Accountable Care Reforms Improve Women’s And Children’s Health In Nepal

Despite accelerated progress in reducing maternal and child mortality around the world,¹ eight hundred women die from preventable causes related to childbirth every day, and nearly six million children under age five die each year.^2,3 The vast majority of these deaths could have been prevented through low-cost and proven interventions such as antenatal and safe delivery care; neonatal resuscitation; management of acute childhood illnesses like diarrhea, pneumonia, and malaria; and achieving appropriate spacing between births.^4–7

While low- and middle-income countries continue to face the disproportionate burden of these deaths, wide disparities in outcomes exist even across countries with similar income levels.⁸ Economic resources are not the only determinants of the health of populations. Low- and middle-income countries that have made progress in reducing child and maternal mortality, such as Nepal and Bangladesh, have developed and adapted a set of strategies that enable better outcomes in resource-constrained environments.⁹

Achieving these solutions often entails bypassing traditional medical approaches, such as activity-based point-of-care payments to providers, and instead builds upon existing social or community services to deliver better, more coordinated care. Innovations and reforms in the delivery of care should be encouraged and supported by the alignment of funds toward patient-centered care. These types of innovations and reforms are common features of accountable care, which we define as an arrangement in which “a group of providers are held jointly accountable for achieving a set of outcomes for a defined population over a period of time and for an agreed cost.”¹⁰ Accountable care policy with concomitant payment reforms provides an opportunity for organizations to shift resources to improve care and be held accountable for achieving better health outcomes. These resource shifts can spur the adoption of innovations in care delivery that are traditionally inadequately reimbursed. Many countries are applying accountable care reforms in a variety of local settings, evidenced by the wide geographical diversity in the uptake of the reforms.¹⁰ The range of global cases provides an opportunity to further accelerate adoption by sharing implementation experiences across contexts.

Elsewhere we have presented the core components of accountable care policies. These include population, performance measures and outcomes, continuous improvement, payments and incentives, and coordinated care delivery.¹⁰ We later expanded this framework to encompass the health policy environment, including institutional, political, and regulatory factors and organizational competencies, such as technological capabilities or financial readiness to share risk, that support accountable care implementation.¹¹

Here we apply this accountable care framework to analyze the experience of Possible and the Ministry of Health of Nepal in forming a public-private partnership to deliver an integrated reproductive, maternal, newborn, and child health intervention in Achham district in rural Nepal.¹² Nepal poses a challenging environment for health care delivery in general and public-private partnerships in particular. Limited administrative capacity and acute fiscal shortages have constrained public-sector investments: Health expenditure is 5.8 percent of gross domestic product (GDP)—much less than the global average of 9.9 percent.¹³

In Achham, Possible implements community health care programs and independently manages the government-owned, district-level Bayalpata Hospital, a sixty-five-bed facility that serves over 85,000 outpatients and 2,500 inpatients per year. Through a hub-and-spoke model, Possible employs a cadre of community health workers to deliver care to patients in rural areas and conduct disease surveillance. Through community health workers, Possible has expanded the scope of care delivery in the region, allowing outreach to remote villages.

The public-private partnership has been operational for over eight years and has demonstrated improved outcomes as measured by indicators like the institutional birth rate (defined as the proportion of births taking place in health care facilities). In this article we describe the impact of the model over an eighteen-month period during 2015–16 as Possible improved its system by strengthening active surveillance by community health workers, integrating digital health information, and increasing its monitoring and supervision capabilities. Because Possible and the Ministry of Health developed performance measures as part of program implementation, for this study we were able to use these measures to describe their model. We evaluated population-based indicators at the household level across time within the catchment area, including antenatal care attendance, institutional birth rates, and rates of postpartum contraception use. We also assessed how patients’ knowledge, attitudes, and practices changed during the intervention. Additionally, we analyzed quality improvement data that Possible collects, to better understand how staff members are applying lessons learned. The availability of these data enabled us to conduct an analysis of a recently implemented accountable care system.

