1
|
Wolfson C, Qian J, Creanga AA. Levels, Trends, and Risk Factors for Stillbirths in the United States: 2000-2017. Am J Perinatol 2024; 41:e601-e611. [PMID: 35973798 DOI: 10.1055/a-1925-2131] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 11/01/2022]
Abstract
OBJECTIVE This study documents 2000 to 2017 trends in stillbirth rates and changes in associations between known maternal and fetal risk factors and stillbirths for 2000 to 2002 versus 2015 to 2017 in the United States. STUDY DESIGN We conducted a retrospective, population-based analysis of stillbirths and live-births using national vital statistics data. We calculated annual stillbirth rates overall and by gestational age; and examined stillbirth rates by maternal age, race-ethnicity, and state for 2000 to 2002 versus 2015 to 2017. We used Chi-squared tests to examine associations between maternal and fetal risk factors separately for early (20-27 weeks) and late (28+ weeks) stillbirths compared with live-births for 2000 to 2002 versus 2015 to 2017. RESULTS Stillbirth rates declined by 7.5% (p < 0.001) during 2000 to 2006 but remained flat at approximately 6 stillbirths per 1,000 births thereafter. Throughout 2000 to 2017, there were significant improvements in stillbirth rates at 39+ weeks nationally (p < 0.001), but rates varied greatly between and within states. Sociodemographic (advanced maternal age, Black race, low education, unmarried status, and rural residence), obstetric, and other medical factors (>3 births, use of infertility treatment, maternal obesity, diabetes, chronic hypertension, eclampsia, no prenatal care, and tobacco use) were significantly more prevalent in women with late than early stillbirths or live births. Notably, late and total stillbirth rates were approximately 30% higher for women >35 years than for women <35 years and twice as high for non-Hispanic Black than non-Hispanic White women; American Indian/Alaska Native women represented the only racial-ethnic group with significantly higher late stillbirth rates in 2015 to 2017 than in 2000 to 2002. Pregnancy and fetal factors (multiple pregnancy, male fetus, and breech presentation) were more prevalent in women with early than late stillbirths or live births. CONCLUSION U.S. stillbirth rates have plateaued since 2006. There are persistent differential risk profiles for early versus late stillbirths which can inform stillbirth prevention strategies (e.g., close observation of women with risk factors for stillbirth) and new research into the causes of stillbirths by gestational age. KEY POINTS · U.S. stillbirth rates have plateaued since 2006.. · Stillbirth rates vary between and within U.S. states and by maternal and fetal factors.. · Early versus late stillbirths have different risk profiles which can guide stillbirth prevention strategies..
Collapse
Affiliation(s)
- Carrie Wolfson
- Department of International Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland
| | - Jiage Qian
- Department of International Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland
- Department of Population, Family and Reproductive Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland
| | - Andreea A Creanga
- Department of International Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland
- Department of Population, Family and Reproductive Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland
- Department of Gynecology and Obstetrics, Johns Hopkins School of Medicine, Baltimore, Maryland
| |
Collapse
|
2
|
Mukherjee A, Di Stefano L, Blencowe H, Mee P. Determinants of stillbirths in sub-Saharan Africa: A systematic review. BJOG 2024; 131:140-150. [PMID: 37272228 DOI: 10.1111/1471-0528.17562] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/02/2022] [Revised: 04/07/2023] [Accepted: 05/13/2023] [Indexed: 06/06/2023]
Abstract
BACKGROUND Sub-Saharan African (SSA) countries have high stillbirth rates compared with high-income countries, yet research on risk factors for stillbirth in SSA remain scant. OBJECTIVES To identify the modifiable risk factors of stillbirths in SSA and investigate their strength of association using a systematic review. SEARCH STRATEGY CINAHL Plus, EMBASE, Global Health and MEDLINE databases were searched for literature. SELECTION CRITERIA Observational population- and facility-level studies exploring stillbirth risk factors, published in 2013-2019 were included. DATA COLLECTION AND ANALYSIS A narrative synthesis of data was undertaken and the potential risk factors were classified into subgroups. MAIN RESULTS Thirty-seven studies were included, encompassing 20 264 stillbirths. The risk factors were categorised as: maternal antepartum factors (0-4 antenatal care visits, multiple gestations, hypertension, birth interval of >3 years, history of perinatal death); socio-economic factors (maternal lower wealth index and basic education, advanced maternal age, grand multiparity of ≥5); intrapartum factors (direct obstetric complication); fetal factors (low birthweight and gestational age of <37 weeks) and health systems factors (poor quality of antenatal care, emergency referrals, ill-equipped facility). The proportion of unexplained stillbirths remained very high. No association was found between stillbirths and body mass index, diabetes, distance from the facility or HIV. CONCLUSIONS The overall quality of evidence was low, as many studies were facility based and did not adjust for confounding factors. This review identified preventable risk factors for stillbirth. Focused programmatic strategies to improve antenatal care, emergency obstetric care, maternal perinatal education, referral and outreach systems, and birth attendant training should be developed. More population-based, high-quality research is needed.
