Table 1.

Characteristics of included studies.

Authors	Year	Country	Description of project/aim	Study design	Methods	Major findings	Strengths	Limitations	MMAT scorea
Ashwell and Barclay [28]	2009a	Papua New Guinea (PNG)	Women and Children's Health Project – aimed at increasing quality and coverage of rural health services to reduce maternal and infant mortality. Involved community development and health promotion.	Program evaluation	Mixed methods. Precede-Proceed model. Interviews and Focus Group Discussions. National, provincial and district level data.	Donor-defined objectives and contractual obligations limit project activity and outcomes. Increased interactions between community and maternal child health services. Healthcare workers, community health organisations and district/provincial health administrators not adequately engaged in planning process. Raising awareness of maternal child health issues is not enough, need to provide community leaders with tools/process to address issues. Healthcare workers expected to mobile community without support or consideration of capacity to do this. Insufficient technical advice and local leadership. Project ‘scope’ modified and expanded with inclusion of foreign technical advisor but ongoing lack of local involvement/leadership. Project lacked alignment with existing health system levels. Lack of involvement of National Department of Health from the outset therefore impact on national policies and programs was limited.	National, provincial and district level data included.	Model failed to sufficiently describe barriers to improving quality care. PNG-specific may not be generalisable to the Pacific region.	2
Ashwell and Barclay [49]	2009b	PNG	Women and Children's Health Project – aimed at increasing quality and coverage of rural health services to reduce maternal and infant mortality. Involved community development and health promotion.	Program evaluation	Mixed methods. Qualitative interviews, Focus Group Discussions, site visits, opportunistic observations of behaviours, health record audits.	In villages where volunteers and staff had been trained - new health knowledge led to changes in lifestyle practices, improved physical health and social and emotional wellbeing. Factors influencing success were motivated community members who acted as a catalyst for change, empowered leadership through governance structures, effective visual tools, village health volunteers linking community and rural health workers. Factors limiting success were poor understanding of community development, limited information sharing, ‘top down’ approach to community development and weak community leadership.	Large sample size, several different communities involved, follow up post project implementation sufficient to assess sustainability of project outcomes, broad spectrum of stakeholder perspectives included.	Limited quantitative data to support qualitative findings, unclear outcome measures, limited interviews with community members. PNG-specific may not be generalisable to the Pacific region.	2
Bettiol [39]	2004	PNG	Village Birth Attendant (VBA) program trains women in small, rural PNG villages to be village birth attendants.	Program evaluation	Qualitative. Semi-structured interviews with VBAs.	VBAs have many roles – maternal and child health patrols, antenatal checks and referrals, health promotion. VBAs were motivated to help mothers in community They faced difficulties managing obstetric complications i.e. breech, post-partum haemorrhage and logistical issues i.e. transport access.	Methods well described and outcome data provided regarding VBA perspectives.	Perspectives and experiences from mothers who received care from village birth attendants not included. PNG-specific may not be generalisable to the Pacific region.	5
Choy and Duke [40]	2000	PNG	Village community health worker and child health nurses training in maternal and child health for women from remote villages coordinated by Save the Children with Goroka Base Village hospital.	Program evaluation	Quantitative, community level data.	30 Village community health workers and 25 child health nurses trained. Improvement in village child health worker knowledge including how to recognise a sick child, immunisation requirements, measure growth and infant and young child feeding recommendations.	Strong rationale for program and data provided to highlight need, clear acknowledgement of limitations of the program.	Methods used to assess knowledge of health care workers unclear. PNG-specific may not be generalisable to the Pacific region.	N/A
