1
|
Bai X, Zhou Z, Zheng Z, Li Y, Liu K, Zheng Y, Yang H, Zhu H, Chen S, Pan H. Development and evaluation of machine learning models for predicting large-for-gestational-age newborns in women exposed to radiation prior to pregnancy. BMC Med Inform Decis Mak 2024; 24:174. [PMID: 38902714 PMCID: PMC11188254 DOI: 10.1186/s12911-024-02556-6] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/17/2023] [Accepted: 05/28/2024] [Indexed: 06/22/2024] Open
Abstract
INTRODUCTION The correlation between radiation exposure before pregnancy and abnormal birth weight has been previously proven. However, for large-for-gestational-age (LGA) babies in women exposed to radiation before becoming pregnant, there is no prediction model yet. MATERIAL AND METHODS The data were collected from the National Free Preconception Health Examination Project in China. A sum of 455 neonates (42 SGA births and 423 non-LGA births) were included. A training set (n = 319) and a test set (n = 136) were created from the dataset at random. To develop prediction models for LGA neonates, conventional logistic regression (LR) method and six machine learning methods were used in this study. Recursive feature elimination approach was performed by choosing 10 features which made a big contribution to the prediction models. And the Shapley Additive Explanation model was applied to interpret the most important characteristics that affected forecast outputs. RESULTS The random forest (RF) model had the highest average area under the receiver-operating-characteristic curve (AUC) for predicting LGA in the test set (0.843, 95% confidence interval [CI]: 0.714-0.974). Except for the logistic regression model (AUC: 0.603, 95%CI: 0.440-0.767), other models' AUCs displayed well. Thereinto, the RF algorithm's final prediction model using 10 characteristics achieved an average AUC of 0.821 (95% CI: 0.693-0.949). CONCLUSION The prediction model based on machine learning might be a promising tool for the prenatal prediction of LGA births in women with radiation exposure before pregnancy.
Collapse
Affiliation(s)
- Xi Bai
- Key Laboratory of Endocrine Glucose & Lipids Metabolism and Brain Aging, Department of Endocrinology, Ministry of Education, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, Shandong, China
- Key Laboratory of Endocrinology of National Health Commission, Department of Endocrinology, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College Hospital, Chinese Academy of Medical Science and Peking Union Medical College, Beijing, 100730, China
| | - Zhibo Zhou
- Key Laboratory of Endocrinology of National Health Commission, Department of Endocrinology, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College Hospital, Chinese Academy of Medical Science and Peking Union Medical College, Beijing, 100730, China
| | - Zeyan Zheng
- Key Laboratory of Endocrinology of National Health Commission, Department of Endocrinology, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College Hospital, Chinese Academy of Medical Science and Peking Union Medical College, Beijing, 100730, China
| | - Yansheng Li
- DHC Mediway Technology CO., Ltd, Beijing, China
| | - Kejia Liu
- DHC Mediway Technology CO., Ltd, Beijing, China
| | | | - Hongbo Yang
- Key Laboratory of Endocrinology of National Health Commission, Department of Endocrinology, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College Hospital, Chinese Academy of Medical Science and Peking Union Medical College, Beijing, 100730, China
| | - Huijuan Zhu
- Key Laboratory of Endocrinology of National Health Commission, Department of Endocrinology, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College Hospital, Chinese Academy of Medical Science and Peking Union Medical College, Beijing, 100730, China
| | - Shi Chen
- Key Laboratory of Endocrinology of National Health Commission, Department of Endocrinology, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College Hospital, Chinese Academy of Medical Science and Peking Union Medical College, Beijing, 100730, China.
| | - Hui Pan
- Key Laboratory of Endocrinology of National Health Commission, Department of Endocrinology, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College Hospital, Chinese Academy of Medical Science and Peking Union Medical College, Beijing, 100730, China.
| |
Collapse
|
2
|
Singh D, Manna S, Barik M, Rehman T, Kanungo S, Pati S. Prevalence and correlates of low birth weight in India: findings from national family health survey 5. BMC Pregnancy Childbirth 2023; 23:456. [PMID: 37340388 DOI: 10.1186/s12884-023-05726-y] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/14/2023] [Accepted: 05/20/2023] [Indexed: 06/22/2023] Open
Abstract
BACKGROUND Childhood mortality and morbidity has become a major public health issue in low-middle-income countries. However, evidence suggested that Low birth weight(LBW) is one of the most important risk factors for childhood deaths and disability.This study is designed to estimate the prevalence of low birth weight (LBW) in India and to identify maternal correlates associated with LBW. METHODS Data has been taken from National Family Health Survey 5 (2019-2021) for analysis. 149,279 women belonging to reproductive age group (15-49) year who had last recent most delivery preceding the NFHS-5 survey. RESULTS Mother's age, female child, birth interval of less than 24 months, their low educational level, low wealth index, rural residence, lack of insurance coverage, women with low BMI, anaemia, and no ANC visits during pregnancy are predictors that contribute to LBW in India. After adjusting for covariates, smoking and alcohol consupmtion is strongly correlated with LBW. CONCLUSION Mother's age, educational attainment and socioeconomic status of living has a highly significant with LBW in India. However, consumption of tobacco and cigarrettes are also associated with LBW.
