Prevalence and risk factors of adverse birth outcomes in the Pacific Island region: A scoping review.

BACKGROUND: Prevalence and exposures of adverse birth outcomes is well studied in low-and-middle-income countries but not well-established for the Pacific Island region. Our study mapped the available evidence on low birth weight (LBW), preterm birth, and small for gestational age (SGA)'s prevalence and their corresponding risks in the region.
METHODS: We followed the five-staged Arksey and O'Malley's framework with clinicians' consultation in the region. Five scholarly databases and non-indexed studies were searched and extracted data were analysed as numerical and thematic summaries mapping the outcomes and exposures.
FINDINGS: We included 20 studies representing 11 Pacific Island countries with the following mean prevalence and associations at 95% confidence interval. Estimated mean prevalence for LBW and preterm births were 12% and 13%, respectively. LBW were associated with malaria in pregnancy [aOR 3.3 (1.00, 10.60)], and betel nut and tobacco [aOR 2.4 (1.00, 6.00)]. Preterm births were associated with malaria in pregnancy [aOR 6.6 (2.46, 17.62)] and maternal obesity [aOR 1.5 (1.00, 2.30)]. SGA were associated with short stature [aOR 1.7 (1.22, 2.41)] and no antenatal bookings [aOR 4.0 (2.12, 7.57)].
INTERPRETATION: Several significant factors identified were malaria infection, obesity, betel nut and tobacco and no antenatal care, also validated by clinicians consulted. FUNDING: Australia National Health and Medical Research Council.

Evidence before this study

Adverse birth outcomes such as low birth weight, preterm birth and fetal growth restriction are the primary cause of infant and child mortality and morbidity in low-and middle-income countries, including the Pacific Island region. At the outset, we conducted a literature review and initial searches on scholarly databases and online using Google and Google scholar search engines to survey the current literature on the topic using key concept terms such as adverse “birth outcomes”, “pregnancy risk factors”, and the “Pacific Island region.” We found no systematic or narrative, or scoping reviews conducted in the Pacific region on this topic. We also found a minimal number of primary studies that reported on the prevalence and measures of association of exposures and adverse birth outcomes in the region.

Added value of this study

To our knowledge, this is the first comprehensive scoping review exploring the prevalence and factors of adverse birth outcomes in the Pacific Island regions. There is limited number of research in the region on the prevalence and risk of adverse birth outcomes. Of the few studies conducted, which we reviewed in the current study, malaria in pregnancy, betel nut and tobacco use during pregnancy, obesity in pregnancy, and no and low antenatal care were notable exposures associated with adverse birth outcomes. The findings were also validated by 18 clinicians consulted working in maternal and neonatal health care from 4 countries in the region.

Implications of all the available evidence

Despite limited studies in the Pacific region, we found several important risk factors such as malaria infection, obesity, betel nut and tobacco and no antenatal care during pregnancy to be associated with the adverse birth outcomes, which are essential to inform the literature and evidence for clinical practice. Also, further study is needed to investigate the burden of adverse birth outcomes in the region.

Alt-text: Unlabelled box

Introduction

Adverse birth outcomes contribute to poorer childhood outcomes and are the primary cause of child mortality and morbidity in low-and middle-income countries (LMICs). Adverse birth outcomes, including preterm birth, small for gestational age (SGA), low birth weight (LBW), stillbirths and miscarriage, are associated with a range of social, biological, and environmental risk factors.1, 2, 3, 4

Early birth, most commonly defined as preterm birth or birth before 37 weeks of gestation, is a major determinant for neonatal mortality and morbidity. Preterm birth often necessitates therapy, and its prevention requires treatment, both of which are less accessible in LMICs, and can contribute to excess morbidity burden., Growth restriction while in utero can contribute to morbidity with preterm birth. SGA, a retrospective proxy for fetal growth restriction, occurs when the birth weight falls below a cut-point in the gestational age and sex-specific birth weight distribution. This cut-point is commonly the lowest 10th centile or a multiple of standard deviations below the mean weight. In LMICs, retrospective proxies fetal growth restriction such as SGA is determined at birth due to the lack of fetal monitoring systems during pregnancy, no routine recording of gestational dates and start of the last menstrual period, and limited availability of ultrasound dating. The most common proxy for both preterm birth and fetal growth restriction in LMICs is LBW, which is defined as birth weight less than 2500 g. Term LBW, defined as low birth weights from 37 weeks of completed gestation, and SGA are the preferred indicators for fetal growth restriction as newborn weight is also a consequence of time spent in utero. It is well established that infants that are preterm, term LBW and SGA are at elevated risk of early childhood mortality.,

Most of the deaths attributable to fetal growth restriction and preterm birth occur as miscarriages before the gestational age of viability or stillbirths, occurring from viability to delivery. Although there is no global consensus on when a fetus is viable, the most well adopted gestational age cut-off to discern miscarriages from stillbirths in LMICs is 28 weeks of gestation. According to the global burden of disease study, the prevalence of stillbirths in LMICs has been reported to be as high as 20 per 1000 live births, with Papua New Guinea (PNG) reporting 16.3 per 1000 live births.