Collapse
Affiliation(s)
- Ankita Mukherjee
- Faculty of Public Health and Policy, London School of Hygiene and Tropical Medicine, London, UK
- Monitoring, Evaluation and Research, New Delhi, India
| | | | - Hannah Blencowe
- Department of Infectious Disease Epidemiology, London School of Hygiene and Tropical Medicine, London, UK
| | - Paul Mee
- Department of Infectious Disease Epidemiology, London School of Hygiene and Tropical Medicine, London, UK
- Lincoln International Institute for Rural Health, College of Social Science, University of Lincoln, Lincoln, UK
| |
Collapse
|
3
|
Konstantyner T, Areco KCN, Bandiera-Paiva P, Marinonio ASS, Kawakami MD, Balda RDCX, Miyoshi MH, Sanudo A, Costa-Nobre DT, de Freitas RMV, Morais LCC, Teixeira MLP, Waldvogel BC, Kiffer CRV, de Almeida MFB, Guinsburg R. The burden of inappropriate birth weight on neonatal survival in term newborns: a population-based study in a middle-income setting. Front Pediatr 2023; 11:1147496. [PMID: 37360363 PMCID: PMC10285294 DOI: 10.3389/fped.2023.1147496] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 02/07/2023] [Accepted: 05/10/2023] [Indexed: 06/28/2023] Open
Abstract
Introduction Premature birth, perinatal asphyxia, and infections are the main causes of neonatal death. Growth deviations at birth also affect neonatal survival according to week of gestation at birth, particularly in developing countries. The purpose of this study was to verify the association between inappropriate birth weight and neonatal death in term live births. Methods This is an observational follow-up study with all term live births from 2004 to 2013 in Sao Paulo State, Brazil. Data were retrieved with the deterministic linkage of death and birth certificates. The definition of very small for gestational age (VSGA) and very large for gestational age (VLGA) used the 10th percentile of 37 weeks and the 90th percentile of 41 weeks + 6 days, respectively, based on the Intergrowth-21st. We measured the outcome in terms of time to death and the status of each subject (death or censorship) in the neonatal period (0-27 days). Survival functions were calculated using the Kaplan-Meier method stratified according to the adequacy of birth weight into three groups (normal, very small, or very large). We used multivariate Cox regression to adjust for proportional hazard ratios (HRs). Results The neonatal death rate during the study period was 12.03/10,000 live births. We found 1.8% newborns with VSGA and 2.7% with VLGA. The adjusted analysis showed a significant increase in mortality risk for VSGA infants (HR = 4.25; 95% CI: 3.89-4.65), independent of sex, 1-min Apgar score, and five maternal factors. Discussion The risk of neonatal death in full-term live births was approximately four times greater in those with birth weight restriction. The development of strategies to control the factors that determine fetal growth restriction through planned and structured prenatal care can substantially reduce the risk of neonatal death in full-term live births, especially in developing countries such as Brazil.