Collapse
Affiliation(s)
- Damini Singh
- Division of Public Health Research, ICMR-Regional Medical Research Centre, Bhubaneswar-23, Bhubaneswar, Odisha, India
| | - Sayantani Manna
- Division of Public Health Research, ICMR-Regional Medical Research Centre, Bhubaneswar-23, Bhubaneswar, Odisha, India
| | - Manish Barik
- Division of Public Health Research, ICMR-Regional Medical Research Centre, Bhubaneswar-23, Bhubaneswar, Odisha, India
| | - Tanveer Rehman
- Division of Public Health Research, ICMR-Regional Medical Research Centre, Bhubaneswar-23, Bhubaneswar, Odisha, India
| | - Srikanta Kanungo
- Division of Public Health Research, ICMR-Regional Medical Research Centre, Bhubaneswar-23, Bhubaneswar, Odisha, India.
| | - Sanghamitra Pati
- Division of Public Health Research, ICMR-Regional Medical Research Centre, Bhubaneswar-23, Bhubaneswar, Odisha, India.
| |
Collapse
|
3
|
Cho KH, Yoon SJ, Lim J, Eun H, Park MS, Park KI, Jo HS, Lee SM. Epidemiology of Macrosomia in Korea: Growth and Development. J Korean Med Sci 2021; 36:e320. [PMID: 34873886 PMCID: PMC8648607 DOI: 10.3346/jkms.2021.36.e320] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 07/07/2021] [Accepted: 10/14/2021] [Indexed: 12/23/2022] Open
Abstract
BACKGROUND Macrosomia, as an infant with birth weight over 4 kg, can have several perinatal, and neonatal complications. This study aimed to estimate the incidence of macrosomia in Korea and to identify the growth and developmental outcomes and other neonatal complications. METHODS In total, 397,203 infants who were born in 2013 with birth weight ≥ 2.5 kg and who underwent infant health check-up between their 1st and 7th visit were included from the National Health Insurance Service database. The information was obtained by the International Classification of Diseases-10 codes or self-reported questionnaires in the National Health Screening Program. RESULTS The distribution of infants by birth weight was as follows: 384,181 (97%) infants in the 2.5-3.99 kg (reference) group, 12,016 (3%) infants in the 4.0-4.49 kg group, 772 (0.2%) infants in the 4.5-4.99 kg group, and 78 (0.02%) infants in the ≥ 5 kg group. Macrosomia showed significantly higher incidence of sepsis, male sex, and mothers with GDM and birth injury. There was a significant difference in weight, height, and head circumference according to age, birth weight group, and combination of age and birth weight, respectively (P < 0.001). The number of infants with the weight above the 90th percentile in macrosomia at each health check-up showed higher incidence than in reference group. The mean body mass index significantly differed among the groups, as 50.6 in infants with 2.5-3.99 kg of birth weight, 63.5 with 4.0-4.49 kg, 71.0 with 4.5-4.99 kg, and 73.1 with ≥ 5 kg. There was a significant difference in the incidence of poor developmental results between infants with macrosomia and the reference group at 24, 36 and 48 month of age. CONCLUSION Macrosomia was significantly associated with the risk of sepsis, birth injury, obesity and developmental problem especially in a boy born from mothers with gestational diabetes mellitus. Careful monitoring and proper strategies for monitoring growth and development are needed.