Context And Model

Nepal’s nominal per capita GDP of roughly US$730 places it among the poorest countries of Asia; its public-sector health care expenditure of US$40 per capita places it among the lowest in the world.^13,14 Within Nepal, Achham is characterized by remoteness, poverty, and social disruption due to mass migration. As a result, it has among the lowest human development indicators among the seventy-five districts of Nepal, with over 40 percent of the population living on less than $3.10 (purchasing power parity) a day.¹⁵

In this context, Possible and the Ministry of Health pioneered a public-private partnership to deliver high-quality, affordable care in areas where the public sector has traditionally struggled. The partnership builds off of Nepal’s longstanding Female Community Health Volunteer program by integrating community health workers into facility-based care to improve reproductive, maternal, newborn, and child health across the district. The community health workers provide three core functions: surveillance of conditions in the community; triage, referral, and care coordination with facilities; and community-based counseling, education, and diagnosis. They serve the villages surrounding Bayalpata Hospital and referral facilities, including government village clinics (known as health posts). The scope of care includes comprehensive reproductive, maternal, newborn, and child health services and chronic disease services, though this article focuses on the former.

The program uses accountable care principles to drive improvements in care. The Nepal government pays Possible to provide care for women through pregnancy and delivery. To risk-stratify and identify the population of pregnant women, community health workers continuously survey the population for new pregnancies, help women obtain laboratory and ultrasound testing to identify high-risk pregnancies, and follow those pregnancies through the postpartum period. Patients are then enrolled in an open-source electronic health record (EHR) platform, to remotely monitor each one’s care over time and link the health workers’ data with hospital data. Possible uses key performance measures tied to patient outcomes and process improvements that are incorporated into the financial contract. Community health workers generate data at the point of care, via a smartphone app, which provides care coordination and decision support. This enables providers to analyze key performance results weekly, quarterly, and annually to make continuous improvements. Twenty percent of the partnership contract is linked to performance indicators, which creates a system-level incentive for high-quality care. Possible uses electronic records and linked data analytics to help coordinate and transform the delivery of care, enabling community health workers to collaborate with hospital providers to ensure that women receive appropriate care for pregnancy and delivery. Possible has also implemented new steps for patient involvement by engaging mothers in developing birth plans, educating families on peripartum danger signs, and soliciting input from a community advisory board consisting equally of both men and women and of representatives of historically marginalized groups.

Study Data And Methods

Participants

The data presented here were collected as part of a separate study assessing the impact of group antenatal care and home visits by community health workers.¹⁶ For this article we assessed the results from fourteen village development committees (hereafter, village clusters), which served as the original pilot of the model before its expansion. Participants eligible for inclusion in the study included reproductive-age (ages 15–49) female patients residing within the study village clusters. Although the catchment area remained constant, the methods for counting the population for a continuous census through digital means were under development throughout the study. This led to discrepancies in census households within the catchment area, since different women enrolled in the pre and post periods.

Community health workers identified women at Bayalpata Hospital or within the community, regardless of economic status or caste.

Data Source And Analysis

We used four data sources to assess the program. The online Appendix provides greater detail on data sources, storage, and access.¹⁷

To evaluate population-based indicators, we developed a pre-post census survey. We summarized one-year baseline (November 2014 to February 2015) and endline (February to July, 2016) outcomes and mortality using frequency tables and rates per thousand deliveries or live births (as applicable). All outcomes are binary (coded 0/1): institutional birth rate (the primary outcome); four or more visits of antenatal care; contraceptive use at one year after delivery; and all infant deaths (defined as neonatal or infant deaths at less than one year, excluding stillbirths). For each outcome, we calculated crude percentage differences of that outcome in periods of pre- versus post-intervention.

We followed a similar approach to assess the significance of pre-post changes. For each outcome we fitted logistic models with a single predictor indicating pre-post time period and used the p value of the estimate associated with that parameter as the basis for a test of significance of pre- versus post-intervention change in outcome. To better visualize the data, we present outcomes as percentages.

We assessed patients’ knowledge using a Knowledge-Attitudes-Practice (KAP) survey (see the online Appendix),¹⁷ using convenience sampling to select participants. We conducted the baseline assessment from November 2014 to February 2015 and the endline assessment from February to July 2016. We scored baseline responses based on correctness and then added these scores to obtain a total score in each knowledge category. We then conducted Wilcoxon rank sum tests of the median change in knowledge scores over time (baseline score subtracted from endline score to compare the change in knowledge of antenatal care). The level of significance for all analyses was $\alpha =0.05$.

To analyze improvements in quality, we used data from EHRs from care delivery and follow-up visits. We portrayed the data in a manner consistent with the accountable care approach, where data serve the primary purpose of program improvement and are made available for policy analysis.

To assess program costs, we analyzed expenditures for 576 deliveries over one year using Possible’s financial system (Intacct software).