Collapse
Affiliation(s)
- Tulio Konstantyner
- Departamento de Pediatria, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, São Paulo, Brazil
| | - Kelsy Catherina Nema Areco
- Departamento de Pediatria, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, São Paulo, Brazil
- Departamento de Informática em Saúde, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, São Paulo, Brazil
| | - Paulo Bandiera-Paiva
- Departamento de Informática em Saúde, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, São Paulo, Brazil
| | | | - Mandira Daripa Kawakami
- Departamento de Pediatria, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, São Paulo, Brazil
| | - Rita de Cássia Xavier Balda
- Departamento de Pediatria, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, São Paulo, Brazil
| | - Milton Harumi Miyoshi
- Departamento de Pediatria, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, São Paulo, Brazil
| | - Adriana Sanudo
- Departamento de Medicina Preventiva, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, São Paulo, Brazil
| | - Daniela Testoni Costa-Nobre
- Departamento de Pediatria, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, São Paulo, Brazil
| | - Rosa Maria Vieira de Freitas
- Diretoria Adjunta de Produção e Análise de Dados, Fundação Sistema Estadual de Análise de Dados, São Paulo, São Paulo, Brazil
| | - Liliam Cristina Correia Morais
- Diretoria Adjunta de Produção e Análise de Dados, Fundação Sistema Estadual de Análise de Dados, São Paulo, São Paulo, Brazil
| | - Monica La Porte Teixeira
- Diretoria Adjunta de Produção e Análise de Dados, Fundação Sistema Estadual de Análise de Dados, São Paulo, São Paulo, Brazil
| | - Bernadette Cunha Waldvogel
- Diretoria Adjunta de Produção e Análise de Dados, Fundação Sistema Estadual de Análise de Dados, São Paulo, São Paulo, Brazil
| | - Carlos Roberto Veiga Kiffer
- Departamento de Medicina, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, São Paulo, Brazil
| | | | - Ruth Guinsburg
- Departamento de Pediatria, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, São Paulo, Brazil
| |
Collapse
|
4
|
Dube K, Lavender T, Blaikie K, Sutton CJ, Heazell AEP, Smyth RMD. Identification of factors associated with stillbirth in Zimbabwe - a cross sectional study. BMC Pregnancy Childbirth 2021; 21:662. [PMID: 34587922 PMCID: PMC8482658 DOI: 10.1186/s12884-021-04102-y] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/15/2021] [Accepted: 08/29/2021] [Indexed: 11/10/2022] Open
Abstract
INTRODUCTION 98% of the 2.6 million stillbirths per annum occur in low and middle income countries. However, understanding of risk factors for stillbirth in these settings is incomplete, hampering efforts to develop effective strategies to prevent deaths. METHODS A cross-sectional study of eligible women on the postnatal ward at Mpilo Hospital, Zimbabwe was undertaken between 01/08/2018 and 31/03/2019 (n = 1779). Data were collected from birth records for maternal characteristics, obstetric and past medical history, antenatal care and pregnancy outcome. A directed acyclic graph was constructed with multivariable logistic regression performed to fit the corresponding model specification to data comprising singleton pregnancies, excluding neonatal deaths (n = 1734), using multiple imputation for missing data. Where possible, findings were validated against all women with births recorded in the hospital birth register (n = 1847). RESULTS Risk factors for stillbirth included: previous stillbirth (29/1691 (2%) of livebirths and 39/43 (91%) of stillbirths, adjusted Odds Ratio (aOR) 2628.9, 95% CI 342.8 to 20,163.0), antenatal care (aOR 44.49 no antenatal care vs. > 4 antenatal care visits, 95% CI 6.80 to 291.19), maternal medical complications (aOR 7.33, 95% CI 1.99 to 26.92) and season of birth (Cold season vs. Mild aOR 14.29, 95% CI 3.09 to 66.08; Hot season vs. Mild aOR 3.39, 95% CI 0.86 to 13.27). Women who had recurrent stillbirth had a lower educational and health status (18.2% had no education vs. 10.0%) and were less likely to receive antenatal care (20.5% had no antenatal care vs. 6.6%) than women without recurrent stillbirth. CONCLUSION The increased risk in women who have a history of stillbirth is a novel finding in Low and Middle Income Countries (LMICs) and is in agreement with findings from High Income Countries (HICs), although the estimated effect size is much greater (OR in HICs ~ 5). Developing antenatal care for this group of women offers an important opportunity for stillbirth prevention.
