Evidence-based training and mentorship combined with enhanced outcomes surveillance to address the leading causes of neonatal mortality at the district hospital level in Ghana

Authors


Abstract

Objective

To evaluate the impact of a district hospital intervention focused on enhancing healthcare provider capacity to address leading causes of neonatal death: birth asphyxia, infection and prematurity.

Methods

The neonatal quality improvement initiative was launched at two intervention referral district hospitals in Ghana. Local Health and Demographic Surveillance Systems were enlisted to enhance recording of neonatal and infant deaths in the community and at the facility. After baseline site assessments, a team of local paediatric experts conducted three clinical trainings on-site at each intervention hospital. Assessments were conducted to evaluate participant knowledge before and after participation in training modules. Monthly mentorship visits provided additional training to support the adoption of essential early neonatal care practices.

Results

In the first year of implementation, the initiative provided focused clinical training to 278 participants. A comparison of pre- and post-training test results demonstrates significant improvement in provider knowledge (73% vs. 89% correct, < 0.001), with even greater improvement among trainees receiving recurrent refresher training (86% vs. 94% correct, < 0.001). Participant feedback following training revealed enthusiasm about the programme and improved confidence.

Conclusions

Locally led initiatives that invest directly in healthcare provider education and health systems strengthening represent a promising avenue for reducing neonatal morbidity and mortality. The NQI initiative demonstrates the positive impact of a district hospital intervention that combines on-site training, mentorship and enhanced demographic surveillance.

Abstract

Objectif

Evaluer l'impact d'une intervention au niveau de l'hôpital de district portant sur le renforcement de la capacité des prestataires de soins de santé, pour faire face aux principales causes des décès néonataux: asphyxie à la naissance, infection et prématurité.

Méthodes

L'initiative d'Amélioration de la Qualité Néonatale(AQN) a été lancée dans deux hôpitaux d'orientation de district recevant l'intervention au Ghana. Les Systèmes de Surveillance Démographiques et de Santé locaux ont été mobilisés pour améliorer l'enregistrement des décès néonataux et infantiles dans la communauté et dans les établissements. Après les évaluations des données de référence dans les sites, une équipe d'experts pédiatriques locaux a effectué trois stages cliniques sur place dans chaque hôpital d'intervention. Les évaluations ont été réalisées afin d’évaluer les connaissances des participants avant et après la participation à des modules de formation. Des visites mensuelles de mentorat ont offert une formation supplémentaire pour soutenir l'adoption de pratiques de soins néonataux précoces essentiels.

Résultats

Dans la première année d'implémentation, l'initiative a permis une formation clinique focalisée à 278 participants. Une comparaison des résultats de formation pré- et post- tests a démontré une amélioration significative des connaissances des prestataires (73% versus 89% de réponses correctes, p < 0.001), avec même une plus grande amélioration chez les stagiaires recevant une formation de perfectionnement périodique (86% versus 94% de réponses correctes, p < 0.001). Les perceptions des participants après la formation ont révélé un enthousiasme au sujet du programme et une confiance en soi améliorée.

Conclusions

Les initiatives menées localement qui investissent directement dans l’éducation des prestataires de soins de santé et dans le renforcement des systèmes de santé, représentent une voie prometteuse pour réduire la morbidité et la mortalité néonatale. L'initiative de l’AQN démontre l'impact positif d'une intervention au niveau de l'hôpital de district qui combine une formation sur place, le mentorat et le renforcement de la surveillance démographique.

Abstract

Objetivo

Evaluar el impacto de una intervención a nivel de hospitales distritales, diseñada para mejorar la capacidad del personal sanitario de abordar las principales causas de muerte neonatal: asfixia al nacer, infección y prematuridad.

