Machine Learning Algorithms Versus Classical Regression Models in Pre-Eclampsia Prediction: A Systematic Review.

PURPOSE OF REVIEW: Machine learning (ML) approaches are an emerging alternative for healthcare risk prediction. We aimed to synthesise the literature on ML and classical regression studies exploring potential prognostic factors and to compare prediction performance for pre-eclampsia. RECENT FINDINGS: From 9382 studies retrieved, 82 were included. Sixty-six publications exclusively reported eighty-four classical regression models to predict variable timing of onset of pre-eclampsia. Another six publications reported purely ML algorithms, whilst another 10 publications reported ML algorithms and classical regression models in the same sample with 8 of 10 findings that ML algorithms outperformed classical regression models. The most frequent prognostic factors were age, pre-pregnancy body mass index, chronic medical conditions, parity, prior history of pre-eclampsia, mean arterial pressure, uterine artery pulsatility index, placental growth factor, and pregnancy-associated plasma protein A. Top performing ML algorithms were random forest (area under the curve (AUC) = 0.94, 95% confidence interval (CI) 0.91-0.96) and extreme gradient boosting (AUC = 0.92, 95% CI 0.90-0.94). The competing risk model had similar performance (AUC = 0.92, 95% CI 0.91-0.92) compared with a neural network. Calibration performance was not reported in the majority of publications. ML algorithms had better performance compared to classical regression models in pre-eclampsia prediction. Random forest and boosting-type algorithms had the best prediction performance. Further research should focus on comparing ML algorithms to classical regression models using the same samples and evaluation metrics to gain insight into their performance. External validation of ML algorithms is warranted to gain insights into their generalisability.

Development and validation of a risk prediction model for lost to follow-up among adults on active antiretroviral therapy in Ethiopia: a retrospective follow-up study.

BACKGROUND: Over 420,000 people have initiated life-saving antiretroviral therapy (ART) in Ethiopia; however, lost-to-follow-up (LTFU) rates continues to be high. A clinical decision tool is needed to identify patients at higher risk for LTFU to provide individualized risk prediction to intervention. Therefore, this study aimed to develop and validate a statistical risk prediction tool that predicts the probability of LTFU among adult clients on ART. METHODS: A retrospective follow-up study was conducted among 432 clients on ART in Gondar Town, northwest, Ethiopia. Prognostic determinates included in the analysis were determined by multivariable logistic regression. The area under the receiver operating characteristic (AUROC) and calibration plot were used to assess the model discriminative ability and predictive accuracy, respectively. Individual risk prediction for LTFU was determined using both regression formula and score chart rule. Youden index value was used to determine the cut-point for risk classification. The clinical utility of the model was evaluated using decision curve analysis (DCA). RESULTS: The incidence of LTFU was 11.19 (95% CI 8.95-13.99) per 100-persons years of observation. Potential prognostic determinants for LTFU were rural residence, not using prophylaxis (either cotrimoxazole or Isoniazid or both), patient on appointment spacing model (ASM), poor drug adherence level, normal Body mass index (BMI), and high viral load (viral copies > 1000 copies/ml). The AUROC was 85.9% (95% CI 82.0-89.6) for the prediction model and the risk score was 81.0% (95% CI 76.7-85.3) which was a good discrimination probability. The maximum sensitivity and specificity of the probability of LTFU using the prediction model were 72.07% and 83.49%, respectively. The calibration plot of the model was good (p-value = 0.350). The DCA indicated that the model provides a higher net benefit following patients based on the risk prediction tool. CONCLUSION: The incidence of LTFU among clients on ART in Gondar town was high (> 3%). The risk prediction model presents an accurate and easily applicable prognostic prediction tool for clients on ART. A prospective follow-up study and external validation of the model is warranted before using the model.

Determinants of contraceptive decision making among married women in Sub-Saharan Africa from the recent Demographic and Health Survey data.

BACKGROUND: Different evidence suggested that couples often disagree about the desirability of pregnancy and the use of contraceptives. Increased women's decision-making on contraceptives is identified as a key solution that can change the prevailing fertility and contraceptive utilization pattern in SSA. Therefore, this study aimed to determine determinants of contraceptive decision-making among married women in Sub-Saharan Africa. METHODS: The data source of this study was the standard demographic and health survey datasets of 33 Sub-Saharan Africa countries. Reproductive age group women aged (15-49 years) currently married who are not pregnant and are current users of contraceptive preceding three years the survey was included from the individual record (IR file) file between 2010 and 2018. Since the outcome variable is composed of polychotomous categorical having multiple-choice, the Multinomial logistic regression (MNLR) model was applied. RESULTS: A total of 76,516 married women were included in this study. Maternal age 20-35 and 36-49 years were more likely to had decision making on contraceptive use in both women-only and joint (women and husband/partner category (referance = husband/partner) (RRR = 1.2; 95% CI = 1.05-1.41, RRR = 1.18; 1.04-1.33 and RRR = 1.38; 95% CI = 1.17-1.61, RRR = 1.27; 1.11-1.47)] respectively. Married women with higher education were more likely to decide by women-only category on contraceptive use (referance = husband/partner) (RRR = 1.26; 95% CI = 1.06-1.49). Women only decision-making to use contraceptives relative to the husband/partner only decreases by a factor of 0.86 (95% CI = 0.80-0.93) among rural than urban residences. Women only or joint decision making to use contraceptives was 1.25 and 1.35 times more likely relative to husband/partner decision making respectively among women who had work than that of had no work. The relative risk of women's decision to use family planning relative to husband increased among couples who had a marital duration of ≥ 10 years (RRR = 1.14; 95% CI = 1.06-1.22). But it has no significant effect on joint decision making. Respondents found in the richest wealth index category increase the relative risk of joint decision-making relative to husband/partner (RRR = 1.33; 95% CI = 1.20-1.47) compared to the poorest category. CONCLUSIONS: Decision-making to use contraceptives among married women varies greatly by socio-demographic characteristics. The finding of this study showed that women's age, women educational status, residence, duration of the marriage, family economy, and country income were significantly associated with contraceptive decision-making. Therefore to promote ideal family planning decision making, there is a need to formulate policies and design programs that target women's socio-demographic characteristics and modern contraceptive interventions should be promoted by considering empowering women on decision making.

