Survey (DHS) Program to measure changes in HF readiness for normal low-risk deliveries. This occurred over a time period when Nepal was hit by a 7.8 Richter Scale earthquake and the recent COVID-19 pandemic and amid the country's transition to a federal government system. Methodologically, the NHFS derived a sampling frame from Nepal's national health system that included both public and private HFs. The survey instruments were validated, and verbal responses were confirmed by observations. Data were collected in tablets by thoroughly trained medical, nursing, or public health personnel. Data quality was checked daily by the local research firm and the DHS Program, and the analysis was guided by publicly available code books, weighing techniques, and tutorials at https://​dhsprogram.​com/​Data/​. Although data collection for the 2015 and 2021 NHFSs was impacted by the earthquake and the COVID-19 pandemic, respectively, standard operating procedures were fully followed, and there were no major quality concerns except that data collection was delayed and took longer than planned. However, there are some limitations. These include, for example, the small sample size for private HFs since only private hospitals were surveyed, and the fact that stock data were impacted by the timing of receipt of the stocks at HFs, and that the 2015 earthquake and the COVID-19 pandemic made supply chains vulnerable and irregular. General amenities like waiting areas, sanitation, and drinking water facilities were not analyzed, and newborn care was not comprehensively covered in the analysis. This study focused only on the variables recommended by WHO’s SARA manual and did not analyze other health system factors.

The policy environment for improving QoC is progressing in Nepal. However, several challenges exists that need to be addressed. For instance, the establishment of an accrediting body for quality assurance in the public and private health sector, as envisioned in the NHSSP (2023–2030), has not been fully effective yet; local governments have limited human resource capcity to provide basic healthcare services; monitoring mechanisms for basic healthcare services remain unclear; and the lack of accountability mechanisms is hindering progress which impacts the HF readiness to provide delivery services. The standards for normal low-risk deliveries are equal for all HFs regardless of their size and managing authority [16, 23, 40, 41]. While the private sector in Nepal is expanding, it represents the first point of contact for women seeking delivery services in several parts of the country [1] but monitoring QoC in the private sector is challenging due to unclear regulations. Comparatively, private HFs have slightly better readiness than public HFs, with better availability of essential medicines and equipment. However, they lack adequately trained providers and delivery care guidelines. While they are more ready to provide BEmONC services, and generally more resourceful compared with public HFs, they need a better operating environment and a motivation to implement the Aama Program, which can create a win–win situation for both the public and the private sectors, ensuring high QoC for pregnant women.

The earthquake in 2015 claimed the lives of over 9,000 people, while the COVID-19 pandemic resulted in a death toll of over 12,000 people [42] and continues to impact Nepal's health system. Although Nepal's transition from a unitary to a federal country allowed for local planning and management of basic healthcare services, unclear demarcation of authority between jurisdictions in the three tiers of government led to slow decision-making and adversely affected healthcare service delivery [43]. This study indicates that, following the shift to a federal system, increased emphasis was placed on expanding the number of HFs, but not enough attention was paid to improving their readiness and the QoC. The challenges of inadequate and insufficiently skilled staff at HFs need to be addressed by local governments through robust planning, budgeting, implementation, and oversight of basic healthcare services. Similarly, clarifying the roles and responsibilities of the three tiers of government in the delivery of basic healthcare services and fostering proper coordination among them remain crucial to ensure HF readiness for high-quality services, including normal low-risk births, obstetric complication management, and improved maternal survival.

Data availability for QoC is improving in Nepal [12]. The implementation of previous health sector strategies has informed the government and other stakeholders to develop the NHSSP (2023–2030), the Nepal Safe Motherhood and Newborn Health Road Map 2030, and a nationally representative, comprehensive NHFS to monitor changes in QoC at five-year intervals. The NHFS 2021 is the key data sources to track HF readiness indicators of the NHSSP (2023–2030), its continuity will provide opportunities for researchers and policymakers to analyze trends and rates of changes in HF readiness for delivery and other basic healthcare services. However, Nepal also needs a functional information system that can routinely track QoC indicators in both public and private sectors. While the Health Management Information System and Logistic Management Information System monitor service utilization rates, medicine stock, and Minimum Service Standard scores at HFs, they are limited in monitoring other important QoC indicators.

Improving QoC needs to remain a top priority in the health system in Nepal, with clear roles, evidence-based strategies, and interventions designed for all providers, including the three tiers of government and the private sector. The gaps in HF readiness identified in this study, as well as those in NHFS 2021, the Minimum Service Standard tool implementation in public HFs, and other studies, need to be collated to generate knowledge and translate them into budgeted plans for the relevant governments. These actions should ensure effective supply chain management, training and regular availability of service providers, availability of program guidelines, and functioning quality improvement systems. The local governments and HFs need to leverage internal and external resources and monitor the effective implementation of basic healthcare services, which are key to increasing HF births, managing obstetric complications, preventing women from nosocomial infections, and ultimately reducing maternal deaths. The results of this study can complement the baseline and review of the result framework for the NHSSP (2023–2030), with suggested QoC indicators, monitoring approaches, and support for tracking progress over time.