Infant Death Research Articles

Infant Mortality among the Tea Garden Workers in Upper Assam: Insights from a Field Survey

Infant mortality rate is an important indicator that reflects not only the health status of the infant but also the well-being of society as a whole. The north-eastern state of Assam is one of the states having the highest number of maternal and infant deaths within the country. For the first time, infant mortality in Assam registered a decline to 31.9 compared to the national average of 35.2 per 1,000 live births (NFHS-5). Though the overall rank of the state in terms of infant mortality has improved (from the 34th position to the 25th position), it is not enough and is not evenly distributed across the communities within the state. Human rights violations are particularly prominent among the tea garden workers of Assam and are mostly seen against the Adivasi children and woman, which is reflected by high maternal and infant deaths. The study aims to determine the factors responsible for infant mortality among tea garden workers by using primary data collected from 300 women workers across tea gardens in the Sonitpur and Tinsukia districts of Upper Assam. The binary logistic regression method is used for the analysis. It was observed that among the six probable factors, namely educational qualifications of the mother, marriage age of the mother, daily working hours, daily wage as tea garden labour, total number of births and the source of drinking water, the following three factors—namely the total number of births, age at marriage and the source of drinking water—are found to be the significant determinants of infant mortality among the tea garden labourers of Assam. JEL Codes: I12, I15

Abstract 4112716: The Trend In Racial Differences In Mortality Attributed To Congenital Heart Diseases In Infants In The United States From 2005 To 2019

Background: Deaths from congenital heart disease (CHD) in children have been decreasing in the United States. We examined the differences in mortality trends between Non-Hispanic Black (NHB) and Non-Hispanic White (NHW) infants. Methods: We retrospectively analyzed publicly available data from the Centers for Disease Control and Prevention’s Wide-ranging Online Data for Epidemiologic Research (CDC WONDER). The data was obtained from the linked birth/infant deaths from 2005 to 2019. We evaluated all infant deaths up to 1 year of age with the cause of death listed as CHD (International classification of diseases, 10 th revision (ICD-10) codes Q20-Q26 (except atrial septal defect, Q21.1 and patent ductus arteriosus, Q25. CHD infant mortality rate (IMR) was calculated per 100,000 live births. Race was ascertained based on death certificate reporting. Joinpoint regression was used to examine CHD-IMR by year, including stratification by NHB vs NHW, and neonatal vs postneonatal. The difference between NHB and NHW CHD-IMR was ascertained via the Mann-Whitney U test. P<0.05 was considered statistically significant. Results: Out of 60,243,988 live births, there were 19,004 CHD-related infant deaths. The characteristics of the study population are displayed in Table 1. The overall CHD-IMR declined from 36.1 to 27.0 per 100,000 live births (25.2%) with an average annual percentage change (AAPC) of -2.1 [95% CI -2.6, -1.572] (Figure 1). The CHD-IMR was significantly higher in NHB than in NHW (40 vs 29.3; P<0.0001) and this difference remained when the mortality rate was stratified by neonatal and post-neonatal periods. The CHD-IMR decreased significantly in NHW [AAPC: -2%; 95% CI: -2.5, -1.5%], but not in NHB [AAPC of -1.4%, 95% CI, -3.0, 0.3%]. The NHB-to-NHW CHD-IMR ratio, a measure of CHD-IMR disparity averaged 1.4 over the study period [AAPC: 0.8%; 95% CI: -0.2, 1.8%] (Figure 2). Conclusions: The CHD-IMR declined significantly for NHW, but not in NHB in the US from 2005 to 2019. The CHD-IMR was higher in NHB than in NHW. There was no significant change in the NHB-to-NHW CHD-IMR ratio, signifying no change in the disparity that exists between NHB and NHW IMR. These findings may have implications for patient care and public health policymaking. They lay the groundwork for additional studies to determine and understand the drivers behind these findings such as in prenatal diagnosis.

Parental obesity and the risk of sudden unexpected infant death

AbstractAimTo clarify the possible role of accidental asphyxiation/suffocation for infants in shared sleeping situations.MethodsAs it has recently been shown that there is an association between maternal obesity and sudden unexpected infant death (SUID) an index was calculated by dividing parental body mass index (BMI) by the shared sleeping surface area.ResultsA much higher index was derived for an obese individual on a narrow sleeping area such as a sofa (50) compared to a thin adult on a queen‐sized mattress (5).ConclusionThe index may be a way of semi‐quantitating high, and low, risk shared sleeping situations.