Limitations

The study examined a new integrated system in the early pilot and improvement stage, which presented limitations to our analysis. Possible collected data primarily to facilitate rapid assessment and learning, not for program evaluation. As a result, the assessment was nonrandomized, lacked a control group, and was conducted over a limited period of time.

Possible’s approach to evaluation is to embed digital data collection into the architecture of care delivery. This makes it challenging to collect data in “control” areas where care is not being delivered. This approach is consistent with the Ministry of Health and Possible’s goal of generating evidence for policy in an affordable and ethical manner.

Additionally, it is difficult to isolate the effects of specific community- and hospital-based care reforms on the observed results. The intervention took place concurrently with a study that assessed the impacts of group care and home visits.¹⁶

Lastly, the intervention was conducted within a public-private partnership framework that, by design, operates quite differently than both the private- and public-sector providers that represent the vast majority of the health care ecosystem in Nepal and elsewhere.

Study Results

Pre-Post Population Survey

The pre-post study population characteristics are shown in Exhibit 1. A total of 682 women who delivered during the study period were identified in the community at baseline and 541 at endline eighteen months later. The results at eighteen months are shown in Exhibit 2. The institutional birth rate increased by 11.8 percentage points (95% confidence interval: 5.3, 18.3; $p<0.001$); completion of antenatal care increased by 6.4 percentage points (95% CI: 0.1, 11.8; $p=0.001$); and rates of postpartum contraception increased by 27.5 percentage points (95% CI: 20.8, 34.2; $p<0.001$). Infant mortality decreased from 18.3 per thousand to 12.5 per thousand live births, for a difference of - 5.8 per thousand (95% CI: −27.7, 16.1; $p=0.41$) (results not shown).

SOURCE Authors’ analysis of census survey data. NOTES Baseline: November 2014 to February 2015; endline: February 15 to July 10, 2016. Institutional birth rate is the rate of women giving birth in a health care facility with a health care professional present. Receipt of antenatal care involves the completion of four antenatal visits. Postpartum contraception is the use of contraceptives one year after delivery. ****$p\le 0.001$

Knowledge-Attitudes-Practice Survey

A subset of 136 women were surveyed at baseline and 117 at endline, as a result of 9 percent attrition. According to the Wilcoxon sign-ranked test, significantly more participants improved on the KAP survey than declined or stayed the same for antenatal and post-natal care ($p<0.001$). However, for birth planning scores, there was no statistically significant difference between the number of people whose knowledge improved, declined, or stayed the same (Exhibit 3).

SOURCE Authors’ analysis of data from the Knowledge-Attitudes-Practice (KAP) survey conducted among 117 intervention participants selected through convenience sampling. NOTES “Pre” refers to assessments at baseline (November 2014 to February 2015). “Post” refers to assessments at endline (February 15 to July 10, 2016). The horizontal lines below each box refer to the local minimum; those above each box refer to the local maximum. The boxes depict the first quartile (bottom), median (middle), and third quartile (top) for each score. The y axis represents a numerical score that indicates participants’ knowledge of antenatal care, birth planning, and postnatal care (PNC) using a Knowledge-Attitudes-Practice (KAP) survey. A higher number indicates a better score. Birth planning is the use of contraception one year after delivery.

Internal Quality Improvement Measures

We assessed a number of quality improvement measures. These include institutional birth rate, antenatal care visits (percentage of pregnant women in a catchment area who completed four antenatal visits prior to delivery), and ultrasound usage (percentage of pregnant women in a catchment area who received ultrasound by their expected date of delivery). Measures were tracked in fourteen village clusters over a period of ten months (Exhibit 4). Results were relatively consistent across time for antenatal care and institutional birth rate, though ultrasound usage varied more widely. We summarize the successes and challenges in the Appendix.¹⁷

SOURCE Authors’ analysis of internal (electronic health record–based) quality improvement data from care delivery and follow-up visits. NOTES For antenatal care (ANC), the exhibit displays pregnant women in a catchment area who completed four antenatal visits prior to delivery as a percentage of the total number of pregnant women in a catchment area giving birth. Institutional birth rate is women giving birth in a health care facility in a catchment area with a health care professional present as a percentage of the total number of births reported in a catchment area. Receipt of ultrasound is women in a catchment area who received ultrasound by month 8 or 9 of pregnancy as a percentage of the total number of pregnant women who had an expected date of delivery in the given month.