Datta et al. [30]	2013	PNG	Development of National Action Plan for Elimination of Maternal and Neonatal tetanus in PNG.	Program evaluation	Quantitative descriptive (case study series).	Three cases of neonatal tetanus recorded in PNG in 2011 – 2 survived, 1 died. Al involved village births with untrained attendants, unhygienic cord care and no maternal tetanus vaccinations. National action plan involved national supplementary immunisation rollout – targeting women of child-bearing age (77% of eligible women reach in first round), training for midwives and CHWs around hygienic birth practices and antenatal care services.	Sufficient data supplied to support case series reports, clear rationale for action plan, action plan appears appropriate for targeted outcomes.	Authors, funders and consultation process for development and implementation of action plan unclear. Limited data on measures of success, unclear methods for evaluation of vaccine program. PNG-specific may not be generalisable to the Pacific region.	N/A
Dawson [41]	2016	PNG	Maternal and Child Health Initiative – eight clinical midwifery facilitators placed in four midwifery training schools across PNG to train and mentor midwifery educators.	Program evaluation (first 2 years)	Mixed methods – focus group discussions, semi-structured interviews, regular site visits, meeting minutes, surveys, feedback surveys and assessment pieces.	Increased quantity and quality of midwifery graduates (enrolment numbers, knowledge, clinical skills, compassion towards patients). Increased capacity of midwifery educators and clinicians (Improved student academic results, improved clinical education experiences for students, better maternal and newborn outcomes i.e. neonatal sepsis rates decreased). Accreditation of midwifery curricula in all midwifery schools. Increased networking and collaboration between schools.	Excellent methodology, study design and research questions. Mixed-methods approach and purposive sampling provided breadth and depth of insight. Involvement of a diverse stakeholders, monitoring and evaluation framework were underpinned by program logic model, substantial evaluation period (2 years).	Women's perspectives not included. PNG-specific may not be generalisable to the Pacific region.	N/A
Duke [32]	2017	PNG	Implementation of oxygen delivery systems.	Project protocol	Description of quality improvement (QI) project protocol.	Requires holistic systems approach. Expensive and time-consuming. Needs adequate community engagement. Limited by infrastructure issues (transport and terrain). Requires staff training.	Study sites selected with key stakeholder consultation, clear research questions and outcome measures, metrics suitable for assessing outcome measures, comprehensive evaluation plan, comprehensive discussion about considerations for conducting implementation study.	Women and their families’ perspectives not included. PNG-specific may not be generalisable to the Pacific region.	N/A
Field [31]	2018a	PNG	Community Mine Continuation Agreement Middle and South Fly Health Program aimed at improving health service delivery in remote PNG.	Program evaluation	Mixed methods – pre/post analysis of health service delivery indicators, semi-structured interviews with health workers and assessment of health facility equipment and infrastructure.	Rate of outpatients, outreach clinics, immunisation coverage, antenatal care coverage, all significantly improved with the program except family planning coverage (which improved but not significantly) and supervised births (which decreased despite health worker perceptions). Proportion of facilities with standard equipment, transport and lighting increased. Health worker training, especially obstetric training, was most commonly cited by health workers as leading to improved services. Main barriers included a lack of basic supplies, lack of supervision, lack of community support and cultural barriers that prevented people from accessing services.	Program logic model created, comprehensive mixed methods study designed to triangulate results, appropriate measures taken to ensure accuracy of data when missing/unclear.	No information regarding whether health workers declined to be interviewed, women and their families’ perspectives not included. PNG-specific may not be generalisable to the Pacific region.	5
Field [29]	2018b	PNG	Rural Primary Health Services Delivery project – review development of national policies and standards, establishment of partnerships at provincial level with state and non-state partners in health; health worker training, infrastructure development through construction of two community health posts in each district and community-level health promotion activities to improve demand for services.	Program evaluation	Mixed methods (contextual analysis) – sequential explanatory design involving analysis of baseline quantitative indicator performance, followed up by semi-structured interviews with provincial health administrators.	