Métodos

La iniciativa de Mejora de la Calidad Neonatal (MCN) se lanzó en dos hospitales distritales de referencia en Ghana. Se utilizaron los sistemas de vigilancia sanitaria y demográfica para mejorar los registros de muertes neonatales e infantiles en la comunidad y en el centro hospitalario. Después de una evaluación de base, un equipo de expertos en pediatría realizó tres entrenamientos in-situ en cada uno de los hospitales que participaron en la intervención. Se realizaron evaluaciones para determinar el conocimiento de los participantes antes y después de los módulos de entrenamiento. Se realizaron visitas mensuales de tutoría con el fin de proveer entrenamiento adicional y apoyo en la adopción de prácticas esenciales en cuidados neonatales.

Resultados

Durante el primer año de implementación, la iniciativa ofreció entrenamiento clínico dirigido a 278 participantes. La comparación de los resultados de las pruebas pre- y post-entrenamiento demuestran una mejoría significativa en los conocimientos de los proveedores (73% vs. 89% respuestas correctas, < 0.001), con una mejoría aún mayor entre los participantes que recibían sesiones de refuerzo recurrentes (86% vs. 94% respuestas correctas, < 0.001). La retroalimentación de los participantes después del entrenamiento reveló entusiasmo por el programa y una mejora en su nivel de autoconfianza.

Conclusiones

Las iniciativas locales que invierten directamente en la educación de los proveedores sanitarios y fortalecen los sistemas de salud representan un camino prometedor para reducir la mortalidad y la morbilidad en neonatos. La iniciativa MCN demuestra el impacto positivo de una intervención a nivel de hospitales distritales que combina entrenamiento in-situ, la tutoría y mejoras en la vigilancia demográfica.

Introduction

Globally, more than 3 million infants die in the neonatal period, within the first month of life; nearly all of these deaths occur in poor countries (Rajaratnam et al. 2010; Oestergaard et al. 2011). While under-five mortality has been decreasing steadily, the proportion of child deaths occurring in the neonatal period continues to increase, comprising 40% of under-five deaths (Lawn et al. 2006; Lawn et al. 2010; You et al. 2011 UNICEF 2012). Nearly three quarters of all neonatal deaths take place in the first week of life as a result of common perinatal causes including prematurity (35%), intrapartum events (birth asphyxia) (23%) and infections (23%) (Engmann et al. 2012; Liu et al. 2012). In Ghana, demographic data reveal a neonatal mortality rate of 28 deaths per 1000 live births, with an under-five mortality rate of 74 per 1000 live births (Rajaratnam et al. 2010; You et al. 2011. Despite concerted efforts by the Ministry of Health and Ghana Health Service, including a programme to provide free health care to mothers, newborns and infants, the incidence of neonatal death remains unchanged (Ghana Statistical Service (GSS), Noguchi Memorial Institute for Medical Research (NMIMR), & ORC Macro 2004; Johnson et al. 2005; Pond et al. 2005; Ghana Statistical Service (GSS), Ghana Health Service (GHS), & ICF Macro 2009 Ministry of Health, Government of Ghana, & United Nations Country Team in the Republic of Ghana 2011). Although Ghana has focused on reducing maternal and under-five mortality in recent years, neonatal survival remains a pressing challenge.

Recent reports indicate that neonatal health initiatives have been drastically underfunded at the global scale, accounting for <6% of official development assistance earmarked for neonatal health interventions (Lawn et al. 2012). Meanwhile, it is estimated that universal implementation of relatively low-cost interventions such as Kangaroo Mother Care, prevention of asphyxia and treatment of infection can reduce overall newborn mortality by 75% (Darmstadt et al. 2005; Lee et al. 2011). Thus, a high-impact strategy should focus on the early post-natal period, when newborns, particularly those living in rural and impoverished communities, are most vulnerable. Sustainable and locally led initiatives that address neonatal health are a critical component to reducing morbidity and mortality in children under-five. Monitoring of neonatal outcomes in resource-poor settings represents an additional strategy for improving child health, as it can contribute to the evidence base and direct national policies (Baiden et al. 2006).