Maternal Obesity and Sudden Unexpected Infant Death

Maternal Obesity and Sudden Unexpected Infant Death

Maternal Obesity and Sudden Unexpected Infant Death—Reply

Maternal Obesity and Sudden Unexpected Infant Death—Reply

State and Local Government Expenditures and Infant Mortality.

A previous study reported that increased state and local government expenditures were associated with decreased infant mortality rates (IMRs). However, reported estimates of the association between expenditures and IMR represented the degree to which the association changed each year, not the main effect. We reproduced the original results, reporting this main effect and replicated the analysis using improved methodology and updated data. For the reproduction analysis, we used methods and data identical to the original study: A publicly-posted, state-level data set of expenditures from 2000 to 2014 US Census Bureau survey data linked to 2-year lagged IMR data with a random intercept model including an interaction between time and expenditures. For the replication analysis, we added 5 years of data and adjusted for fixed state differences and inflation. In the reproduction, the main effects of total, environmental, and educational expenditures on IMR were much larger than the interaction effects previously reported as the main effects. For example, a 1-SD increase in per-capita total expenditures was associated with a reduction of 0.35 infant deaths per 1000 live births instead of 0.02 deaths per 1000 live births originally reported. In the updated replication, the main effects were generally even larger (eg, -0.51 deaths per 1000 per SD increase in total expenditures). Increased total expenditures were associated with absolute but not relative reductions in Black-white IMR gaps. State and local government expenditures are associated with greater reductions in IMR than previously reported, underscoring the importance of continued public investment.

Perinatal deaths in twin and singleton infants in Ireland: A comparison of characteristics and causes.

Twin pregnancies are associated with significantly higher perinatal mortality (PM) rates compared to singletons, primarily due to complications like fetal growth restriction, preterm birth, and congenital anomalies. This study aimed to compare the characteristics associated with PM in twin pregnancies and compare maternal and obstetric factors and cause of death among twins and singletons in the Republic of Ireland. Data spanning 2011 to 2022 from the National Perinatal Epidemiology Centre's annual perinatal mortality clinical audit included 4494 perinatal deaths. Maternal characteristics, antenatal care factors and cause of death were analysed with relative risk calculated using national Hospital In-Patient Enquiry data. Pearson's chi-squared tests studied the difference between mortality in twins and singletons. Twins accounted for 10.4% of all perinatal deaths, despite representing only 3.6% of total births. The PM rate for twins was 17.3 per 1000 births, 3.1 times higher than for singletons. Early neonatal deaths (ENNDs) were more frequent in twins (54.2%), while stillbirths predominated among singletons (68.6%). Younger maternal age and lower BMI were associated with higher PM risks in twins. A considerable proportion of twin deaths with major congenital anomalies or birth before 28weeks gestation occurred in non-tertiary hospitals, suggesting limitations in referral pathways to centres with appropriate neonatal expertise. Twin pregnancies pose a higher risk of perinatal mortality, particularly among younger mothers and preterm births. The findings highlight the need for updated guidelines that prioritise early risk assessment, targeted interventions, and improved referral systems.

Analysis of the potential correlation between infant mortality rate and GDP per capita among 217 countries

Abstract. This research explores the connection between GDP per capita and infant mortality rates (IMR) by examining data collected from 217 countries between 2000 and 2022. Regression analysis is utilized to examine the associations among different factors, encompassing healthcare, education, and unemployment, previously identified as influencing the infant mortality rate (IMR). The analysis reveals an inverse relationship between GDP per capita and IMR. Specifically, one dollar growth in GDP per capita corresponds to a 0.107 reduction in infant death, with all other variables held constant. The study also examined interaction terms, revealing a positive correlation between certain variable combinations and infant mortality. These findings indicate that policymakers should prioritize improving reproductive education and allocating more funding to healthcare in order to further decrease infant death. Nevertheless, the study has constraints, such as the absence of certain observations and the necessity for supplementary social and economic factors to acquire a more comprehensive understanding of the components impacting infant death.