Less than half of all included districts met the 2013 national target for each performance indicator (outreach clinics, measles vaccination coverage, supervised births). Large variation in performance indicators between and within project districts. Provincial Health Administrators perceived high performance to be influenced by accessibility of health facilities by road in urban areas, competent staff and health services operated by churches or private companies. Inadequate numbers of staff, poorly skilled staff, funding delays and challenging geography were major contributors noted for poor performance.	Presentation of data by district rather than province enabled comparisons between districts within a province. Sequential explanatory approach strengthened quantitative data.	Data sourced from National Health Information System - issues with quality and completeness noted. Limited description of project. Only baseline data from first year of evaluation presented. PNG-specific may not be generalisable to the Pacific region.	5
Gardiner [37]	2016	Fiji	IMproving Perinatal mortality Review and Outcomes Via Education (IMPROVE) programme – series of workshops designed to improve maternity providers knowledge of and confidence in the management of perinatal deaths.	Program evaluation	Mixed methods – pre/post program questionnaire. Knowledge, confidence and satisfaction post program. Open-ended questions to identify what aspects of the program participants found most and least useful and suggestions for improvement.	Proportion of Australian and international participants who were knowledgeable and confident regarding management of perinatal deaths increased from pre- to post- workshop for all items and stations (p<0.001). High degree of satisfaction (>90%) for the workshops both in Australia and internationally. In the international workshops, participants expressed a desire for more hands-on activities and slight adaptation of the program to align with local investigation and classification practices. Participants from middle-income countries felt that the material was relevant to their workplace and their goals for healthcare provision.	Large sample size, delivery of program to a variety of relevant health professions. Inclusion of open-ended questions to enhance assessment of program and improve future delivery.	Long term follow-up of knowledge and confidence would be useful, no assessment of clinical outcomes.	5
Gupta [44]	2017	PNG	Spacim Pikinini - Implant outreach program coordinated by non-government organisations (NGOs), community leaders and health authorities. Local health workers trained in contraception counselling and implant insertion and removal techniques.	Program evaluation	Quantitative descriptive – cross sectional survey of women in two rural provinces who received a contraceptive implant 12 months prior.	860 women who had had a contraceptive implant inserted 12 months prior were surveyed. 97% still had the device in situ, 95% were very happy with it, 76% reported no side effects. Irregular bleeding was the most reported side effect (20.6% of women) but only 7% said the bleeding was bothersome. A desire to have more children was the most common reason for removal of the implant. 92% of women indicated they would use the implant again in future. Around 50% of women acknowledged they had been discouraged from using the device by church members and/or fellow villagers but this was not a reported reason for removal. Documented failure rates were 0.8%.	Large sample size. Comprehensive outline of project and methods. Included direct patient perspective.	Potential for over- and under-measurement of implant failure. Reliability of the data affected by recall bias given retrospective study. Only two rural provinces included, findings not generalisable to all of PNG. In addition, may not be generalisable to the Pacific region.	5
Kamblijambi [43]	2017	PNG	To examine a PNG University's Bachelor of Clinical Maternal and Child Health programme in respect to macro and micro knowledge and skill transferability to all stakeholders.	Program evaluation	Qualitative – curriculum analysis, interviews with graduates, focus groups with Village Birth Attendants (VBA) and Village Community Health Workers (VCHWs), face-to-face interviews with postnatal women (recipients of targeted VBA education).	Insufficient resources and program too short to meet objectives. Interviewees reported feeling well equipped with core theory and clinical knowledge but often felt unable to implement the knowledge gained in their work environment. Group learning considered particularly useful. Rural community placements had a positive influence on attitudes, behaviour and clinical practice of participants. Working in partnership with the community considered valuable. Participants considered assessment practices to be poor and would have preferred assessments that enabled knowledge retention and enhanced clinical skills acquisition and improvement. Participants felt the program had enhanced their career prospects.	Inclusion of wide range of stakeholders. Content validity of interview guide was ensured through revision and conduction of a pilot study. Purposive sampling employed to ensure diversity of participant location and experience.	Only phases 1 and 2 reported. No reporting of outcome measures related to morbidity, mortality or health practice implementation. PNG-specific may not be generalisable to the Pacific region.	5