The Pacific Island region comprises countries with developing economies with potentially high infant and child mortality rates resulting from poor pregnancy and birth outcomes. The region consists of 22 Melanesian, Polynesian, and Micronesian countries with 12 independent sovereignties and 10 dependent or unincorporated territories either by the United States, France or New Zealand. It is a diverse territory in terms of socioeconomic development, culture, population and geography, with a myriad of risk factors relevant to perinatal health and early childhood mortality. Most Pacific Island countries are under-resourced with inadequate health facilities that affect maternal and infant health to various extents across the region, signified in several health indicators. For example, infant mortality has been reported to be higher in the PNG region of Melanesia (70/1000 live births) compared to the Cook Islands of Polynesia (10/1000 live births)., Antenatal care coverage and skilled birth attendance in Tonga and Samoa are over 95%,18, 19, 20 compared to 49% in PNG. Geographical challenges across land and sea magnified by infrastructure limitations in PNG and countries of Melanesia are some of the barriers contributing to low antenatal attendance and skilled birth., Trends in morbidity and mortality due to communicable and non-communicable diseases (NCD) varies across the different subregions. Obesity-related NCD morbidities and mortalities in Polynesian countries of Tonga and Samoa have significantly increased. While, in the Melanesian region, PNG and Solomon Islands have seen a two-fold increase in the burden of NCD and communicable diseases such as diabetes and malaria across adult populations, and severe malnutrition among women and children.

Maternal health in the Pacific Islands is challenged by geographical disparities and dynamics in the various parts of the region. Most Pacific Island communities are isolated by sea or land, which limits access to health services, which is added to a background of already elevated risk from socioeconomic disadvantage., The low lying islands of Polynesia and Micronesia and elsewhere in the region are continually faced with the health challenges associated with global warming. The wider Pacific island countries are frequently affected by natural hazards such as cyclones, floods and coastal erosion, earthquakes, and tsunamis affecting the people's livelihood and overstretching the limited health resources., Vector-borne disease such as malaria and dengue is prevalent due to the climate, recurrent heavy rains and tropical cyclones., Women and children are often vulnerable to malnutrition due to food shortages and diarrhoeal diseases in the aftermath of natural disasters.,

Cultural and social practices also precipitate various health risks. Betel nut use is practised in PNG, Solomon Islands, Palau and the Marshall Islands. Studies have reported that betel nut use is associated with oral cancers in the region and elsewhere., Studies in LMIC countries outside the Pacific Island region (South East Asia) have also shown betel nut use to be adversely associated with adverse birth outcomes. Kava, a traditional drink made from Piper methysticum, is another unique exposure to many Pacific Islanders, including those from Fiji, Vanuatu and Samoa. There is a dearth of studies on kava use and its impact on birth outcomes, and to date, there has been no systematic assessment of the burden of, and risk factors for adverse birth outcomes in the Pacific Island region. This scoping review aimed to map the available literature on the prevalence of adverse birth outcomes and associated risk factors in the Pacific Island countries and gain additional insights and recommendations from the perspective of health professionals.

Methods

Study design and protocol

The systematic scoping review was registered with the Joanna Briggs Institute, and a detailed description of the methods was published as a protocol. Our review followed the Joanna Briggs Institute Reviewers Manual derived from Arksey and O'Malley's five-staged methodological framework. We also included a consultation exercise, interviewing health professionals from four countries in the region.

Research question

Our review was guided by the following research questions: (i) what is the prevalence of the adverse birth outcomes in the Pacific Island region? And (ii) what are the risk factors for adverse birth outcomes in the Pacific Island region?

Study selection and inclusion

We selected studies that corresponded to the population, concept, and context (PCC) criteria. In the review, ‘population’ includes women and births from the pacific countries, ‘concept’ includes the various social, health and behavioural risk factors associated with adverse birth outcomes. The ‘context’ included 22 sovereign island states and territories of the region, namely: American Samoa, Cook Islands, Easter Islands, Federated States of Micronesia, Fiji, Guam, Kiribati, Commonwealth of the Northern Mariana Islands (CNMI), Marshall Islands, Nauru, New Caledonia, Niue, Palau, PNG, Samoa, Solomon Islands, Tahiti, Tokelau, Tonga, Tuvalu, Vanuatu, and Wallis and Futuna. We considered studies published between 1 January 2000 and 28 February 2021. Studies with populations from high-income countries located in the Pacific region (e.g., Australia, New Zealand, the United States) were excluded.