Etiologies and Comorbidities of Meningitis Deaths in Children Under 5 Years in High-Mortality Settings: Insights from the CHAMPS Network in the Post-Pneumococcal Vaccine Era

BackgroundThe role of meningitis in causing deaths and in children under 5 is unclear, especially since widespread use of vaccines to prevent common causes of meningitis. Child Health and Mortality Prevention Surveillance (CHAMPS) uses post-mortem minimally invasive tissue sampling (MITS) and ante-mortem data to explore death causes. We aimed to assess meningitis’s contribution to mortality and identify causative pathogens in children under 5 within CHAMPS Network sites. MethodIn this observational study, we analyzed deaths in live-born children <5 years of age that occurred between December 16, 2016, and December 31, 2023, in CHAMPS catchments in six sub-Saharan African countries (Ethiopia, Kenya, Mali, Mozambique, Sierra Leone, South Africa) and Bangladesh. MITS was conducted within 24–72hours of death, including blood and cerebrospinal fluid (CSF) culture, multi-organism targeted nucleic acid amplification tests on blood, CSF and lung tissue, and histopathology of lung, liver and brain. Expert panels at each site reviewed data to attribute causes of death following ICD-10 standards. ResultMeningitis was in the causal pathway for 7.0% (270/3857) of deaths; in 4.8% (13/270) meningitis was considered the underlying condition. Neonates accounted for 65.9% (178/270) and infants or children 34.1% (92/270). Among neonatal meningitis deaths, 55.6% (99/178) occurred ≥72hours post-hospital admission; and common pathogens were Acinetobacter baumannii (49.5%, 49/99; mainly from South Africa) and Klebsiella pneumoniae (40.4%, 40/99). Forty-four percent (79/178) of neonatal meningitis deaths were community-associated, primarily due to K. pneumoniae (35.4%, 28/79) and Escherichia coli (13.9%, 11/79). Among infant and child meningitis deaths, 43.5% (40/92) occurred ≥72hours post-admission; and common pathogens were K. pneumoniae (42.5%,17/40) and A. baumannii (17.5%, 7/40). Among community-associated meningitis deaths in infants and children (56.5%, 52/92), Streptococcus pneumoniae (34.6%, 18/52) and K. pneumoniae (19.2%, 10/52) were common pathogens. Pathogen prevalence varied by region. ConclusionOur study highlights meningitis as a significant contributor to under-5 mortality in low-middle-income countries. The prominent role of K. pneumoniae and A. baumannii, particularly in healthcare settings and specific regions, highlights the need for better infection control, targeted interventions, and more effective treatment strategies.

Using Sudden Unexpected Infant Death-Case Registry Data to Drive Prevention.

Sudden unexpected infant death (SUID) is a major contributor to infant death and a persistent public health issue. After an initial decline after the 1994 "Back to Sleep" campaign, SUID numbers plateaued. Currently, ∼10 infants die suddenly and unexpectedly each day in the United States. In 2019, we established a surveillance system for SUID in Cook County, Illinois, partnering our academic medical center, the Cook County Medical Examiner's Office, and child death review to create the Cook County SUID Case Registry. Our data show that, in Cook County, including the city of Chicago, ∼1 infant dies unexpectedly during sleep every week. Of these SUID, ∼25% were because of suffocation/possible suffocation. SUID peaks at 30 to 60 days old. SUID rates are 15 times higher in non-Hispanic Black infants and 3 times higher in Hispanic infants, compared with white infants. Nearly all involved 1 or more unsafe sleep factors. SUID are concentrated in community areas experiencing high hardship. Through our Community Partnership Approaches for Safe Sleep-Chicago team, we have developed collaborative prevention approaches in affected communities, allowing for conversations with families and those who support them to better understand barriers to safe sleep that they experience. These partnerships and our data allow for tailoring of informed prevention approaches to address upstream factors driving disproportionate infant mortality in historically disinvested communities, as well as optimizing the immediate risks posed by the infant sleep environment. Data from our system show the number of SUID declining modestly since our prevention work began.

Unexplained Infant Deaths Without Unsafe Sleep Factors: 2011 to 2020.