Kirby [48]	2013	PNG	Mother Baby Bundle Gift Program aims to encourage mothers to birth in health facilities. Program consists of mother and baby support packages, staff incentives, staff emergency obstetric and newborn care training and emergency supplies. Installation of waiting houses, solar lighting and water supplies are also included in the program.	Brief report	Data source not clear – analysis of outcome data for mothers and newborns.	Three health facilities included. Implementation of the Baby Bundle Gift Program resulted in 55-143% increase in supervised deliveries across all three facilities. No increase in complications. 2 maternal deaths over 10-month study period.	Informed by survey of women in the village - prospective program recipient involvement in project design. Simple, relatively cheap intervention with impact. Comprehensive description of QI activity.	Intervention multi-pronged but results presented in a way which attributed all outcomes to the mother baby gift program. PNG-specific may not be generalisable to the Pacific region.	4
Kirby [33]	2015	PNG	To identify causes of maternal mortality and identify appropriate and sustainable solutions in Milne Bay province. Also to review whether the Mother Baby Bundle Gift Program can prevent maternal deaths.	Descriptive study	Mixed methods - Surveys and interviews with a focus on reviewing maternal deaths. Health care workers and family members asked to identify causes of maternal mortality and classification of factors according to a three-phases-of-delay model.	31 maternal deaths identified from maternal mortality survey. 20 cases (65%) related to haemorrhage: postpartum haemorrhage (11), retained placenta (6) and antepartum haemorrhage (3). Other causes included sepsis, ectopic pregnancy, eclampsia, cerebrovascular accident, pneumonia and tuberculosis. Phase one delays (deciding to seek care) responsible for 80% of cases. MBG program was associated with increases in the number of supervised health centre births across all health centres. 3 maternal deaths in the village and 3 maternal deaths at health centres over the study period.	All major health centres in the region included. Intervention targeted towards specific barriers to appropriate perinatal care in the region.	Short term follow-up. Intervention multi-pronged but results presented in a way which attributed all outcomes to the mother baby gift program. Value of program beyond individual province not discussed. PNG-specific may not be generalisable to the Pacific region.	5
Mannering [73]	2013	Samoa	To evaluate the POINTS* education program to improve neonatal care from the perspective of neonatal nurses. *POINTS – pain management, optimal oxygenation, infection prevention, nutritional and temperature management, and supportive care.	Brief report	Reports from neonatal nurses involved in the program.	Increase in neonatal nurse neonate. ‘5 moments of hand hygiene’ prominently displayed and promoted throughout unit. Increase in pain relief measures practised Improved developmental support for premature babies. Increased enthusiasm for neonatal education. Increased profile of neonatal nursing.	Clear description of programme and outcomes.	Evaluation findings not included in the report, anecdotal reports of improvement only. Data collection processes unclear.	N/A
Moores [42]	2016	PNG	To explore the impact of strengthening midwifery education in PNG.	Descriptive study	Mixed methods - surveys, focus groups and interviews were via telephone phone or in person.	89.3% of respondents were working as midwives, with an additional 3% working as midwifery or nursing educators. Graduates located in 21 of the 22 provinces of PNG, predominantly in urban areas with 41% working in rural areas. Most graduates felt that the theoretical, clinical and rural practice components of their education course had prepared them for their current clinical practice.
						Many felt they had improved basic midwifery skills and emergency obstetric skills, however some felt there were inadequate opportunities to acquire the necessary skills to feel competent and confident. Graduates exhibited increased skills acquisition and confidence, leadership in maternal and newborn care services and a marked improvement in the provision of respectful care to women.	Large sample size. Almost all provinces included. Appropriate use of quantitative data to support qualitative findings.	Graduates working in remote areas not included. Limited outcome data (e.g. number of women giving birth at facilities, morbidity, mortality) and patient perspectives. PNG-specific may not be generalisable to the Pacific region.	5