Literature search

The authors consulted the Curtin University Health Science librarian to design the search strategy. Prior to the search, a preliminary literature review was undertaken to understand the extent of literature on the subject in the context. The literature search was carried out in three-stages as outlined by the Joanna Briggs Institute., In stage one, key concept terms were selected from CINAHL to identify Medical Subject Headings (MeSH) or text terms contained within the titles and abstracts of articles. Key concept terms used were adverse “birth outcomes”, “pregnancy risk factors” and the “Pacific Island region.” In stage two, all MeSH terms, key concept terms, and their synonyms were combined with Boolean operators, truncations, and wildcards to develop search strings and applied across the selected databases such as CINAHL, Medline, ProQuest, SpringerLink and Scopus. In this stage, search strings were either general or specific with their corresponding key concept terms and synonyms combined with MeSH terms identified before applying them to the database. Stage three involved searching using Google and Google Scholar for non-indexed studies on regional websites of the World Health Organisation (WHO), the Secretariat of Pacific Community and United Nation International Children Emergency Fund and hand searching reference lists of studies initially retrieved. A table of full search strategies is outlined in Supplementary File 1.

Data charting process

Once all retrieved records that matched the PCC criteria were exported to Endnote 9, we conducted data extraction based on a priori data extraction tool developed by LSKK and discussed with the co-authors (Supplementary File 2). During the data extraction, all results were entered into Excel spreadsheets alongside standard bibliographic information that included author(s), year of publication, country where the study were conducted, aims and purpose, study population, intervention type, intervention duration, outcomes, and details of key findings, including prevalence on the outcomes identified in the study and associated risk factors. LSKK conducted the data extraction and later reviewed and discussed by GAT, JJ, and GP. All arising questions and uncertainty during the process was discussed with the research team to reach an agreement.

Summarising and reporting the results

The findings, presented as tables, were obtained by mapping the data extracted from the selected articles and guided by Arksey and O'Malley. Information on the prevalence of LBW, preterm birth and SGA were presented in a table (Supplementary File 3). We also summarised and presented the risk factors associated with adverse birth outcomes from the included studies. We presented our findings following the preferred reporting items for scoping review and meta-analysis (PRISMA-ScR, Supplementary File 4).

Health professionals’ consultation

The consultation stage was conducted via one-on-one face to face or telephone interviews. Consultants were required to be health professionals (nurses, midwives, and obstetricians) from Pacific Island countries working in antenatal care, births, and neonatal units. Recruitment was purposeful and incorporated snowball sampling, with 18 consultants originally recruited from hospitals and health clinics. The interview schedule was informed by the literature review preliminary findings and the research team, experienced in qualitative research and the topic area. The interview questions asked about risk factors, adverse birth outcomes and recommendations to improve birth outcomes in Pacific Island countries.

After informed consent was obtained, interviews were conducted in English, Solomon Island Pidgin English, PNG Tok Pisin or Vanuatu Bislama, as the principal investigator was fluent in all languages and Creoles. Collected data were both quantitative (counts of nominated risk factors and adverse outcomes) and qualitative data (health professionals’ perspectives on issues). Recorded data were translated from the Creoles to English. NVIVO version 20 was used to manage the interview data and undertake the content and thematic analysis, which involved data familiarisation, formulation of codes, and development of themes., Content and thematic analysis were deemed appropriate for this study as this approach enabled the identification and quantifying of nominated adverse birth outcomes and risk factors and supported the generations of themes. Narrative summaries and quotes were used to illuminate the meaning of the themes., Demographic data (age, qualification, profession, country) were collected. Ethical approval was obtained from Curtin University.

Ethics approval

The part of the study which involved consultation of clinicians was approved by Curtin University human research committee with approval number: HRE2020-0530.

Role of the funding source

Funding sources was not involved in the study design; collection, analysis or interpretation of the data; and in writing of the manuscript or in the decision to submit the manuscript for publication.

Results

Selection process and characteristics of studies

A total of 959 records were identified in the initial search of which 918 were retrieved from scholarly databases and 41 from additional website searches. After excluding 255 duplicates, 704 remained for titles and abstract screening, of which 62 was retrieved for full-text screening. Next, we excluded studies that did not meet the PCC criteria (n = 36), those conducted before 2000 (n = 3), not able to be accessed (n = 2) and abstract articles (n = 1). After all exclusions, 20 studies remained for the final review (Figure 1 illustrates the selection process using PRISMA flow diagram).

Figure 1

PRISMA flow diagram

From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(7): e1000097. doi:10.1371/journal.pmed1000097.