To describe sudden unexpected infant deaths (SUIDs) occurring in safe sleep environments and explore differences in selected characteristics. We examined SUID from 22 jurisdictions from 2011 to 2020 and classified them as unexplained, no unsafe sleep factors (U-NUSF). Data were derived from the Sudden Unexpected Infant Death and Sudden Death in the Young Case Registry, a population-based Centers for Disease Control and Prevention surveillance system built on the National Center for Fatality Review and Prevention's child death review program. SUID classified as U-NUSF included infants who were (1) awake, under supervision, and witnessed to become unresponsive or (2) found unresponsive in a safe sleep environment after sleep (unwitnessed). We calculated frequencies and percentages for demographics, birth and environmental characteristics, medical history, and death investigation findings. Most of the 117 U-NUSF SUID occurred before 4 months of age. Witnessed deaths most commonly occurred at <1 month of age (28%), whereas unwitnessed deaths most commonly occurred at ages 2 to 3 months (44%) Among all U-NUSF, 69% occurred in the infant's home (62% witnessed, 77% unwitnessed). All unwitnessed deaths occurred in a crib; most witnessed deaths occurred while being held (54%) or in a car seat traveling (18%). Most infants (84%) had no history of abuse or neglect. Abnormal autopsy findings were reported in 46% of deaths (49% witnessed, 42% unwitnessed). Characterizing these deaths is key to advancing our knowledge of SUID etiology. Our study revealed a heterogeneous group of infants, suggesting physiologic, genetic, or environmental etiologies.

Improving Consistency in Classifying Child Maltreatment for Sudden Unexpected Infant Deaths.

Develop guidelines for child death review teams that will improve the consistency in classifying child maltreatment (CM) and distinguish between classification of exposure to hazards and neglect for sleep-related sudden unexpected infant deaths (SUID). Sleep-related SUID (n = 25 797) were identified from the National Fatality Review-Case Reporting System between 2004 and 2018. Key variables considered when classifying CM among sleep-related SUID were identified. Logistic regression was used to assess the strength of associations and identify factors that distinguished between exposure to hazards and neglect. Odds ratios (OR) and 95% confidence intervals (CI) were calculated. Guidelines were developed based on these findings. Among the 13 034 sleep-related SUID with CM identified, hazards in the infant sleep environment were strongly associated with classification of both exposure to hazards and neglect, as were supervisor impairment and child welfare substantiation of the death. Comparing neglect to exposure to hazards, there was no association with sleep environment hazards with ORs ranging from 0.8 to 1.3 (95% CI 0.4-3.0), but sleep-related SUID were approximately twice as likely to be classified as neglect when the supervisor was impaired (OR 2.3, 95% CI 2.0-2.7) or child welfare action was taken (OR 1.8, 95% CI 1.5-2.0). The guidelines recommend classification of exposure to hazards for sleep environment hazards with elevation to neglect if the supervisor was impaired or child welfare substantiated the death. Among sleep-related SUID, classification of exposure to hazards is based on presence of sleep environment hazards, whereas neglect is based on supervisor impairment and child welfare action.

Collaboration to Reduce Sudden Unexpected Infant Death With Child Fatality Review and Outreach.

Sudden unexpected infant death (SUID) is the leading cause of postneonatal infant mortality in the United States, with disproportionately high rates in Baltimore City and Baltimore County in Maryland. This Advocacy Case Study describes the collaboration between the City and County Child Fatality Review teams to decrease infant mortality. B'more for Healthy Babies, formed in 2009 by the Bureau of Maternal Child Health in Baltimore City with the goal of reducing infant mortality through policy change, service improvements, community mobilization, and behavior change has had a sustained effort to respond to SUID. Recognizing that infants born in Baltimore City often reside in Baltimore County (and vice versa), collaboration with Baltimore County has been essential to increasing B'more for Healthy Babies' scope, reach, and effectiveness. Public health messaging campaigns, creation, and dissemination of a "SLEEP SAFE" testimonial video, and Safe Sleep Summits have involved caregivers, healthcare providers, and community partners. Outcomes of this ongoing collaboration demonstrate a decrease in sleep-related infant mortality rates. Our findings also highlight the need for continued, real time monitoring of sleep-related infant mortality trends with a coordinated, multidisciplinary, and crossjurisdictional response. This initiative can serve as a model of cross-sector communication that can be replicated in other similar geographic locations to further reduce SUID.