Narayan [34]	2009	Fiji	A report reviewing ten cases of neonatal Enterobacter aerogenes infections in a Fijian NICU and the measures implemented to prevent transmission and stop the outbreak.	Case series and description of QI activity	Examination of medical records, medical and microbiological tests, description of QI activity.	10 of the 18 infants admitted to the unit in May 2007 developed septicaemia with extended-spectrum beta-lactamase-producing Enterobacter organisms. Three infants died. Infants ranged in age from 4 hours to 19 days. The most common reason for admission to the NICU was respiratory distress with premature birth. A communally used bag containing normal saline was deemed to be the source in all 10 cases. Three major issues were identified as contributing to the outbreak: poor adherence of physicians and nurses to hand hygiene protocol; prolonged use of normal saline and dextrose 10% bags and for multiple patients; infrequent disinfection of the IV rubber ports on fluid bags and lines before withdrawal and injection. Infection control measures introduced including simple hand hygiene, use of single-use vials of IV fluid bags and strict aseptic technique for injections - no new cases were identified after the implementation of these control measures.	Identified contributing factors to outbreak and designed infection control measures targeted specifically towards the outbreak. Clear description of protocol. Ongoing monitoring of infection control compliance.	Follow up period not specified. Sustainability of the program not discussed. Issues within one hospital may not reflect wider health system issues related to infection prevention and control. Fiji-specific may not be generalisable to the Pacific region.	N/A
O'Keefe [38]	2011	PNG	To evaluate the effectiveness of the East Sepik Women and Children's Village Health Volunteers program.	Before and after study	Quantitative – review of medical and health records.	In East Sepik province, the proportion of women receiving their first antenatal care visit from a village health volunteer increased from 6% in 2007 to 15% in 2010. Proportion of women whose childbirth was attended by a village health volunteer increased from 8% to 13.2%. Reported health facility coverage of women attending a health facility for childbirth decreased from 28% in 2007 to 22% in 2010.	Large sample size, data cleaning performed before analysis in attempt to ensure accuracy, highlights contribution of village health worker to maternal and newborn care.	Potentially inaccurate population estimates – based on census data collected 9 years prior. Quantitative data only – limited ability to explain causes behind increased proportion of births attended. PNG-specific may not be generalisable to the Pacific region.	5
Olita'a [74]	2019	PNG	To evaluate the safety and effectiveness of a protocol based on giving minimal or no antibiotics to well term babies born after premature rupture of membranes and its effect on neonatal sepsis at Port Moresby General Hospital in PNG.	Descriptive study (case report)	Quantitative. Data collected from medical records, clinical assessment and information provided by healthcare staff.	133 babies born to mothers who had premature rupture of membranes were assessed at 7 days. Signs of sepsis occurred in 10 babies (7.5%; 95% CI 4.4% to 13.2%) in the first week. An additional four (3%) had any sign of sepsis between 8 and 28 days. There was one case of bacteraemia and no deaths. 37 were lost to follow-up, but hospital records did not identify any subsequent admissions for infection. Sepsis rates documented were found to be comparable with other studies in low-income countries.	Used definition of neonatal sepsis in line with WHO recommendations (ie including clincial features). Use of different management protocol depending on mother's intrapartum antibiotic status.	No background information provided on previous sepsis rates in neonates. Study conducted in peri-urban setting with health centre and hospital access, limited generalisability across PNG. In addition, PNG-specific may not be generalisable to the Pacific region.	3
Raman [35]	2015	Fiji	Development and evaluation of a perinatal mortality audit system in 3 hospitals in Fiji.	Program evaluation	Mixed methods – perinatal audit data and key informant interviews.	141 stillbirths and neonatal deaths were analysed from 2 hospitals (57 from hospital A and 84 from hospital B; forms from hospital C excluded because incomplete/illegible). 28 (49%) stillbirths were recorded in hospital A compared with 53 (63%) in hospital B. 82% of stillbirths from hospital A were due to a delay in receiving care.