We identified 20 studies, 11 peer-reviewed articles44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54 and nine demographic health survey reports,,55, 56, 57, 58, 59, 60, 61 conducted in 11 countries in the Pacific Island region, with a total population of 38,148 singleton births. Of the reported studies, seven were cohort studies,,,,,52, 53, 54 and 11 were cross sectional studies,,,,,55, 56, 57, 58, 59, 60, 61one case control study and one randomised clinical trial. Nine of the studies were conducted in PNG,,47, 48, 49, 50, 51, 52, 53, 54 and others were conducted in other sub-regions: CNMI (n = 1), Kiribati (n = 1), Marshall Islands (n = 1), Nauru (n = 1), Palau (n = 1), Samoa (n = 1), Solomon Islands (n = 2),, Tonga (n = 1), Tuvalu (n = 1), and Vanuatu (n = 1) (Table 1).

Table 1

Summary of studies included in the review.

(i) Eleven journal articles
Author(s)	Year	Study context	Study aims	Study design	Population (Mother and infants, Births)	Adverse birth outcomes	Risk factors investigated
Journal
Unger et al.54	2019	PNG Madang Province	To evaluates the associations between malaria infection and hemoglobin in PNG women.	Prospective cohort study	1976	LBW, preterm birth, small for gestational age (SGA)	Plasmodium Falciparum malaria infection. (microscopic submicroscopic) and plasmodium vivax malaria infection
Cruz et al.46	2018	CNMI	To explore the racial/ethnic disparities that exist among Pacific Islander women residing in the CNMI and newborns.	Retrospective cohort study	8918	preterm birth, LBW	Race of CNMI, Filipino, Chinese, other Pacific Islander, maternal age and number of antenatal number of visits
Fowkes et al.47	2018	PNG Madang Province	To determine the association between iron deficiency and birth outcomes, and malaria	Longitudinal cohort study	279	preterm birth, LBW	Iron deficiency anaemia, iron deficiency (Hb < 11 g/dL), moderate anemia, severe anaemia (Hb < 7 g/dL*), plasmodium falciparum malaria (peripheral blood)
Lufele et al.48	2017	PNG Madang Province	To investigate the prevalence of Plasmodium malaria its risk factors, maternal anaemia and birth outcomes associated with women receiving at least one dose of Intermittent preventive treatment in pregnancy (IPTp).	Randomised clinical trial	1451	LBW, preterm birth	Plasmodium malaria infection, active placenta malaria, acute malaria infection, chronic malaria infection, past plasmodium malaria infection and anaemia
Berger et al.44	2015	Palau	To determine the effects of betel nut with tobacco use on pre-pregnancy obesity and adverse birth outcomes.	Retrospective cohort	1,171	preterm birth, LBW, Preterm LBW, Term LBW, Preterm normal weight	Betel nut chewing, tobacco use, combined use of tobacco and betel nut, and obesity in pregnancy-Body Mass Index (BMI) >30)
Cafaro et al.45	2015	Kirakira Solomon Islands Solomon Islands	To calculate the incidence of LBW and prematurity and proportion of women receiving antenatal care.	Retrospective audit and cross-sectional study	1295	LBW, preterm birth	Betel nut use, tobacco use and urinary tract infection in pregnancy
Ome-Kaius et al.49	2015	PNG Madang Province	To access the effects of areca nut chewing on pregnancy outcomes, birth weight, anaemia in a large cohort of pregnant women.	Longitudinal cohort	1769	LBW, Stillbirth, preterm birth	Betel nut, smoking, alcohol use, low mid upper arm circumference (MUAC), maternal height, malaria prophylaxis, primigravida and fewer antenatal visits
Stanisic et al.52	2015	PNG
Madang
Province	To identify risk factors for LBW, anaemia, and preterm in pregnant women of PNG.	Prospective cohort study	328	LBW, preterm birth	Malaria infection, parasitemia, severe anaemia, multigravida, tobacco smoke, education and MUAC
Unger et al.53	2015	PNG Madang Province	To evaluate factors with FGR among pregnant woman enrolled in randomised control trial evaluating IPTp malaria prophylaxis in pregnancy.	Cohort study with clinical trial	671	LBW, SGA, preterm birth low weight gain	Low MAUC (<22 cm), short stature (<150 cm), low BMI, anaemia (<90 g/l)⁎⁎, malaria infections, plasmodium falciparum malaria, plasmodium vivax malaria (peripheral or Sub-microscopic infections)
Senn et al.51	2009	PNG Madang Province	To investigates the habits of betel nut chewing and possible impact on pregnancy.	Cross-sectional study	310	Reduced mean birth weight	Primigravity, betel nut, smoke, alcohol, and low BMI
Peters et al.50	2001	PNG	To identify risk factors associated with LBW in the Western highland Province of PNG.	Prospective case-control	299	LBW	Maternal age, birth interval, no antenatal bookings, low antenatal attendance, fever during Pregnancy, pre-eclampsia, antepartum haemorrhage, maternal smoking, short stature, Low BMI, low haemoglobin, low education
(ii)Nine grey literatures
National Statistics Office PNG21	2019	PNG	To provide information indicators of fertility, fertility preferences, family planning practices, childhood mortality, maternal and child health, knowledge, and awareness of human immunodeficiency virus- acquired immunodeficiency syndrome (HIV/AIDS), domestic violence, and other related health issues.	Demographic health survey	4672	LBW, Smaller than average babies	Not applicable⁎⁎⁎
National Statistic Office SI61	2015	Solomon Islands	To provide current and reliable data on fertility and family planning behavior, child mortality, adult and maternal mortality, children's nutritional status, the use of maternal and child healthcare services, knowledge of HIV and AIDS, and other health-related issues.	Demographic health survey	3535	LBW, Smaller than average baby, very small baby	Not applicable
Bureau of Statistics56	2014	Samoa	To provide information for policymakers, planners, researchers, and program managers, for use in planning, implementing, monitoring, and evaluating population and health programs within the country.	Demographic health survey	3192	LBW, Smaller than average baby, very small baby	Not applicable
Ministry of Health VNSO59	2013	Vanuatu	To furnish policymakers and planners with detailed information on fertility, family planning, infant and child mortality, maternal and child health and nutrition, and knowledge of HIV and AIDS and other sexually transmitted infections.	Demographic health survey	1562	LBW	Not applicable
Ministry of Health TDoS20	2012	Tonga	To ensure better understanding and use of these data and widely dissemination of results at different planning levels.	Demographic health survey	1703	LBW, Smaller than average baby, very small baby	Not applicable
National Statistic Office KaSotPC60	2009	Kiribati	To provide information for policymakers, planners, researchers, and program managers, for use in planning, implementing, monitoring, and evaluating population and health programs in the country.	Demographic health survey	1099	LBW, very small baby	Not applicable
Bureau of Statistics NatSotPC55	2007	Nauru	To provide information for policymakers, planners, researchers, and program managers, for use in planning, implementing, monitoring, and evaluating population and health programs in the country.	Demographic health survey	322	LBW, very small baby	Not applicable
Central Statistics Division57	2007	Tuvalu	To provide information for policymakers, planners, researchers, and program managers, for use in planning, implementing, monitoring, and evaluating population and health programs in the country.	Demographic health survey	447	LBW, Smaller than average baby, very small baby	Not applicable
Economic Policy58	2007	Marshall Islands	To provide information for policymakers, planners, researchers, and program managers for use planning, implementation, monitoring and evaluation of population and health programs in the country.	Demographic health survey	1173	LBW, Smaller than average baby, very small baby	Not applicable