The Ghanaian-led Neonatal Quality Improvement (NQI) initiative is a collaborative programme that was launched by the Ghana Health Service, WHO and the Johns Hopkins University. The programme combines evidence-based teaching about early neonatal care, focusing on management of sepsis, asphyxia and preterm birth with ongoing on-site mentorship, review of essential medicines and equipment, and enhanced outcomes monitoring using the existing demographic surveillance systems. The overall goal of the NQI initiative is to demonstrate the impact of a locally led community and district hospital initiative that combines cost-effective and evidence-based practices to target the leading causes of neonatal death.

Background

Ghana's public healthcare system is operated through a National Health Insurance Scheme designed to ensure access to basic health care regardless of socio-economic status. While recipients are charged a nominal fee, this health coverage is largely funded through government investment (Agyepong & Adjei 2008). A national free maternal healthcare programme was also introduced in 2008, providing pre- and post-natal health care to all mothers free of charge. This policy was implemented to help accelerate the Millennium Development Goals for maternal and child health (MDG 4-5), enhancing services provided to mothers and their newborn babies (National Health Insurance Authority 2011).

Despite these efforts, improving neonatal outcomes remains a challenge in Ghana, particularly in rural settings (Ministry of Health, Government of Ghana, & United Nations Country Team in the Republic of Ghana 2011). Addressing neonatal mortality comprehensively requires interventions that cover all levels of care, from community health centres and primary healthcare clinics to district hospitals and tertiary care centres (Blencowe & Cousens 2013). While there are presently interventions underway to improve care at the community and primary healthcare levels (Kirkwood et al. 2010, 2013; Project Fives Alive! Team 2011), neonatal care at the district hospital level remains under-prioritised. A recent baseline assessment conducted by Ghana Health Service at three of the country's district hospitals using a WHO assessment tool (WHO 2006; GHS & WHO 2010) revealed inadequacies in the medicines, supplies, equipment and case management practices for sick neonates and children. Around the same time, Ghana Health Service was working in collaboration with the regional WHO country office to launch a revised version of the Pocketbook of Hospital Care for Children, a standardised WHO volume which outlines guidelines for management of common childhood illnesses in resource-limited settings (WHO 2005).

To facilitate the implementation of neonatal guidelines, a technical consulting team from Johns Hopkins partnered with the Ghana Health Service and the WHO to develop a scalable package of interventions designed to improve neonatal care at the district hospital level drawing from the WHO Pocket Book and from other universally accepted resources and guidelines. The resulting NQI programme combined enhanced demographic surveillance with a district hospital mentorship and training initiative. Specifically, the programme focused on addressing and managing the leading causes of neonatal mortality, including prematurity, asphyxia and infection.

Methods

A steering committee was established to oversee the development and implementation of the NQI initiative. The committee was led by Dr. Cynthia Bannerman, Deputy Director of the Institutional Care Division of Ghana Health Service, with representation from the WHO and Johns Hopkins. A supporting committee of local paediatric experts and community stakeholders was also identified to provide guidance and oversight to the team.

Intervention sites were selected based on the recommendation of the Ghana Health Service. Sites included Navrongo War Memorial District Hospital in Ghana's Upper East region and Kintampo District Hospital in the Brong Ahafo region (Figure 1). Site selection was guided by Ghana Health Service and informed by the presence of a first-referral district hospital and a Health and Demographic Surveillance System (HDSS) equipped to record community and hospital-based maternal and neonatal outcomes.

Figure 1.

Ghanaian neonatal quality improvement intervention and control sites.

Navrongo War Memorial District Hospital is a first-referral district hospital serving a population of approximately 156 000 in Kassena-Nankana district (Ghana Statistical Service 2008). The other healthcare facilities in the region are 23 Community Based Health Planning and Service compounds, seven community health centres, three rural clinics and one maternity home (Ghana Health Service Council 2012). Kintampo District Hospital serves Kintampo North and South districts, which have a catchment area of approximately 140 000 residents (Ghana Statistical Service 2008). The population from this district is also served by seven CHPS compounds, six community health centres, four rural clinics, two privately managed maternity homes and one primary hospital (Ghana Health Service Council 2012).