						Factors that led to delays in receiving appropriate healthcare included patient; sociocultural and family; resource; social and human infrastructure; health system; care; and process factors. Substantial health system factors contributing to preventable deaths were identified, and included inadequate staffing, problems with medical equipment, and lack of clinical skills. Leadership, teamwork, communication, and having a standardised process were associated with increased uptake of perinatal mortality audit system.	Perinatal mortality audit datasheet refined for use in Fiji. Mixed methods design - investigation of factors affecting success of perinatal mortality audit. Exploration of factors that led to delays in accessing appropriate health care.	Most results derived from Hospital A (tertiary hospital, which could bias results). Incomplete data from Hospitals A and B. Fiji-specific may not be generalisable to the Pacific region.	5
Rumsey [46]	2016	Cook Islands, Kiribati, PNG, Solomon Islands, Tonga, Vanuatu, Fiji, Nauru, Samoa, Tokelau, Tuvalu, Niue	To describe the experiences of 34 nursing and midwifery leaders in the South Pacific region who undertook the Australian Award Fellowship (AAF) program.	Descriptive study	Qualitative – semi-structured interviews (individually or in pairs).	Thirty-four nurses and midwives from 12 countries participated. Four main themes were identified: having a country-wide objective, learning how to be a leader, negotiating barriers and having effective mentorship. Participants had a good understanding of their country's needs and objectives for improvement of their health care and workforce and described positive learning experiences from involvement in the AAF program. Participants deemed their mentorship from country leaders highly valuable in relation to completing their projects, networking and role modelling.	Inclusion of participants from multiple years of the program and a variety of countries.	Limited outcome data to determine if program translated into improved care.	5
Sa'avu [57]	2014	PNG	To understand the quality of care provided for care in five rural district hospitals in the highlands of PNG.	Descriptive study and protocol development	Quantitative (baseline assessment of quality of care provided) – structured survey forms, medical and admissions records, oxygen/electricity records.	Many district hospitals are run by under-resourced NGOs. Most hospitals had general wards in which both adults and children were managed together. Paediatric case numbers ranged between 232 and 840 patients per year with overall case-fatality rates (CFR) of 3–6% and up to 15% among sick neonates. Pneumonia accounts for 28–37% of admissions with a CFR of up to 8%.
						There were no supervisory visits by paediatricians, and little or no continuing professional development of staff. Essential drugs were mostly available, but basic equipment for the care of sick neonates was often absent or incomplete. Infection control measures were inadequate in most hospitals. Cylinders were the major source of oxygen for the district hospitals, and logistical problems and large indirect costs meant that oxygen was under-utilised. Multiple electricity interruptions, but hospitals had back-up generators to enable the use of oxygen concentrators. After 6 months in each of the five hospitals, high-dependency care areas were planned, oxygen concentrators installed, staff trained in their use, and a plan was set out for improving neonatal care.	Detailed data obtained from reliable sources. Included plan for future QI improvement.	Unclear study design. PNG-specific may not be generalisable to the Pacific region.	4
Sandakabatu [36]	2018	Solomon Islands	To evaluate a new child mortality review process introduced at the National Referral Hospital, Honiara, Solomon Islands.	Program evaluation	Mixed methods – audit data from clinical records and observations of audit meetings.	33 child mortality review meetings were conducted over 6 months, reviewing 66 neonatal and child deaths. Areas for improvement included use of systematic classification of causes of death, inclusion of social risk factors and community problems in the modifiable factors and more follow-up with implementation of action plans. To avoid preventable deaths, greater emphasis needed with communication, clinical assessment and treatment, availability of laboratory tests, antenatal clinic attendance and equipment for high dependency neonatal and paediatric care. Causes of deaths were classified, although there was a need among health workers to clarify what constitutes an immediate and underlying cause of death, as there was overlap. Classification of modifiable factors was performed, but it was identified that there was a lack of consideration for social risks and community-based factors.	Sufficient study duration to assess change over time. Suggestions to improve the audit process offered.	Mortality cases often discussed many weeks/months after event, potentially affecting recall. Solomon Islands-specific may not be generalisable to the Pacific region.	5