Notes:

grams per decilitre.

grams per litre.

Not applicable; these surveys did not present measure of associations between the independent and dependent variables.

Prevalence of adverse birth outcomes

Prevalence of the various adverse birth outcomes was reported for 11 countries in the region (Supplementary File 3). Nineteen studies,,44, 45, 46, 47, 48, 49, 50,52, 53, 54, 55, 56, 57, 58, 59, 60, 61 representing 11 countries reported the prevalence of LBW which ranged from 3% in CNMI to 27% in Nauru with a mean of 12%. The Polynesian and Micronesian countries such as CNMI, Palau, Samoa, Tonga and Tuvalu reported LBW prevalence of less than 10%. Six studies44, 45, 46, 47,, reported the prevalence of preterm births which ranged from 7% in CNMI to 24% in the Solomon Islands with estimated mean of 13%. The Palau study reported a prevalence for preterm LBW of 5%, term LBW of 4% and a prevalence for preterm normal weight of 4%.

Risk factors for adverse birth outcomes

Low birth weight

Six studies, five from PNG,50, 51, 52, and one from Palau, reported the risk factors for LBW. LBW was associated with malaria infection in pregnancy.,, These studies investigated acute and chronic, and microscopic and sub-microscopic malaria infections, which were all associated with an increased risk of LBW.,, Southeast Asian ovalocytosis infection in pregnancy was associated with LBW, as was betel nut and tobacco use in pregnancy., Low maternal mid-upper arm circumference (<23 cm) and short stature (<150 cm) were also associated with LBW. Young and advanced maternal ages (< 22 years and > 35 years) were associated with LBW. Other risk factors associated with LBW were birth intervals less than 2 years, no antenatal booking, fever during pregnancy, and female infant and primigravida (Supplementary file 5).