Evaluating Educational Patterns and Methods in Infant Sleep Care: Trends, Effectiveness, and Impact in Home Settings—A Systematic Review

Background: Sleep care is crucial for the health and development of infants, with proper sleep patterns reducing the risk of sudden infant death syndrome (SIDS) and other sleep-related incidents. Educational interventions targeting caregivers are essential in promoting safe sleep practices. Methods: This systematic review adhered to PRISMA guidelines, searching databases such as PubMed, MEDLINE, Scopus, and the Cochrane Library. Inclusion criteria focused on studies involving home-based interventions for infants aged 0–12 months, including parental education and behavioral interventions. Exclusion criteria included studies in clinical settings and non-peer-reviewed articles. Data extraction and synthesis were performed by two independent reviewers, using a narrative approach to categorize interventions and outcomes. Results: Twenty-three studies met the inclusion criteria. Key findings indicate that home-based educational interventions, including hospital-based programs, home visits, and mobile health technologies, significantly improve parental knowledge and adherence to safe sleep practices. These interventions also enhance parental satisfaction and contribute positively to infant health outcomes. Conclusions: Educational interventions have demonstrated effectiveness in promoting safe sleep practices among caregivers, particularly in home settings. These interventions, including hospital-based programs, home visits, and digital tools, improve parental knowledge, adherence to guidelines, and overall satisfaction. The impact is evident in the reduction of unsafe sleep behaviors and enhanced infant health outcomes. However, variability in the intervention methods and delivery, cultural contexts, and geographic focus suggest a need for more tailored, long-term, and comprehensive studies. Future research should standardize outcome measures and assess the sustained impact of these educational strategies on infant sleep patterns and caregiver practices over time. This will provide deeper insights into the trends and long-term effectiveness of educational patterns and methods in diverse home environments.

Assessing universal considerations in infant mortality across the globe: A descriptive observational study of sudden infant death syndrome knowledge and reduction coverage on YouTube

Assessing universal considerations in infant mortality across the globe: A descriptive observational study of sudden infant death syndrome knowledge and reduction coverage on YouTube

Implementation of Primary Immune Thrombocytopenia Clinical Practice Guidelines for Management of Pregnancy

Background: Managing primary immune thrombocytopenia (ITP) in pregnancy is challenging. Providers must balance bleeding risk against medication toxicity. The evaluation of the implementation of pregnancy-specific ITP clinical guidelines has not been widely studied. The goal of this study was to describe the implementation of pregnancy-specific ITP guidelines at an academic health center. Methods: We conducted a retrospective chart review at the University of Oklahoma Health system from 2011 to 2020. Descriptive statistics were calculated to summarize the characteristics of the study population. Management, according to the clinical guidelines (American Society of Hematology; American College of Obstetricians and Gynecologists) was evaluated during pregnancy and during/for delivery. Results: A total of 85 pregnant persons with ITP were included. The majority (68%; 58/85) delivered vaginally. There were 0 maternal deaths and 2 infant deaths. No patients had major bleeding during pregnancy. Postpartum hemorrhage was experienced by 14%. The management of thrombocytopenia during pregnancy was 100% adherent to the strong recommendation for severe (n = 13) and mild (n = 11) thrombocytopenia. However, 18/50 (36%) asymptomatic persons with moderate thrombocytopenia received treatment despite the strong recommendation that treatment was unnecessary. Additionally, 8/21 (38%) persons with moderate thrombocytopenia received treatment to increase platelet counts for epidural anesthesia despite the guideline’s suggestion that it was unnecessary. Conclusions: During pregnancy, patients with severe thrombocytopenia (i.e., most at risk of bleeding) received treatment. On the other hand, approximately 40% of pregnant persons with ITP received unnecessary treatment for moderate asymptomatic thrombocytopenia either during pregnancy or for an epidural. Utilizing clinical practice guidelines would reduce the overtreatment of pregnant persons which would reduce the potential side effects of therapy for the mother and infant.