Thiessen [47]	2018	PNG	To evaluate facilitators and barriers to effectiveness of a Public Private Partnership (PPP) funded Reproductive Health Training Unit (RHTU) in PNG.	Descriptive study	Qualitative – interviews with key stakeholders.	Features of the PPP that enabled the RHTU to be more effective included understanding and agreeing with national plan for PPPs and maternal child health, strong champions & strong relationships with decision making bodies and creating autonomy and branding. Features of the PPP that created barriers to the effectiveness of the RHTU included lack of governance leading to confused decision making and roles & responsibilities, differing institutions, cultures and ownership struggles, lack of capacity within institutes especially National Department of Health.	Large number of stakeholders interviewed (85) from range of disciplines, participants interviewed at multiple timepoints to track changes in perception over time.	Evaluation unable to determine if partnership improved maternal and newborn health outcomes. Community perspectives not included. PNG-specific may not be generalisable to the Pacific region.	5
Tynan [45]	2018	PNG	To examine healthcare worker perceptions of health system factors impacting on the performance of Prevention of Parent to Child (PPTCT) programmes.	Descriptive study	Qualitative – semi-structured interviews with healthcare workers involved in the program.	Sixteen interviews undertaken with healthcare workers involved in the PPTCT program. Major factors reported as barriers for implementing a successful PPTCT programme included broken equipment; problems with supply and access to HIV test kits, ART and HIV prophylaxis for children; significant shortage of appropriately trained and supported health care workers; and the absence of leadership and coordination of this complex, multi-staged national programme from central government. The need for women and children to engage with the health system over an extended period and often at different services with different health care workers was highlighted. Challenges were specifically faced in accomplishing initial diagnosis of the expectant mother and delivering postnatal and paediatric follow‐up services. A few participants also identified several weaknesses with health information systems including timely access to data and ineffective reporting for monitoring and evaluation.	Sound methodological rigour. Classification of results according to the framework of the building blocks of a health system developed by WHO.	Patient perspectives limited. Inclusion of 2 provinces with high burden of HIV, may not reflect issues affecting provinces with lower burdens or different barriers to care. PNG-specific may not be generalisable to the Pacific region.	5
Usher [54]	2003	Fiji	To conduct an impact evaluation of the Nurse Practitioner role in Fiji. Nurse Practitioners must hold midwifery and public health qualifications, have approximately 15 years of service and have successfully completed a 14-month course run by the Fiji School of Nursing.	Descriptive study	Qualitative – semi-structured interviews and focus group discussions.	18 nurse practitioners and 54 stakeholder interviews and three community focus group (22 community members) were conducted. Villagers who currently have access to the service provided by these NPs find the role to be extremely beneficial to the health of their community. They, and key stakeholders, expressed a high degree of satisfaction with the role. Issues raised related to the practice of using nurse practitioners to replace medical officer positions at hospitals, lack of formal career path and lack of involvement in decisions about placement in remote areas.	Inclusion of community and all key stakeholder perspectives. Good rationale for and rigorous methodology. Clear description of Nurse Practitioner role.	Lack of qualitative data (interview or focus group quotes) to support findings. Implications of study may be overstated given lack of data provided. Ethics approval authority not specified. Fiji-specific may not be generalisable to the Pacific region.	4
West [55]	2017	PNG	To determine how the PNG Maternal and Child Health Initiative approach contributed to strengthening midwifery education in PNG.	Program evaluation	Qualitative – semi-structured interviews.	26 midwifery educators were interviewed. Individual and relationship factors perceived as enabling improved midwifery teaching and learning were knowing your own capabilities; being able to build relationships; and being motivated to improve the health status of women. Four themes identified as constraining midwifery educators were lacking a mutual understanding of capacity building; not feeling adequately prepared to work together; not feeling culturally competent and lack of a supportive environment.	Criterion sampling used to ensure diversity of participant perspectives. Data collection and analysis underpinned by a theoretical framework. Strong use of quotes to support findings.	PNG-specific may not be generalisable to the Pacific region.	5
WHO [50]	2005	Regional (including PNG))	WHO regional 'Making pregnancy safer' strategy.	Strategy review	-	Strategy contains 4 strategic areas with 12 component strategies. Pacific relevant ones selected below: Service capacity increased at the community and referral levels in all priority countries. Development of guidelines and services protocols around Managing Complications in Pregnancy and Childbirth. Birth attendants trained in Solomon Islands. Midwives trained in Vanuatu. Training curriculum on Pregnancy, childbirth, postnatal and newborn care (PCPNC) prepared for 10 Pacific island countries.