Cost

The annual cost per capita of delivery of reproductive, maternal, newborn, and child health services is estimated US$3.40 per person in the population and US$185 per pregnant woman in the prior twelve months (Exhibit 5). Costs were split evenly between the facility and community health worker levels. Costs varied by delivery type: spontaneous vaginal (51 percent), complicated (60 percent), cesarean section (77 percent), and out-of-hospital (38 percent). The facility cost for delivery outside Bayalpata Hospital was US$58, largely because personnel costs were much lower than in Bayalpata Hospital deliveries. The use of electronic health records at Bayalpata Hospital also contributed to the cost difference.

Discussion

As health care systems expand their focus to include efficiency and quality, innovative approaches are required. In reproductive, maternal, newborn, and child health, studies have begun to describe new approaches to measuring, tracking, and improving quality.^18–20 In this article we have assessed a public-private partnership in a resource-limited setting that aims to improve quality and outcomes using accountable care principles. We found that the partnership between the Government of Nepal and the private nonprofit Possible was associated with improvements in process measures of population health linked to maternal and neonatal mortality. The rates of these measures at the end of the study show an impressive improvement over national statistics for both antenatal care completion (90 percent at the end of the study versus 50 percent nationally) and institutional births (92 percent at study end versus 37 percent nationally). The endline rates were higher in part because elements of the public-private partnership have been in operation for eight years. This study’s results provide encouraging evidence that the application of an accountable care framework in resource-limited areas can lead to improved access as well as more efficient and higher-quality care, and thus better outcomes.

Despite the progress achieved by the Possible partnership, opportunities remain to improve care. Staff members are still learning to incorporate and respond to performance metrics in their work routines. Limitations in skill sets, such as computer literacy, slow Possible’s efforts to effectively use the available technology. Providers’ lack of familiarity with the public-private partnership model may also undercut Possible’s ability to expand to new providers and areas.

Enabling Factors

Possible’s experience can inform accountable care implementation, particularly in rural or remote areas with limited resources. Here we discuss enabling factors that could support the expansion of these reforms to new providers and settings.

Teams Linked To The Community

The first enabling factor was the use of teams with community and social linkages. Possible’s model emphasizes preventive care to improve outcomes, which departs from the reactive-based health care practice and cultural mindset that are traditionally present in many settings. To succeed with this changed emphasis, Possible relied on community health workers to engage and educate patients. Since these workers are selected from the villages in which they serve, they tend to enjoy greater trust and understanding by the community. Using these health workers, Possible conducted home visits and built referral relationships with local clinics, expanding women’s access to a network of primary and secondary physicians. Health workers also educated patients to become effective managers of their own health by, for instance, monitoring their symptoms and proactively engaging with the health care system earlier to avoid complications.

Incremental Approach

The second enabling factor was Possible’s incremental approach to implementing and expanding accountable care. Wholesale care and payment reforms are capital and time intensive. Alternatively, reconfiguring services gradually enables organizations to innovate amid the resource constraints in low-income settings. For instance, Possible expanded its capabilities over a period of several years. Instead of deploying a complex information-telecommunications infrastructure, Possible developed its own open-source EHR system and increased that system’s functionality over time. Possible also created multidisciplinary teams by iteration, supplementing partnerships over time to include specialists such as mental health providers, depending on the needs of the system and community. Additionally, Possible augmented the scope of its public-private partnership contract over time, including the quality and importance of performance measures.

Generating Early Evidence

A third and related enabling factor was the generation of evidence early on to guide and refine reforms. Evidence on the effectiveness of financial and care delivery reforms is needed to scale or refine innovative approaches. While more resources for data collection (for example, in matched control populations) and analysis are desirable, in the absence of such resources it is important to use performance measurement as a means to generate evidence on program impact. Such timely studies can help organizations identify and scale positive outcomes.²¹

Possible is establishing the infrastructure to conduct comparative trials on different variations of its integrated care components. It aims to ask targeted questions within the routine course of clinical care about different elements of the system and thus have a stronger evidence foundation from which to guide further care reforms. These evaluations of low-cost innovations will be both rapid and affordable and could therefore serve higher-income settings as well, where such innovations may be too costly to implement and test. As Possible has exhibited, collecting health care data is itself an intervention that leads to broader systems improvement. The integrated, household-to-facility EHR that Possible developed represents the kind of architecture necessary to conduct such trials. The affordable creation of this resource, within the constraints of a rural region in one of the world’s most impoverished countries, suggests that there is possibility for developing similar health care data systems throughout the world.