Collapse
Affiliation(s)
- Kushupika Dube
- Mpilo School of Midwifery, PO Box 2096, Vera Road, Bulawayo, Zimbabwe.
| | - Tina Lavender
- Centre for Childbirth, Women's and Newborn Health, Liverpool School of Tropical Medicine, Liverpool, UK
| | - Kieran Blaikie
- Centre for Biostatistics, School of Health Sciences, University of Manchester, Manchester, UK
| | - Christopher J Sutton
- Centre for Biostatistics, School of Health Sciences, University of Manchester, Manchester, UK
| | - Alexander E P Heazell
- Maternal and Fetal Health Research Centre, School of Medical Sciences, Faculty of Biological, Medical and Human Sciences, University of Manchester, Manchester, UK.,National University of Science & Technology, P. O. Box AC 939, Ascot, Bulawayo, Zimbabwe
| | - Rebecca M D Smyth
- Division of Nursing, Midwifery and Social Work, School of Health Sciences, The University of Manchester, Manchester, UK
| |
Collapse
|
5
|
Ssegujja E, Ddumba I, Andipartin M. Prioritization of interventions in pursuit of maternal health policy objectives to mitigate stillbirth risks. An exploratory qualitative study at subnational level in Uganda. BMC Health Serv Res 2021; 21:53. [PMID: 33430858 PMCID: PMC7802206 DOI: 10.1186/s12913-020-06046-z] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/19/2020] [Accepted: 12/26/2020] [Indexed: 12/04/2022] Open
Abstract
Background Global calls for renewed efforts to address stillbirth burden highlighted areas for policy and implementation resulting in national level translations. Information regarding adapted strategies to effect policy objectives into service delivery by frontline health workers remains scanty especially at subnational level. The study explored strategies prioritized to mitigate stillbirth risk in the context of operationalizing recommendations from the global campaigns at a subnational level in Uganda. Methods A cross-sectional qualitative exploratory study was conducted among a purposively selected sample of sixteen key informants involved in delivery of maternal and child health services in Mukono district. Analysis followed thematic content analysis deductively focusing on those policy priorities highlighted in the global stillbirth campaigns and reflected at the national level in the different guidelines. Results. Interventions to address stillbirth followed prioritization of service delivery aspects to respond to identified gaps. Efforts to increase uptake of family planning services for example included offering it at all entry points into care with counseling forming part of the package following stillbirth. Referrals were streamlined by focusing on addressing delays from the referring entity while antenatal care attendance was boosted through provision of incentives to encourage mothers to comply. Other prioritized aspects included perinatal death audits and improvements in data systems while differentiated care focused on aligning resources to support high risk mothers. This was in part influenced by the limited resources and skills which made health workers to adapt routine to fit implementation context. Conclusions The resource availability determined aspects of policy to prioritize while responding to stillbirth risk at subnational level by frontline health workers. Their understanding of risk, feasibility of implementation and the desire for optimal health systems performance worked to define the nature of services delivered calling for purposeful consideration of resource availability and implementation context while prioritizing stillbirth reduction at subnational level. Supplementary Information The online version contains supplementary material available at 10.1186/s12913-020-06046-z.