Both regions have a reported fertility rate of 4.1 (Ghana Health Service 2010). While skilled delivery coverage is reportedly higher than the national average (52.23% in 2011) in these two regions, with 63.2% coverage in the Brong Ahafo region and 67.3% coverage in the Upper East region, addressing neonatal and infant mortality remains difficult (Ghana Health Service 2011).

Implementation

In November 2011, district health and demographic surveillance systems in the Navrongo and Kintampo regions began collecting enhanced demographic data from district hospital sites and the community to record all births, stillbirths, neonatal and infant deaths. A baseline district hospital assessment had previously been conducted to review the availability of hospital guidelines, healthcare provider practices and case management for children under 5 years of age at each site. The assessment also included an appraisal of essential paediatric medicines and equipment. After this baseline evaluation and 2 months of preliminary demographic monitoring at all sites, the NQI initiative was officially launched in January 2012 with a series of on-site clinical trainings. A longitudinal approach was adopted, combining frequent on-site clinical education with monthly day-long mentorship and clinical refresher visits (Figure 2). During mentorship visits, trainers reviewed case management and observed patient care, providing coaching and feedback to trained participants and other healthcare providers at the hospital. The site visits also provided an opportunity to review the availability of essential equipment and medicines for newborn care. NQI trainers and regional supervisors worked with administrative staff, including the head of pharmacy, the hospital matron, as well as the medical superintendent at each hospital to address supply chain issues and management.

Figure 2.

Neonatal quality improvement flow chart.

Data collection

A data collection plan was established by senior representatives from the Health Research Centers that run Ghana's HDSS sites. HDSS centres fall under the direction of the Ghana Health Service and are part of the INDEPTH network, an international network of sentinel demographic data collection systems that record population-based longitudinal demographic information and provide a platform for health research. Each HDSS employs a group of trained fieldworkers who conduct home visits every 4 months to collect regional data for a district population of approximately 100 000–150 000. Additionally, a network of trained volunteers called community key informants (CKIs) reports key events occurring in the community, such as pregnancies, births and deaths. For the purposes of the NQI programme, CKIs and field workers worked together to record new pregnancies, births and deaths in real time. CKIs were contacted by field workers every 2 weeks to identify new events. Subsequently, HDSS field supervisors verified each of these events within a week of the reporting. As part of the regular HDSS system, CKIs are compensated with a small token for each event that is accurately reported. Once verified, these data were submitted to the HDSS for inclusion in monthly and annual demographic reports. A dedicated HDSS fieldworker was also assigned to each intervention district hospital to record a daily log of admissions, births, deliveries and deaths. The purpose of community and facility-based recording was to ensure precise and harmonised data collection.

NQI training

A team consisting of Ghanaian paediatricians and nurses, along with clinical consultants from Johns Hopkins University, conducted clinical training sessions at both Navrongo and Kintampo District Hospitals. The objective of the NQI clinical training was to enhance district hospital workforce capacity through neonatal care training and frequent refresher sessions. Training was offered bi-annually to medical officers, nurses, midwives, medical assistants and all auxiliary staff working with neonates and infants, including staff working in the Labor and Delivery Ward, Maternity Ward, Neonatal Unit, Children's Ward, Emergency Ward and Outpatient Department. Additionally, healthcare providers from neighbouring health facilities were invited to attend the trainings. Each 2-day session was repeated twice at the intervention sites to maximise participation without disrupting patient care.

Members of the NQI training team were selected through the Ghana Health Service and included a core team of paediatricians with a clinical focus in neonatal care, and experience in paediatric training and teaching within their institutions and hospitals. The majority were also experienced national trainers for newborn care. A regional clinical supervisor was also assigned to each intervention site to provide oversight, support and mentorship, through monthly clinical visits and feedback. The core team of paediatricians and regional clinical supervisors participated in all on-site trainings and shared responsibilities for conducting monthly mentorship visits. During training visits, this team was also accompanied by a consulting neonatologist from Johns Hopkins as well as a support team of nurses with extensive paediatric experience, trained medical students, neonatology trainees and support staff from the three partner organisations, Ghana Health Service, WHO Ghana and Johns Hopkins. In general, the average size of each training team ranged between 8 and 10 members.