Preterm birth

Six studies reported risk factors for preterm births from PNG,,,, CNMI, and Palau.

Three studies reported that malaria infections during pregnancy were associated with preterm birth.,, Two of the studies showed that acute and chronic placental malaria were associated with preterm birth, with chronic infection showing a stronger association., Microscopic malaria infections were also associated with preterm birth. Three studies reported preterm birth were associated with demographic, health risk factors such as maternal obesity, maternal age (age <20 years and ≥ than 35 years) and insufficient antenatal care (1–8 visits) in the CNMIs. Pacific Islands, Filipino women and CNMI women living in CNMI were associated with higher risk of preterm birth than women in CNMI of Chinese descent (Supplementary file 6).

Small for gestational age

Only one study showed the risk factors associated with SGA. This study reported that low maternal haemoglobin level (Hb<9 mg/L) during pregnancy, maternal upper arm circumference (<22 cm), short maternal stature (<150 cm) and primigravida were associated with SGA (Supplementary File 7).

Changes in mean birth weight

Six studies47, 48, 49,51, 52, 53 reported associations with mean birth weight. Acute placental malaria during pregnancy was associated with reductions in mean birth weight. Primigravity and tobacco smoking during pregnancy were also reported as factors that were associated with reductions in mean birth weight (Supplementary File 8).

Stillbirth and miscarriage

Only one study investigated associations with stillbirth and miscarriage. In this study, there was marginal evidence for an association between heavy betel nut use and stillbirth and miscarriage (Supplementary file 7).

Health professionals

Demographic profile

We interviewed 18 health professionals from the Pacific Island countries of the Solomon Islands (n = 11), PNG (n = 3), Fiji (n = 2) and Vanuatu (n =  2). They were midwives (n = 7), neonatal care nurses (n = 5), obstetricians (n = 3), registered nurses (n = 2), and a paediatrician (n = 1) practising in major tertiary hospitals (n = 15) and rural community practice (n = 3). Eleven health professionals had qualifications equivalent to a master's degree, post-graduate or above, while seven had an undergraduate degree or diploma. The health professional mean age was 42 years, and their average work experience was 17 years. The data provided by the health professionals were quantified and categorised and presented under the following themes: (a) adverse birth outcomes and risk factors; (b) mitigating risk factors during pre-pregnancy; and (c) mitigating risk factors during pregnancy.

Adverse birth outcomes and risk factors. The most frequent adverse birth outcomes reported by the health professionals were preterm birth (n = 14), followed by LBW (n = 12) and small for gestational age (n = 11). The most frequently nominated risk factors for adverse birth outcomes were physical and emotional stress (n = 13), teenage pregnancy (n = 11) malaria (n = 11), and poverty (n =11) (Supplementary file 9).

Mitigating of risk factors during pre-pregnancy

Most health professionals (n = 13) reported the risk of adverse birth outcomes could be reduced during pre-pregnancy. They believed that actions to reduce the risk of adverse birth outcomes could be developing pre-pregnancy health care policy, providing holistic health care, effective health education for women and girls, and incorporating sexual and reproductive health into the school curriculum. Health education orientated to informing women of the importance of, family planning, antenatal care, good nutrition, substance use avoidance, and raising awareness on sexually transmitted infections (STIs) would be beneficial.

“Educate young women and girls on good care before pregnancy… collaborate and implement program on sexual and reproductive health on safe sex practice… establish community participation with churches and social organisations” (Midwife, PNG).

“Health education is essential for mothers before pregnancy, on good nutrition, danger signs, early antenatal care, personal hygiene, prophylaxis and delivery at the health facilities” (Midwife, Vanuatu)

Health professionals also suggested that the current health system should focus on health promotion. This was reflected when a midwife from Solomon Islands said:

“The health system should also focus on public health and care for women during pre-pregnancy… pre-pregnancy health care is the public health profession's role and should be integrated with clinical care” (Midwife, Solomon Islands)

“We should have a special clinic for women to seek advice if they [women] planning to get pregnant” (Neonatal care nurse, Solomon Islands)

Mitigating risk factors during pregnancy

Health providers reflected on how risk factors leading to adverse birth outcomes could be mitigated at the time of pregnancy. Most health professionals (n = 13) called for comprehensive and ethical antenatal care involving health assessments, physical examinations, provisions of screening for STIs and malaria and provision of malaria prophylaxis and drug supplements (e.g., iron tablets and folic acid).