Collapse
Affiliation(s)
- Eric Ssegujja
- Makerere University School of Public Health, Kampala, Uganda. .,University of the Western Cape, School of Public Health, Cape Town, South Africa.
| | - Isaac Ddumba
- Mukono District Local Government, Mukono Town, Uganda
| | - Michelle Andipartin
- Department of Psychology, University of the Western Cape, Cape Town, South Africa
| |
Collapse
|
6
|
Nove A, Friberg IK, de Bernis L, McConville F, Moran AC, Najjemba M, Ten Hoope-Bender P, Tracy S, Homer CSE. Potential impact of midwives in preventing and reducing maternal and neonatal mortality and stillbirths: a Lives Saved Tool modelling study. Lancet Glob Health 2021; 9:e24-e32. [PMID: 33275948 PMCID: PMC7758876 DOI: 10.1016/s2214-109x(20)30397-1] [Citation(s) in RCA: 129] [Impact Index Per Article: 43.0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/03/2020] [Revised: 08/27/2020] [Accepted: 09/01/2020] [Indexed: 12/21/2022]
Abstract
BACKGROUND Strengthening the capacity of midwives to deliver high-quality maternal and newborn health services has been highlighted as a priority by global health organisations. To support low-income and middle-income countries (LMICs) in their decisions about investments in health, we aimed to estimate the potential impact of midwives on reducing maternal and neonatal deaths and stillbirths under several intervention coverage scenarios. METHODS For this modelling study, we used the Lives Saved Tool to estimate the number of deaths that would be averted by 2035, if coverage of health interventions that can be delivered by professional midwives were scaled up in 88 countries that account for the vast majority of the world's maternal and neonatal deaths and stillbirths. We used four scenarios to assess the effects of increasing the coverage of midwife-delivered interventions by a modest amount (10% every 5 years), a substantial amount (25% every 5 years), and the amount needed to reach universal coverage of these interventions (ie, to 95%); and the effects of coverage attrition (a 2% decrease every 5 years). We grouped countries in three equal-sized groups according to their Human Development Index. Group A included the 30 countries with the lowest HDI, group B included 29 low-to-medium HDI countries, and group C included 29 medium-to-high HDI countries. FINDINGS We estimated that, relative to current coverage, a substantial increase in coverage of midwife-delivered interventions could avert 41% of maternal deaths, 39% of neonatal deaths, and 26% of stillbirths, equating to 2·2 million deaths averted per year by 2035. Even a modest increase in coverage of midwife-delivered interventions could avert 22% of maternal deaths, 23% of neonatal deaths, and 14% of stillbirths, equating to 1·3 million deaths averted per year by 2035. Relative to current coverage, universal coverage of midwife-delivered interventions would avert 67% of maternal deaths, 64% of neonatal deaths, and 65% of stillbirths, allowing 4·3 million lives to be saved annually by 2035. These deaths averted would be particularly concentrated in the group B countries, which currently account for a large proportion of the world's population and have high mortality rates compared with group C. INTERPRETATION Midwives can help to substantially reduce maternal and neonatal mortality and stillbirths in LMICs. However, to realise this potential, midwives need to have skills and competencies in line with recommendations from the International Confederation of Midwives, to be part of a team of sufficient size and skill, and to work in an enabling environment. Our study highlights the potential of midwives but there are many challenges to the achievement of this potential. If increased coverage of midwife-delivered interventions can be achieved, health systems will be better able to provide effective coverage of essential sexual, reproductive, maternal, newborn, and adolescent health interventions. FUNDING New Venture Fund.