Training sessions lasted 2 days and combined didactic teaching with interactive sessions, hands-on skills practice and demonstration of competency. The training curriculum incorporated a review of treatment guidelines, hands-on patient care, appropriate administration of medicines, and proper use and maintenance of paediatric equipment and supplies. The curriculum was developed using evidence-based materials and guidelines produced by the World Health Organization, the American Academy of Pediatrics (AAP Helping Babies Breathe®) and UNICEF (WHO 2003, 2005; WHO & UNICEF 2008; Singhal et al. 2012). Teaching modules incorporated the AAP's Helping Babies Breathe® curriculum, the WHO's Kangaroo Mother Care, and guidelines on neonatal resuscitation, management of perinatal asphyxia, detection of danger signs and serious newborn bacterial infections, as well as supportive care for the sick neonate.

Participants completed a multiple-choice questionnaire at the beginning of each training session to assess their baseline knowledge of the subject matter. Throughout the session, trainers engaged trainees to ensure active participation. The training curriculum began with didactic sessions, incorporating interactive sessions as the trainees became more comfortable with the course content. All sessions concluded with hands-on skills training conducted in small supervised groups. While in small groups, trainers had the opportunity to directly observe the skills of each participant, monitor progress over the 2-day session and conduct individual evaluations to ensure that each healthcare provider could apply the resuscitation techniques presented in the Helping Babies Breathe curriculum, as well as demonstrate effective Kangaroo Mother Care practice. At the end of each 2-day session, participants completed the multiple-choice questionnaire to assess their understanding and retention of the subject matter. This was followed by a review of all questions and answers by trainers, with an opportunity for participants to ask questions and clarify items that were not understood.

A monthly clinical and site assessment was conducted by regional clinical supervisors and trainers between biannual clinical trainings. The purpose of these visits was to provide mentorship to healthcare providers on-site and to assess the adoption of essential neonatal care practices.

Results

Neonatal quality improvement training sessions took place in January 2012, July 2012 and January 2013. Trainers conducted two separate sessions during each of their weeklong site visits to allow maximal participation by healthcare providers without disrupting patient care. A total of 185 providers were trained over the course of the initiative. Of these, 93 participated in two or more of the NQI training sessions. The majority were nurses and midwives working on-site or in neighbouring health clinics. Additional participants included medical officers, anaesthetists, pharmacists, medical assistants and auxiliary staff (Figure 3). Male-to-female ratio was 0.395, indicating that a greater proportion of women participated in the training.

Figure 3.

Profession of NQI clinical training participants.

NQI clinical training evaluation

Knowledge assessments were performed before and after each training session to track improvement in fund of knowledge and retention of information and to identify the impact of recurrent training. Results of multiple-choice knowledge tests were shared with the trainees at the end of each session to highlight score trends. In addition, there was a session to review all test items, reveal correct responses and address any questions from the trainees.

Significant improvement in knowledge assessment results was observed when comparing scores on tests administered before and after the clinical training at both sites (Figure 4). During the initial January 2012 training, the mean score rose from 72.3% to 88.4% (< 0.001), representing a 16.1-point increase across both sites. Comparable results were observed at the second and third NQI trainings held in July 2012 and January 2013. Mean scores increased from 79.9% to 91.5% (11.6 points, < 0.001) and from 79.4% to 91.8% (12.4 points, < 0.001), respectively. The relatively higher pre- and post-training test scores achieved during the second and third visits reflect the knowledge retention of the 93 participants who repeated the training. Repeat participants consistently scored higher than first-time participants, on pre-training tests (85.8% vs. 73.4% < 0.001) and on post-training tests (93.9% vs. 88.9%, < 0.001). Not only does this suggest that participation in the NQI training is associated with long-term knowledge retention, but also that trainees benefit from repeated participation. Overall, pre- and post-training test results for all trainings demonstrated a significant improvement in provider knowledge with average scores increasing from 73% to 89% (< 0.001).