The Role Of Accountable Care Payment Reform

The enabling factors discussed above are, in turn, sustained and reinforced by accountable care payment reforms. Implementing payments that are population focused and tied to performance, as Possible did, provides a system-level incentive to meet a specific standard of care. It also allowed Possible to orient resources toward value-based activities that legacy payment models may neglect or stymie. For instance, Possible used its partnership contract with the Nepal Ministry of Health to hire nonmedical personnel, such as community health workers. The contract spurred the evolution of Possible’s capabilities, services, and ability to generate data in lockstep with a payment model based on performance. This experience demonstrates that it is feasible to link care delivery models with greater accountability, even in resource-constrained settings with limited measurement capabilities.

Conclusion

In this public-private partnership in Nepal, we found that an accountable care systems approach can lead to improvement in population-level indicators linked to maternal and infant mortality reduction: the receipt of four antenatal care visits, the rate of institutional births, and postpartum contraceptive prevalence. Despite the limitations and challenges involved with the partnership, we found and developed empirical evidence on accountable care systems and other approaches to integrated care delivery in a resource-limited setting where such evidence is severely lacking. An important next step will be to develop a shared set of metrics and evaluation approaches to compare accountable care systems across contexts globally. This will help national health systems identify policies, frameworks, and financing mechanisms to enable health care policy makers, payers, and providers to be more responsive and accountable to the local populations that they aim to serve.

ACKNOWLEDGMENTS

The authors acknowledge the Commonwealth Fund for funding a portion of the research reported in this article. Duncan Maru received support from the National Institutes of Health, Award No. DP5OD019894. Sheela Maru received support from the Mary Horrigan Connors Center for Women’s Health and Gender Biology at Brigham and Women’s Hospital. Krishna Udayakumar has received the following research grants through Duke University or grants/membership support for Innovations in Healthcare, a Duke University–affiliated nonprofit for which Udayakumar serves as executive director: Pfizer Foundation, USAID, the Commonwealth Fund, Medtronic, Novartis Foundation, UNFPA, World Innovation Summit for Health/Qatar Foundation, Health Foundation (UK), Gates Foundation, and Cardinal Health. Isha Nirola and Prajwol Nepal both received compensation from Possible. Indira Basnett and Pushpa Chaudary both received compensation from the Nepal Ministry of Health. Mark McClellan received compensation from the following Duke Sponsored Research Funders: MITRE Corporation, Health Foundation (UK), the Commonwealth Fund, the Food and Drug Administration, Laura and John Arnold Foundation, Novartis, Allergan, Amgen, Bluebird Bio, Spark Therapeutics, Editas, Pfizer, Medtronic, Edwards Life Sciences, Boston Scientific, CEOi, Lilly USA, National Pharmaceutical Council, and Duke-NUS.