Collapse
Affiliation(s)
| | | | | | - Fran McConville
- Department of Maternal, Newborn, Child and Adolescent Health and Ageing, World Health Organization, Geneva, Switzerland
| | - Allisyn C Moran
- Department of Maternal, Newborn, Child and Adolescent Health and Ageing, World Health Organization, Geneva, Switzerland
| | - Maria Najjemba
- Uganda Country Office, United Nations Population Fund, Kampala, Uganda
| | | | - Sally Tracy
- Faculty of Medicine and Health, University of Sydney, Sydney, NSW, Australia
| | - Caroline S E Homer
- Maternal, Child and Adolescent Health, Burnet Institute, Melbourne, VIC, Australia
| |
Collapse
|
7
|
Trinh T, Leal AF, Mello MB, Taylor MM, Barrow R, Wi TE, Kamb ML. Syphilis management in pregnancy: a review of guideline recommendations from countries around the world. Sex Reprod Health Matters 2020; 27:69-82. [PMID: 31884900 PMCID: PMC7888020 DOI: 10.1080/26410397.2019.1691897] [Citation(s) in RCA: 7] [Impact Index Per Article: 1.8] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/19/2022] Open
Abstract
Guidelines can help healthcare practitioners manage syphilis in pregnancy and prevent perinatal death or disability. We conducted systematic reviews to locate guidance documents describing management of syphilis in pregnancy, 2003–2017. We compared country and regional guidelines with current World Health Organization (WHO) guidelines. We found 64 guidelines with recommendations on management of syphilis in pregnancy representing 128 of the 195 WHO member countries, including the two WHO guidelines published in 2016 and 2017. Of the 62 guidelines, 16 were for countries in Africa, 21 for the Americas, two for Eastern Mediterranean, six for Europe and 17 for Asia or the Pacific. Fifty-seven (92%) guidelines recommended universal syphilis screening in pregnancy, of which 46 (81%) recommended testing at the first antenatal care visit. Also, 46 (81%) recommended repeat testing including 21 guidelines recommended this during the third pregnancy trimester and/or at delivery. Fifty-nine (95%) guidelines recommended benzathine penicillin G (BPG) as the first-line therapy for syphilis in pregnancy, consistent with WHO guidelines. Alternative regimens to BPG were listed in 42 (68%) guidelines, primarily from Africa and Asia; only 20 specified that non-penicillin regimens are not proven-effective in treating the fetus. We identified guidance recommending use of injectable penicillin in exposed infants for 112 countries. Most guidelines recommended universal syphilis testing for pregnant women, repeat testing for high-risk women and treatment of infected women with BPG; but several did not. Updating guidance on syphilis testing and treatment in pregnancy to reflect global norms could prevent congenital syphilis and save newborn lives.
Collapse
Affiliation(s)
- Thuy Trinh
- Public Health Analyst, Division of Sexually Transmitted Disease Prevention, Centers for Disease Control and Prevention, Atlanta, GA, USA
| | - Alexis F Leal
- Project Associate, Department of Communicable Diseases and Environmental Determinants of Health, Pan American Health Organization/World Health Organization, Washington, DC, USA
| | - Maeve B Mello
- Medical Epidemiologist, Regional Adviser for HIV/STI Prevention, Department of Communicable Diseases and Environmental Determinants of Health, Pan American Health Organization/World Health Organization, Washington, DC, USA
| | - Melanie M Taylor
- Medical Epidemiologist, Division of Sexually Transmitted Disease Prevention, Centers for Disease Control and Prevention, Atlanta, GA, USA; Medical Officer, Department of Reproductive Health and Research, World Health Organization, Geneva, Switzerland
| | - Roxanne Barrow
- Medical Epidemiologist, Division of Sexually Transmitted Disease Prevention, Centers for Disease Control and Prevention, Atlanta, GA, USA
| | - Teodora E Wi
- Medical Officer, Department of Reproductive Health and Research, World Health Organization, Geneva, Switzerland
| | - Mary L Kamb
- Medical Epidemiologist, Division for Parasitic Diseases and Malaria, Center for Global Health, Atlanta, GA, USA
| |
Collapse
|
8
|
Malacova E, Tippaya S, Bailey HD, Chai K, Farrant BM, Gebremedhin AT, Leonard H, Marinovich ML, Nassar N, Phatak A, Raynes-Greenow C, Regan AK, Shand AW, Shepherd CCJ, Srinivasjois R, Tessema GA, Pereira G. Stillbirth risk prediction using machine learning for a large cohort of births from Western Australia, 1980-2015. Sci Rep 2020; 10:5354. [PMID: 32210300 PMCID: PMC7093523 DOI: 10.1038/s41598-020-62210-9] [Citation(s) in RCA: 23] [Impact Index Per Article: 5.8] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/19/2019] [Accepted: 03/11/2020] [Indexed: 11/30/2022] Open