Figure 4.

NQI knowledge assessments.

Participant feedback

Neonatal quality improvement clinical training feedback was collected from participants using anonymous evaluation forms adapted from a Ghana Health Service Continuing Professional Development form. Feedback response from participants was 100%. Taken together, comments and suggestions were largely positive. Respondents indicated that the NQI training programme had reinforced their knowledge of essential newborn care practices, while also boosting confidence and encouraging peer mentorship in the hospital wards. Over 100 participants requested additional and more frequent training sessions. Additionally, most respondents recommended that the NQI training programme should be scaled up to the surrounding regions, reaching out to Community Based Health Planning and Service compounds, health training and education facilities, and other health centres or clinics in more remote areas).

Neonatal and infant outcomes

Monthly community and district hospital demographic reports were prepared by each HDSS site, providing data on births, deaths, stillbirths and hospital admissions (Figure 5. Reporting occurred in real time, with HDSS systems conveying monthly data by the 10th day of the following month. Given the brief data collection period (November 2011–October 2012) and the early stage of implementation of the NQI programme, the demographic data provide a snapshot of neonatal and infant outcomes in the Navrongo and Kintampo districts. In total, 1764 live births were recorded at Navrongo District Hospital. At Kintampo District Hospital, 1426 live births were reported. At Navrongo District Hospital, there were 179 neonatal and 379 infant admissions reported. Additionally, field workers recorded 35 neonatal deaths, 60 infant deaths and 25 stillbirths. At Kintampo District Hospital, there were 199 neonatal and 395 infant admissions reported. In total, 25 neonatal deaths, 9 infant deaths and 65 stillbirths were recorded by fieldworkers. These data were collected for the period between November 2011 and October 2012. Data collection was limited to this period of collection due to financial restrictions of the project.

Figure 5.

Neonatal deaths and stillbirths by location.

While the available comparison data were limited to neonatal mortality rates for the entire Upper East and Brong Ahafo regions, we observed lower recorded neonatal mortality at both intervention sites. Data collected from Navrongo indicated a neonatal mortality rate of 14 per 1000 live births; the reported neonatal mortality rate is 17 for the entire Upper East region Ghana Statistical Service (GSS), Ghana Health Service (GHS), & ICF Macro 2009. Likewise, the neonatal mortality rate observed at the Kintampo site was 24 per 1000 live births, while the neonatal mortality rate for the entire Brong Ahafo region is reported to be 27. Although these data seem to indicate a lower neonatal mortality rate at the two intervention sites, a long-range study of the demographic data would be required to observe the longer-term impact of the NQI intervention on neonatal mortality outcomes.

Discussion

The NQI programme was developed by the Ghana Health Service through an unprecedented partnership between WHO and Johns Hopkins University. The programme adopts a multipronged approach to addressing the quality of neonatal care at the district hospital level by focusing on building healthcare provider capacity. A novel feature of the programme is that it provides training opportunities for all staff providing care and support to newborns and infants. By addressing care on-site at the district hospital level, the programme recognises the critical importance that first-referral facilities play in the continuum of neonatal and infant care.

Healthcare providers participating in the NQI training demonstrated significant improvement in their post-training knowledge evaluation scores. Furthermore, participants described improved levels of confidence in terms of their essential newborn care skills and competence. These findings are supported in the literature, as clinical training programmes are associated with improved provider knowledge over the short term (Baker 2009; Forsetlund et al. 2009; Bookman et al. 2010; Anderson & Warren 2011; Singhal et al. 2012). Continued cognitive improvement among repeat trainees also suggested that repeat trainings further reinforced learning. Not only does this finding support the importance of recurrent refresher training but it also has implications for future implementation of on-site clinical training.