						Training on PCPNC conducted in 10 Pacific Is countries. Information systems strengthened to monitor the progress of achieving quality of care. Maternal child health surveillance system software has been developed and tested in Solomon Islands. Training conducted for 6 island countries. Pilot test of reproductive health surveillance system conducted in 2003 in Solomon Islands. Training on information systems conducted in Kiribati. Family planning promoted to reduce unwanted pregnancy: 10 Pacific Island countries have revised the service protocol of family planning to improve the quality of service. Seminars on STIs and HIV/AIDs conducted in Tuvalu. Partnerships strengthened for Making Pregnancy Safer programmes in the priority countries UNICEF, UNFPA, IPPF, PATH and the Secretariat of the Pacific Community (SPC) were actively involved in the development and finalisation of the Regional Framework for Accelerated Action for Sexual and Reproductive Health of Young People.	Regional strategy, promoting partnerships and collaborations between agencies, scaling up known effective interventions.	One size fits all approach may not be appropriate for all countries in the region.	N/A
WHO [51]	2009	Regional	Meeting Report -Situation of Maternal and Newborn Health in the Pacific.	Strategy review	Meeting with key stakeholders to review current situation and progress made in Pacific Island countries and country inputs required for strategy document.	Country inputs to strategy documents to: reflect Pacific perspectives, strengthen newborn care in ‘Making Pregnancy Safer framework’, check data for accuracy and cite sources, review technical terms, improve access and availability of services, identify minimum competencies for skill birth attendants, ensure coherence and coordination among reproductive health and other health programs, identify role of men, gender based violence, improving emergency care, costing for countries to have evidence on how much maternal and newborn health interventions and programmes cost, improve monitoring and evaluation.	Regional strategy with emphasis on pacific solutions.	Costing of initiatives/framework and/or potential funding support not included.	N/A
WHO [52]	2013	Regional	Draft action plan for Healthy Newborns in the Western Pacific 2014-2018.	Draft action plan	Consultation meeting - held in Philippines. Plenary and small group discussions on the regional action plan. 5 small groups reviewed sections of the plan. Country teams reviewed implementation of Early Essential Newborn Care and identified key actions to move forward.	Ensure consistent adoption and implementation of EENC. Improve political and social support to ensure an enabling environment for EENC. Ensure availability, access, use of SBAs and essential MN commodities in a safe environment. Engage and mobilise families and communities to increase demand. Improve the availability and quality of perinatal information.	Regional strategy, focus on implementation, use of group work.	Limited number of Pacific countries participated in consultation meeting.	N/A
WHO [75]	2014	Regional	Final Action Plan for Healthy Newborn Infants in WPRO (2014-2020).	Action plan	Consultation meeting - held in Philippines. Plenary and small group discussions on the regional action plan. 5 small groups reviewed sections of the plan. Country teams reviewed implementation of Early Essential Newborn Care and identified key actions to move forward.	Five strategic actions support full implementation of Early Essential Newborn Care (EENC): 1. Ensure consistent adoption and implementation of EENC. 2. Improve political and social support to ensure an enabling environment for EENC. 3. Ensure availability, access and use of SBAs and essential maternal and newborn commodities in a safe environment. 4. Engage and mobilise families and communities to increase demand. 5. Improve the quality and availability of perinatal information.	Regional strategy, developed in collaboration, upstream/downstream initiatives included.	Only 2 Pacific countries included.	N/A
WHO [53]	2016	Regional (PNG, Solomon Islands)	First biennial progress report - Action Plan for Healthy Newborn Infants in the Western Pacific Region (2014-2020). To review progress of action plan for healthy newborn infants in Western Pacific Regional Area.	Progress report	Independent review group.	Report recommendations for next phase: 1. Plan and secure long-term funding. 2. Build consensus and create demand. 3. Further scale up coaching of the basic EENC package to 28,000 health facilities in the region. 4. Strengthen monitoring and evaluation for EENC. 5. Develop, test and introduce new methods and guidelines in collaboration with WHO and other partners.	Coaching approach - emphasis on changing health worker practice, emphasis on quality of care during labour and childbirth.	Only 2 Pacific countries included.	N/A

^aThe Mixed Methods Assessment Tool (MMAT) assesses the quality of qualitative, quantitative, and mixed methods studies. It focuses on methodological criteria. It is a 1–5 scale and includes five core quality criteria for each of the following five categories of study designs: (a) qualitative, (b) randomised controlled, (c) nonrandomised, (d) quantitative descriptive, and (e) mixed methods. N/A represents not assessable rather than not applicable.