Supine Sleep-Promoting Baby Mattress: SIDS Prevention and Athlete Injury Relief

Sudden Infant Death Syndrome (SIDS) is a leading cause of death in infants aged one to 12 months, primarily occurring during sleep. Consequently, there is considerable interest among researchers and healthcare professionals regarding the sleeping environment and potential risk factors. One identified risk factor is infants sleeping prone (on their stomach), making supine (on the back) sleeping the recommended position. Associate Research Fellow Shing-Jye Chen from the Department of Sports Science Research at the Taiwan Institute of Sports Science (TISS) is developing a hammock-like air-based baby mattress covered with breathable mesh. This mattress is designed to promote supine sleeping and prevent babies from rolling onto their stomachs during sleep. It features air tubing within tensioned mesh fabric that automatically adjusts to suspend and support the baby’s weight. This design increases the contact surface area between the baby and the mattress, reduces contact pressure, and provides evenly distributed support as the baby grows. Chen takes an interdisciplinary approach, collaborating with experts in baby breathing, sleeping patterns, and biomechanics to develop solutions aimed at reducing the risk of SIDS based on scientific evidence, safety, and effectiveness. In a pilot study, Chen collected data on the mattress’s performance using weights to simulate babies of different ages. His team will explore the relationship between weight and the measured pressure values at different internal air pressures within the inflatable tubing design. The potential application of this product design may extend to injured athletes, particularly in areas where injuries contact the mattress surface, thereby reducing stress.

Insights into the aetiology and temporal trends in cardiac mortality in the young: a 21-year review of national and registry data in the united kingdom

Abstract Background An accurate representation of the incidence and aetiology of cardiac and sudden cardiac death (SCD) is vital for healthcare policymakers to effectively allocate resources for preventative strategies. Purpose We aimed to report on the incidence and causes of cardiac and SCD among individuals under 35-years-old in England and Wales for 21 years (2001-2021), with a focus on identifying any temporal trends. Methods Annual mortality data from the Office for National Statistics (ONS) pertaining to cardiovascular (CV) and possible CV deaths in individuals under 35-years-old were analyzed using International Classification of Diseases-10 codes. Deaths were categorized into four-classes: A1- definitive cardiac deaths without identifiable structural heart disease (consistent with sudden arrhythmic death syndrome, SADS); A2- definitive cardiac deaths with identified structural heart disease; A3- definitive cardiac deaths with indeterminate cause; and B- potential cardiac deaths. Incidence rates were computed based on annual ONS census data. Additionally, insights were gleaned from the Cardiac Risk in the Young Centre for Cardiac Pathology (CRYCCP) nationwide registry. Results The mean annual number of definitive cardiac deaths (classes A1+A2+A3) was 414 (SD 27.4), resulting in an incidence of young cardiac and SCD of 1.68/100,000 individuals/year. Ischaemic heart disease (28.4%), cardiomyopathies (25.9%), and SADS (21.8%) were the most prevalent cardiac conditions. Cardiomyopathies and SADS related mortality peaked in the 10-to-19 age group, while ischaemic heart disease peaked in the 30-to-35 age group. A mean of 573 (SD 76.6) possible cardiac deaths (class B) occurred, these primarily included ill-defined causes (33.5%), epilepsy (32.4%), sudden infant death syndrome (24.2%), and drowning (7%). Definitive and possible cardiac deaths showed a male preponderance (male to female ratios: 2.3:1 and 1.6:1, respectively). A declining trend in cardiac mortality was observed with a 1.3% (95%CI: 1.0%-1.7%) incidence rate reduction/year (p&amp;lt;0.001). This was associated with a 2.5% annual incidence rate reduction for deaths attributed to structural heart disease (p&amp;lt;0.001), but an 8.5% annual incidence rate increase for deaths attributed to SADS (p&amp;lt;0.001). These trends corelated with increasing numbers of deaths referred to the CRYCCP where an expert cardiac pathologist diagnosed SADS in over 52% of cases. Conclusions Over 21 years in England and Wales, national datasets suggest the incidence of young cardiac and SCD is 1.68/100,000/year, with a small but appreciable declining mortality trend. The rise in SADS rates and contemporaneous decline in deaths attributed to structural heart disease may signify increased awareness and accessibility to expert cardiac pathologists. Ensuring accurate mortality coding is crucial for thoroughly assessing first-degree family members in cases of suspected inherited cardiac conditions.