Participants were also required to apply their essential neonatal care knowledge in a simulated clinical setting during the trainings. The NQI training team evaluated each healthcare provider on their ability to demonstrate appropriate resuscitation skills using infant and neonatal simulator manikins. In order to receive a certificate of completion, each participant was required to successfully demonstrate an appropriate grasp of resuscitation skills, based upon the Helping Babies Breathe checklist. To further reinforce lessons from the training, on-site clinical mentorship was provided on a monthly basis at each intervention site. Members of the training team conducted monthly full-day visits to each intervention district hospital to provide support and guidance to NQI training participants. Mentorship visits also consisted of direct observation of patient care as well as coaching and feedback sessions with providers. Additionally, mentors conducted ad hoc teaching sessions particularly with regards to resuscitation and also assisted with equipment and logistical challenges. This direct mentorship was followed up with consistent monitoring by regional and national NQI supervisors who helped provide evaluation and feedback to both the practitioners and the district hospital administration.

One limitation to this study was a lack of direct clinical observation pre- and post-intervention. Although baseline assessments and monthly mentorship visits included chart review and observation of case management, there was no formal tracking of clinical parameters beyond recording neonatal and infant mortality. Future studies may consider incorporating pre- and post-intervention review of case-specific morbidity and mortality pertaining to intrapartum-related events such as asphyxia and sepsis.

An innovative and scalable approach

The strength of the Neonatal Quality Improvement initiative is that it combines clinical teaching and mentorship with outcomes monitoring within the context of a district hospital capacity strengthening initiative. While the NQI programme focused on district hospital interventions, it is anticipated that improved newborn case will also have a positive impact on the general population, improving overall infant outcomes in the longer term. Additionally, the strategy employed in the NQI programme can be cost-effectively replicated at other district hospitals and community health facilities where newborns receive care. It is estimated that training and mentorship as well as demographic surveillance incurred an average annual cost of $55 000 USD per site. Integrating the NQI programme as part of existing health system strengthening initiatives represents a relatively low-cost intervention that can have an important impact on improving the care of neonates and infants at the district hospital level. It is, however, important to note that a comprehensive scale-up of the NQI programme in Ghana or elsewhere would require consideration of regional factors, such as perinatal health-seeking behaviour, accessibility of care and birth attendance patterns. For instance, in Ghana 55% of births take place at a healthcare facility and the rest at home (WHO 2013). This means that effective scale-up of the NQI programme will need to adopt a community-based component, incorporating principles of neonatal quality improvement across the continuum of care, from household to hospital.

Conclusion

Sustainable initiatives that address newborn health at the district hospital level are a critical step to ensuring that neonatal outcomes are addressed across the continuum of care. The NQI programme was designed to strengthen capacity for essential newborn care at first-referral facilities in under-resourced settings. Specifically, the NQI initiative focused on addressing the leading causes of neonatal death by developing a multipronged district hospital approach that combines on-site clinical refresher training, essential medicines review, monitoring and feedback as well as enhanced health and demographic surveillance. Findings from the first year of implementation demonstrate increased provider knowledge and confidence as well as improved skills in simulated clinical settings. The NQI programme provides a low-cost and scalable model for improving essential newborn care and provider capacity at the district hospital level in under-resourced settings.

Acknowledgements

We would like to acknowledge the following people for their invaluable contribution to the Neonatal Quality Improvement initiative: Selina Dussey, Serwah Amoah, Ghana Health Service and Regional Health Authority representatives, District Hospital healthcare providers and staff from Navrongo, Kintampo and Dodowa, the network of researchers, fieldworkers and CKIs at Navrongo, Kintampo and Dodowa Health Research Centers and support staff at the WHO country office in Accra. We thank the WHO Essential Medicines and Pharmaceutical Policies Secretariat, which provided funding for this initiative through a subcontract with the Bill and Melinda Gates Foundation.

Ancillary