“We must monitor vital signs such as blood pressure, fetal heart rate, glucose, and protein to detect any underlying health condition… we monitor mothers’ weight, no weight gain would mean growth restriction [to the fetus], if the weight increases abnormally, it could mean gestational diabetes” (Midwife, Vanuatu)

“Screen all antenatal mothers for syphilis, malaria and anaemia and treat them early” (Midwife, Solomon Islands)

“….provide malaria prophylaxis, ferrous sulphate with folic acid, albendazole and tetanus vaccine to prevent illness during pregnancy (neonatal nurse, Solomon Island)

Health professionals also suggested the need to improve antenatal health services in rural areas, improve health professionals’ practices, and the need for the provision of health education during antenatal care in local dialects as a means of reducing exposure to risk factors and adverse birth outcomes.

“….problems with birth outcomes occur mostly in rural area… therefore the government should focus on rural health” (Midwife, PNG)

“Trained midwives should see the pregnant woman not as a task but as a whole person” (Midwife, Solomon Islands)

“Use brochures, posters, and pamphlets in local dialects… where lacking in staff, use of pre-recorded speech and videos in antenatal education” (Neonatal care nurse, Vanuatu).

Discussion

Summary of findings from this scoping review

To our knowledge, this scoping review is the first to examine the available evidence on the burden of adverse birth outcomes and their risk factors in the Pacific Island region. Our review of 22 countries and territories located in the region found a dearth of published peer reviewed studies (n = 11) and grey literature (n = 9) that met our inclusion criteria. Studies that did not follow standard reporting guidelines were excluded. Nonetheless, the included studies were informative in (i) identifying the adverse birth outcomes most and least well-studied; (ii) providing an indication of the expected burden of adverse birth outcomes; and (iii) identifying the range of risk factors that have been investigated in the Pacific Island region. In summary, we identified that (i) LBW and preterm birth were the most well-studied outcomes but there were relatively few studies conducted on mortality endpoints; (ii) the prevalence of the adverse birth outcomes varied considerably both within and between regions, and very high for some locations such as the Solomon Islands; and that (iii) although health, obstetric, social and behavioural risk factors have been investigated, several, such as betel nut chewing, are relatively unique to the region and require further investigation. Outside PNG few studies have been conducted.

Prevalence of adverse birth outcomes in the Pacific Island region

There are several plausible explanations for the wide range for prevalence of LBW. Firstly, relatively lower prevalence of LBW and preterm birth in the Polynesia and Micronesia may reflect better maternal health, higher antenatal coverage and greater likelihood of presence of a skilled birth attendant.18, 19, 20 The estimated mean LBW prevalence (12%) in the Pacific island countries is comparable to those from LMICs of Africa and Asia. The mean estimated prevalence for preterm birth was 13% is comparable to WHO estimates of 14% for Pacific region, which is also similar to estimated prevalence for preterm birth (10–15%) in LMIC by WHO. The true prevalence of preterm birth and LBW were hampered by lack of routine accurate records of pregnancy dates, underreporting of stillbirths, lack of availability of measurement instruments and small sample sizes. These two adverse birth outcomes were both the most investigated and identified by the health professionals as those most frequently occurring of concern in the region.

Risk factors for adverse birth outcomes in the Pacific Island region

Malaria

Malaria infections reported in PNG showed significant associations with LBW, preterm births, and changes in mean birth weights.,, Malaria infection was also reported as a major cause of preterm birth, LBW, miscarriage, and stillbirth by the consulted health professionals. Malaria is endemic in all coastal areas of the three Melanesian countries of PNG, Solomon Islands and Vanuatu64, 65, 66, 67 with PNG as the epicentre of malaria infection in the Western Pacific region., Malaria is also endemic in Vanuatu but infection fluctuates across the different seasons, while most countries of Polynesia and Micronesia have been declared as malaria-free by WHO. Successful provision of malaria prophylaxis to pregnant women in the region has been a challenge due to drug shortages and poor access to woman in isolated rural areas., We conclude that malaria is a risk factor in pregnancy and adverse birth outcomes in the malaria-endemic areas of the Pacific Island region, although the magnitude of the association remains unclear.

Anaemia and worm infestation

Health professionals interviewed nominated anaemia and worm infestation to be associated with the adverse birth outcomes of which both conditions were prevalent in the region., Oral iron supplements and albendazole stat doses during pregnancy are part of the antenatal care protocol in most Pacific Island countries. However, accessing these medications can be challenging due to consistent drug shortages, poor access to women in rural and isolated areas, and low antenatal care attendance resulting in a significant portion of pregnant women not taking these medications. The health professionals called for improved antenatal care and health education on prophylaxis and supplements during pre-pregnancy and pregnancy, at all levels, including public health and social organisations, and churches as an innovative mitigation strategy.