NOTES

• 1 Ki-Moon B. Global strategy for women’s and children’s health [Internet]. New York (NY): United Nations; 2010 [cited 2017 Aug 30]. Available from: http://www.everywomaneverychild.org/images/content/files/global_strategy/full/20100914_gswch_en.pdf Google Scholar
• 2 World Health Organization. Maternal mortality: fact sheet [Internet]. Geneva: WHO; 2016 [cited 2017 Aug 30]. Available from: http://www.who.int/mediacentre/factsheets/fs348/en/ Google Scholar
• 3 You D, Hug L, Ejdemyr S, Idele P, Hogan D, Mathers Cet al. Global, regional, and national levels and trends in under-5 mortality between 1990 and 2015, with scenario-based projections to 2030: a systematic analysis by the UN Inter-agency Group for Child Mortality Estimation. Lancet. 2015;386(10010):2275–86. Crossref, Medline, Google Scholar
• 4 El Arifeen S, Blum LS, Hoque DM, Chowdhury EK, Khan R, Black REet al. Integrated Management of Childhood Illness (IMCI) in Bangladesh: early findings from a cluster-randomised study. Lancet. 2004;364(9445):1595–602. Crossref, Medline, Google Scholar
• 5 Arifeen SE, Hoque DM, Akter T, Rahman M, Hoque ME, Begum Ket al. Effect of the Integrated Management of Childhood Illness strategy on childhood mortality and nutrition in a rural area in Bangladesh: a cluster randomised trial. Lancet. 2009;374(9687):393–403. Crossref, Medline, Google Scholar
• 6 Oluwole D, Mason E, Costello A. Management of childhood illness in Africa. Early evaluations show promising results. BMJ. 2000;320(7235):594–5. Crossref, Medline, Google Scholar
• 7 Conde-Agudelo A, Rosas-Bermúdez A, Kafury-Goeta AC. Birth spacing and risk of adverse perinatal outcomes: a meta-analysis. JAMA. 2006;295(15):1809–23. Crossref, Medline, Google Scholar
• 8 Requejo JH, Bryce J, Barros AJ, Berman P, Bhutta Z, Chopra Met al. Countdown to 2015 and beyond: fulfilling the health agenda for women and children. Lancet. 2015;385(9966):466–76. Crossref, Medline, Google Scholar
• 9 Kuruvilla S, Schweitzer J, Bishai D, Chowdhury S, Caramani D, Frost Let al. Success factors for reducing maternal and child mortality. Bull World Health Organ. 2014;92(7):533–44B. Crossref, Medline, Google Scholar
• 10 McClellan M, Kent J, Beales SJ, Cohen SIA, Macdonnell M, Thoumi Aet al. Accountable care around the world: a framework to guide reform strategies. Health Aff (Millwood). 2014;33(9):1507–15. Go to the article, Google Scholar
• 11 McClellan M, Thoumi A, Udayakumar K, Patel H, Abou Samra AB. Implementing accountable care to achieve better health at a lower cost [Internet]. Doha (Qatar): World Innovation Summit for Health; 2016 [cited 2017 Aug 30]. Available from: http://www.wish-qatar.org/app/media/download/2738 Google Scholar
• 12 McClellan M, Udayakumar K, Gonzalez-Smith J, Kadakia K, Thoumi A, Maru Det al. Case study: Possible, Nepal. Accountable care in practice: global perspectives [Internet]. Durham (NC): Duke–Margolis Center for Health Policy; 2017 [cited 2017 Oct 5]. Available from: https://healthpolicy.duke.edu/sites/default/files/atoms/files/nepal_25jan2017.pdf Google Scholar
• 13 World Health Organization. Global health expenditure database: 2014 [Internet]. Geneva: WHO; 2014 [cited 2017 Aug 30]. Available from: http://apps.who.int/nha/database Google Scholar
• 14 World Bank. GDP per capita (current US$), World Bank national accounts data, and OECD national accounts data files [Internet]. Washington (DC): World Bank; [cited 2017 Oct 5]. Available from: http://data.worldbank.org/indicator/NY.GDP.PCAP.CD Google Scholar
• 15 World Bank. Poverty and equity: regional dashboard: South Asia [Internet]. Washington (DC): World Bank; [cited 2017 Aug 30]. Available from: http://povertydata.worldbank.org/poverty/region/SAS Google Scholar
• 16 ClinicalTrials.gov. Group antenatal care: the power of peers for increasing skilled birth attendance in Achham, Nepal [Internet]. Bethesda (MD): National Institutes of Health; 2004 Dec 31 [cited 2017 Aug 30]. Available from: https://clinicaltrials.gov/ct2/show/NCT02330887?term=NCT02330887&rank= Google Scholar
• 17 To access the Appendix, click on the Details tab of the article online.
• 18 Sharma J, Leslie HH, Kundu F, Kruk ME. Poor quality for poor women? Inequities in the quality of antenatal and delivery care in Kenya. PLoS One. 2017;12(1):e0171236. Crossref, Medline, Google Scholar
• 19 Larson E, Vail D, Mbaruku GM, Mbatia R, Kruk ME. Beyond utilization: measuring effective coverage of obstetric care along the quality cascade. Int J Qual Health Care. 2017;29(1):104–10. Medline, Google Scholar
• 20 Kruk ME, Leslie HH, Verguet S, Mbaruku GM, Adanu RMK, Langer A. Quality of basic maternal care functions in health facilities of five African countries: an analysis of national health system surveys. Lancet Glob Health. 2016;4(11):e845–55. Crossref, Medline, Google Scholar
• 21 Farmer P. Clinical trials and global health equity. Lancet Global Health Blog [blog on the Internet]. 2013 Jul 8 [cited 2017 Aug 30]. Available from: http://globalhealth.thelancet.com/2013/07/08/clinical-trials-and-global-health-equity Google Scholar