On the interpretability of the SVM model for predicting infant mortality in Bangladesh

BackgroundAlthough machine learning (ML) models are well-liked for their outperformance in prediction, greatly avoided due to the lack of intuition and explanation of their predictions. Interpretable ML is, therefore, an emerging research field that combines the performance and interpretability of ML models to create comprehensive solutions for complex decision-making analysis. Conversely, infant mortality is a global public health concern affecting health, social well-being, socio-economic development, and healthcare services. The study employs advanced interpretable ML techniques to anticipate and understand the factors affecting infant mortality in Bangladesh, overcoming the shortcomings of the conventional logistic regression (LR) model.MethodsBy utilizing the global surrogate model and local individual conditional expectation (ICE) interpretability technique, the interpretable support vector machine (SVM) has been used in this study to reveal significant characteristics of infant mortality using data from the Bangladesh Demographic and Health Survey (BDHS) 2017–18. To investigate intricate decision-making analysis of infant mortality, we adapted SVM and LR techniques with the hyperparameter tuning parameters. These models’ performances were initially assessed using the receiver operating characteristics (ROC) curve, run-time, and confusion matrix parameters with 100 permutations. Afterward, the SVM model’s model-agnostic explanation and the LR model’s interpretation were compared to enhance advanced comprehension for further insights.ResultsThe results of the 100 permutations demonstrated that the LR model (Average: accuracy = 0.9105, precision = NaN, sensitivity = 0, specificity = 1, F1-score = 0, area under the ROC curve (AUC) = 0.6780, run-time = 0.0832) outperformed the SVM model (Average: accuracy = 0.8470, precision = 0.1062, sensitivity = 0.0949, specificity = 0.9209, F1-score = 0.1000, AUC = 0.5632, run-time = 0.0254) in predicting infant mortality, but the LR model had a slower run-time and it was unable to predict any positive cases. The interpretation of LR analysis revealed that infant mortality rates decrease when mothers give birth after over two years, with higher educational attainment, overweight or obese mothers, working mothers, and families with polluted cooking fuel having lower rates. The local ICE interpretability technique, which depicts individual influences on the average likelihood of dying within the first birthday, explores the interpretable SVM model that mothers with normal BMIs, giving birth within two years, using less polluted cooking fuel, working mothers, and having male infant were more likely to experience infant death. The interpretable SVM model based on the global surrogate model also reveals that working mothers who used polluted cooking fuel at home and working women who used less polluted cooking fuel but had a longer period between pregnancies than two years would have higher infant death rates. Even among non-working mothers who used polluted cooking fuel and gave birth within two years of the preceding one, infant death rates were higher.ConclusionsThe interpretable SVM model reveals global interpretations help clinicians understand the entire conditional distribution, while local interpretations focus on specific instances, providing different insights into model behavior. Interpretable ML models aid policymakers, stakeholders, and families in understanding and preventing infant deaths by improving policy-making strategies and establishing effective family counseling services.

Sociodemographic and Pathological Study of Sudden Death Among Neonates and Infants: An Autopsy Based Study

Children dying within the first year of life is referred to as infant mortality and is an important key indicator for a country’s health and standard of living. Every year, there is an estimated 4.5 million infant deaths around the world, accounting for almost 75% of deaths in children younger than 5 years. The studies based on infant deaths are relatively less in number worldwide.This study investigates the sociodemographic and pathological aspects of infant deaths through autopsy data. The objective of the study was to evaluate the pathology and causes of infant/neonatal deaths and analyze related social and demographic factors. A cross-sectional descriptive autopsy study was conducted at the Government Medical College, Kozhikode, covering 66 medicolegal infant death cases from April 2018 to October 2022. Data were collected via questionnaires and analyzed using SPSS software. Respiratory infections (37.14%) were the leading cause of infant deaths, while perinatal complications and congenital abnormalities each accounted for 25% of neonatal deaths. Most deaths occurred in the postneonatal period (77.1%), predominantly among male infants (63%) from rural areas (84.3%). Key findings included consolidation and bronchopneumonia in lung examinations. Significant associations were found with various factors, including the period of infancy, delivery type, co-sleeping history, and prior hospitalizations. Birth weight and histopathological lung findings were linked to sudden infant death syndrome (SIDS).