Substance use

Betel nut and tobacco use in PNG and Palau were associated with LBW,, birth weight reduction, stillbirth and miscarriage. Studies from Southeast Asia have also reported betel nut chewing during pregnancy as having an impact on birth weight, fetal length and preterm birth.,, The magnitude of the effects of betel nut chewing on birth outcomes remains unclear. Reductions in birth weight from betel nut use appear to be at least as large as those reported for smoking during pregnancy. The health professionals that we consulted also perceived betel nut, tobacco and kava use as having a negative impact on pregnancy and birth outcomes. Although the impact of kava use has been rarely studied in the peer-reviewed literature, health professionals reported its detrimental effects on pregnancy outcomes. Given the increase in kava use in the region, there is a need for future investigation on its impact on births and pregnancies.

Maternal obesity and obesity-related conditions

A study in Palau reported maternal obesity to be positively associated with preterm birth. This finding was also supported by the interviewed health professionals who reported obesity-related diseases such as: type 2 diabetes, gestational diabetes, hypertension, and pregnancy-induced hypertension as risk factors for adverse birth outcomes. The health professionals also reported pre-eclampsia, a hypertensive disorder of pregnancy, as a notable cause of LBW and preterm birth. We conclude that obesity, which occurs among 36% of women in the Pacific Island region,, is likely to have a role in maternal health during pregnancy, whether or not it is a direct cause of adverse birth outcomes in the region.

Poor antenatal care and access

Our literature review found no antenatal bookings and lower number of antenatal visits was associated with LBW and preterm birth. This was also reflected in the health professionals concerns about poor presentation for antenatal care, influenced by poor access to health facilities and health seeking behaviours. Health professionals were concerned that shortage of staff and lack of health facilities especially in rural areas have contributed to lack of quality in antenatal care. Our findings were also validated by recent studies in the region which showed poor quality maternal care in the region leads to poor outcomes in newborns.,

Quality of reporting and number of studies in the Pacific Island region

Our review found a dearth of studies for all countries, for all risk factors and for all birth outcomes that have been investigated to date in the Pacific Islands region. Of the all 20 studies representing the 11 Pacific Island countries (CNMI, Kiribati, Marshall Islands, Nauru, Palau, PNG, Samoa, Solomon Islands, Tonga, Tuvalu, Vanuatu), the measure of association for risk factors and adverse birth outcomes were reported in ten studies, which came from only three countries (PNG, Palau, and CNMI).44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54 Also, the association of risk factors and adverse birth outcomes was reported by nine population-based cohort studies,,,,49, 50, 51, 52, 53, 54 and a randomised control trial of which four had small study populations.,50, 51, 52 Two DHS reports from Nauru and Tuvalu also used small samples to produce prevalence estimates. Many studies did not report measures of uncertainty (confidence intervals or variance), control for confounding or attempt to identify potential sources of bias. This assessment stems from limited research capacity in the region,, and limited resources for health administration. Consultation with health professionals was essential, assisting in illuminating the findings from the literature, contributing additional insights and recommendations to reduce exposure to risk factors and adverse birth outcomes in the Pacific region.

Limitations of this scoping review

We could not access health reports and grey literature that were only published in hardcopy. It is plausible that many health systems in the region do not disseminate health information online. We were also unable to obtain the views of health professionals from the Polynesian and Micronesian regions

Conclusion

Despite the limited studies, the review identified a range of risk factors and adverse birth outcomes relevant to the Pacific Island region. The prevalence of adverse birth outcomes is not well-ascertained for countries within the region, although LBW and preterm birth were relatively more frequently investigated. There were a range of risk factors for adverse birth outcomes - such as, malaria, substance use, obesity and poor antenatal care – that are either relatively prevalent or somewhat unique to the Pacific Islands region. These risk factors have not been comprehensively investigated in the region and were confirmed by health professionals as risk factors of concern.

Contributors

LSKK, GP, GT and JJ conceived and conceptualised study design. LSKK conducted searches and initial study selection. LSKK, GP, GT and JJ did the final study selection. LSKK conducted study analysis under guidance of GP and GT. LSKK and JJ designed stakeholder consultation study, LSKK conduct interviews and data analysis. LSKK wrote first draft and reviewed by GP, GAT and JJ. HB and GD read and approved the final paper.

Supplemental Material: Supplemental material for this article is available online.

Funding

GP was supported with funding from the National Health and Medical Research Council Project and Investigator Grants #1099655 and #1173991, institutional funding for the WA Health and Artificial Intelligence Consortium, and the Research Council of Norway through its Centres of Excellence funding scheme #262700. GAT was supported with funding from the National Health and Medical Research Council Investigator Grant #1195716. LSSK is recipient of scholarship from the Department of Foreign Affairs, Australia Award.

Declaration of interests

The authors declared no potential conflicts of interest.