Collapse
Affiliation(s)
- Kee Hyun Cho
- Department of Pediatrics, Kangwon National University Hospital, Chuncheon, Korea
| | - So Jin Yoon
- Department of Pediatrics, Yonsei University College of Medicine, Seoul, Korea
| | - Joohee Lim
- Department of Pediatrics, Yonsei University College of Medicine, Seoul, Korea
| | - Hoseon Eun
- Department of Pediatrics, Yonsei University College of Medicine, Seoul, Korea
| | - Min Soo Park
- Department of Pediatrics, Yonsei University College of Medicine, Seoul, Korea
| | - Kook In Park
- Department of Pediatrics, Yonsei University College of Medicine, Seoul, Korea
| | - Heui Seung Jo
- Department of Pediatrics, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea
| | - Soon Min Lee
- Department of Pediatrics, Yonsei University College of Medicine, Seoul, Korea.
| |
Collapse
|
4
|
Modzelewski J, Pokropek A, Jakubiak-Proć M, Muzyka-Placzyńska K, Filipecka-Tyczka D, Kajdy A, Rabijewski M. Large-for-gestational-age or macrosomia as a classifier for risk of adverse perinatal outcome: a retrospective cross-sectional study. J Matern Fetal Neonatal Med 2021; 35:5564-5571. [PMID: 33602007 DOI: 10.1080/14767058.2021.1887127] [Citation(s) in RCA: 5] [Impact Index Per Article: 1.7] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/22/2022]
Abstract
OBJECTIVE Small for gestational age (SGA) fetuses and neonates are of great interest, while those who are too big are much less studied. The aim was to analyze the classifiers described by ACOG "Fetal macrosomia" practice bulletin as predictors of adverse perinatal outcomes for overgrown fetuses and their mothers. MATERIALS From a database of 53,586 singleton term births, appropriate-for-gestational-age (AGA), large for gestational age (LGA), and macrosomic deliveries were selected. AGA served as a control. The crude and adjusted odds ratios (aORs) were calculated for large-for-gestational-age >90th centile, and macrosomia >4000 g, >4250 g, and >4500 g. Patients with and without diabetes were analyzed separately. RESULTS Macrosomia >4000 g performed poorer than other classifiers. LGA performed comparably to other definitions of macrosomia. Diabetes carries a severe risk of complications for overgrown neonates, but those non-diabetic also have increased risk. CONCLUSIONS Definition of macrosomia as weight >4000 g should be reconsidered. LGA >90th centile should be used as a definition of fetal overgrowth along with other definitions of macrosomia.
Collapse
Affiliation(s)
- Jan Modzelewski
- Department of Reproductive Health, Centre of Postgraduate Medical Education, Warsaw, Poland
| | - Artur Pokropek
- Department of Sociology, Institute of Philosophy and Sociology Polish Academy of Sciences, Warsaw, Poland
| | - Monika Jakubiak-Proć
- Department of Reproductive Health, Centre of Postgraduate Medical Education, Warsaw, Poland
| | | | | | - Anna Kajdy
- Department of Reproductive Health, Centre of Postgraduate Medical Education, Warsaw, Poland
| | - Michał Rabijewski
- Department of Reproductive Health, Centre of Postgraduate Medical Education, Warsaw, Poland
| |
Collapse
|
5
|
Postnatal Outcomes and Risk Factors for In-Hospital Mortality among Asphyxiated Newborns in a Low-Resource Hospital Setting: Experience from North-Central Nigeria. Ann Glob Health 2020; 86:63. [PMID: 32587813 PMCID: PMC7304451 DOI: 10.5334/aogh.2884] [Citation(s) in RCA: 3] [Impact Index Per Article: 0.8] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/22/2022] Open
Abstract
Background: Birth asphyxia accounts for a third of global newborn deaths and 95 percent of these occur in low-resource settings. A key to reducing asphyxia-related deaths in these settings is improving care of these newborns and this requires an understanding of factors associated with adverse outcomes. Objectives: In this study, we report outcomes and risk factors for mortality among newborn infants with birth asphyxia admitted to a typical low-resource hospital setting. Methods: We prospectively followed up 191 asphyxiated newborn infants admitted to a referral tertiary hospital in North-central Nigeria. At baseline, care-givers completed a structured questionnaire. Using univariable analysis, we compared baseline characteristics between participants who died and those who survived till discharge. We also fitted a multivariable logistic regression model to identify risk factors for mortality among the cohort. Results: Majority (60.7%) of the study participants presented to the hospital within the first six hours of life. Despite this, mortality among the cohort was 14.7% with a third dying within the first 24 hours of admission. The presence of respiratory distress at admission increased the risk for mortality (AOR = 3.73, 95% CI 1.22 to 11.35) while higher participant weight at admission decreased the risk (AOR = 0.11, 95% CI 0.03 to 0.40). Intrapartum factors such as duration of labour and maternal age, although significant on univariable analysis, were not significant on multivariable analysis. Conclusions: Hospital mortality among newborns with birth asphyxia is high in North-central Nigeria and majority of deaths occur during acute care. Respiratory distress at presentation and admission weights were identified as key risk factors for asphyxia mortality. Intrapartum factors on the other hand might have indirect effects on mortality through an increased risk for neonatal complications.
Collapse
|