Abstract
Quantification of stillbirth risk has potential to support clinical decision-making. Studies that have attempted to quantify stillbirth risk have been hampered by small event rates, a limited range of predictors that typically exclude obstetric history, lack of validation, and restriction to a single classifier (logistic regression). Consequently, predictive performance remains low, and risk quantification has not been adopted into antenatal practice. The study population consisted of all births to women in Western Australia from 1980 to 2015, excluding terminations. After all exclusions there were 947,025 livebirths and 5,788 stillbirths. Predictive models for stillbirth were developed using multiple machine learning classifiers: regularised logistic regression, decision trees based on classification and regression trees, random forest, extreme gradient boosting (XGBoost), and a multilayer perceptron neural network. We applied 10-fold cross-validation using independent data not used to develop the models. Predictors included maternal socio-demographic characteristics, chronic medical conditions, obstetric complications and family history in both the current and previous pregnancy. In this cohort, 66% of stillbirths were observed for multiparous women. The best performing classifier (XGBoost) predicted 45% (95% CI: 43%, 46%) of stillbirths for all women and 45% (95% CI: 43%, 47%) of stillbirths after the inclusion of previous pregnancy history. Almost half of stillbirths could be potentially identified antenatally based on a combination of current pregnancy complications, congenital anomalies, maternal characteristics, and medical history. Greatest sensitivity is achieved with addition of current pregnancy complications. Ensemble classifiers offered marginal improvement for prediction compared to logistic regression.
Collapse
Affiliation(s)
- Eva Malacova
- School of Public Health, Curtin University, Perth, WA, Australia
- QIMR Berghofer Medical Research Institute, Brisbane, QLD, Australia
- Faculty of Health and Medical Sciences, School of Population and Public Health, Perth, WA, Australia
| | - Sawitchaya Tippaya
- School of Public Health, Curtin University, Perth, WA, Australia
- Curtin Institute for Computation, Curtin University, Perth, WA, Australia
| | - Helen D Bailey
- Telethon Kids Institute, The University of Western Australia, Perth, WA, Australia
| | - Kevin Chai
- Curtin Institute for Computation, Curtin University, Perth, WA, Australia
| | - Brad M Farrant
- Telethon Kids Institute, The University of Western Australia, Perth, WA, Australia
| | | | - Helen Leonard
- Telethon Kids Institute, The University of Western Australia, Perth, WA, Australia
| | | | - Natasha Nassar
- Child Population and Translational Health Research, The Children's Hospital at Westmead Clinical School, The University of Sydney, Sydney, NSW, Australia
| | - Aloke Phatak
- Curtin Institute for Computation, Curtin University, Perth, WA, Australia
- Centre for Transforming Maintenance through Data Science, Curtin University, Perth, WA, Australia
| | | | - Annette K Regan
- School of Public Health, Curtin University, Perth, WA, Australia
- Telethon Kids Institute, The University of Western Australia, Perth, WA, Australia
- School of Public Health, Texas A&M University, Texas, USA
| | - Antonia W Shand
- Child Population and Translational Health Research, The Children's Hospital at Westmead Clinical School, The University of Sydney, Sydney, NSW, Australia
- Department of Maternal Fetal Medicine, Royal Hospital for Women, Randwick, NSW, Australia
| | - Carrington C J Shepherd
- Telethon Kids Institute, The University of Western Australia, Perth, WA, Australia
- Ngangk Yira: Murdoch University Research Centre for Aboriginal Health and Social Equity, Perth, WA, Australia
| | - Ravisha Srinivasjois
- School of Public Health, Curtin University, Perth, WA, Australia
- Telethon Kids Institute, The University of Western Australia, Perth, WA, Australia
- Department of Neonatology, Ramsay Health Care, Joondalup Health Campus, Joondalup, WA, Australia
| | | | - Gavin Pereira
- School of Public Health, Curtin University, Perth, WA, Australia.
- Telethon Kids Institute, The University of Western Australia, Perth, WA, Australia.
- Centre for Fertility and Health (CeFH), Norwegian Institute of Public Health, Oslo, Norway.
| |
Collapse
|