Spatiotemporal distribution and determinants of overweight or obesity among urban women in Ethiopia: a multivariate decomposition analysis.

BACKGROUND: Overweight /obesity is a global public health concern. It is higher among women than men in most continents of the world. This study aimed to determine the spatiotemporal distribution and determinants of changes in overweight/obesity over time among urban women in Ethiopia. METHODS: We used data from three consecutive Demographic and Health Surveys in Ethiopia (2005, 2011, and 2016). The total weighted sample of 1112 in 2005, 3569 in 2011, and 3071 in 2016 urban women were included in the analysis. The primary outcome measure of this study was the spatiotemporal distribution and trends over time in overweight/obesity. Factors contributing to change in overweight/obesity were examined using a logit-based multivariate decomposition analysis. RESULTS: Overweight/obesity increased from 14.2% in 2005 to 21% in 2016. Approximately 61.3% of the overall increase in overweight/obesity among urban women was due to the difference in coefficient (difference in the effect of characteristics) across the surveys. Changes in the composition of women aged 25-49 years (ß = 0.012, 95% CI 0.008, 0.015), married women (ß = 0.010, 95% CI 0.006, 0.014), women with formal education (primary: ß = 0.007, 95% CI 0.003, 0.011, higher education: ß = 0.014, 95% CI 0.006, 0.022), women with formal employment (ß = 0.006, 95% CI 0.001, 0.011), and women with informal employment (ß = - 0.002, 95% CI - 0.003, - 0.0004) were factors contributing to the change in overweight/obesity from 2005 to 2016. The risk difference (RD) in women's overweight/obesity significantly varied across regions in urban Ethiopia. Furthermore, a high proportion of overweight/obesity was found mainly in Tigray, Oromia, Amhara, and Addis Ababa. CONCLUSIONS: The rate of overweight/obesity among women in urban Ethiopia has shown a significant increase over the last 11 years. This rate change was due to changes in the composition of women's age, educational status, marital status, and employment status. Therefore, program interventions should be targeted at older (> 25 years), educated, married, Addis Ababa residents, and formally employed women.

Prevalence and determinants of overweight/obesity among under-five children in sub-Saharan Africa: a multilevel analysis.

INTRODUCTION: Childhood obesity has become a major public health problem for both developed and developing nations. It is uncommon to find under-nutrition in many low and middle-income countries; as well, obesity is a double burden in these settings. This study aimed to investigate the pooled prevalence of overweight /obesity among under-five (under-5) children in sub-Saharan Africa (SSA). METHODS: Data were accessed from the recent nationally representative demographic and health survey datasets from 33 SSA Countries. A total of 192,132 under-five children were recruited for this study. The pooled prevalence of overweight /obesity among under-5 was done using random-effects meta-analysis command. Multivariable multi-level mixed-effects logistic regression analysis was used to identify determinants for the prevalence of under-5 overweight and/or obesity. A P-value less than 0.05 was used to declare statistical significance. RESULTS: The pooled prevalence of overweight /obesity among under-5 was 5.10% (9% CI: 4.45 - 5.76) in SSA. South Africa region (8.80%, 95% CI: 4.18 - 13.42) had a higher prevalence of under-5 overweight and/or obesity followed by the East Africa region. Male under-5 children (adjusted odds ratio (AOR) = 1.09, 95 confidence interval (CI): 1.02 - 1.25), Larger birth weight under-5 children (AOR = 1.39, 95% CI: 1.26 - 1.54), under-5 children aged older two to three years (AOR = 0.85, 95% CI: 0.76 - 0.94), under-5 children born from educated mothers (secondary and above) (AOR = 1.12, 95% CI: 1.01 - 1.25), and under-5 children living in the West Africa (AOR = 0.67, 95% CI: 0.56 - 0.81) and South Africa (AOR = 1.87, 95% CI: 1.09 - 3.21) were significant determinants for under-5 overweight and/or obesity. CONCLUSION: Childhood obesity is becoming a great challenge and double burden in developing nations. In SSA Africa 1 in 20 under 5 children were overweight and/or obese. Male under-5 children, older aged, under-5 children born from educated mothers, and under-5 children living in the South Africa region were at higher risk for developing overweight and/or obesity. Thus, SSA countries should implement early to pause these consequences preventing the double burden of undernutrition.

Prognostic risk score development to predict birth asphyxia using maternal and fetal characteristics in South Gondar zone hospitals, north West Ethiopia.

BACKGROUND: Birth asphyxia leads to profound systemic and neurological sequela to decrease blood flow or oxygen to the fetus followed by lethal progressive or irreversible life-long pathologies. In low resource setting countries, birth asphyxia remains a critical condition. This study aimed to develop and validate prognostic risk scores to forecast birth asphyxia using maternal and neonatal characteristics in south Gondar zone hospitals. METHODS: Prospective cohorts of 404 pregnant women were included in the model in south Gondar Zone Hospitals, Northwest Ethiopia. To recognize potential prognostic determinants for birth asphyxia, multivariable logistic regression was applied. The model discrimination probability was checked using the receiver operating characteristic curve (AUROC) and the model calibration plot was assessed using the 'givitiR' R-package. To check the clinical importance of the model, a cost-benefit analysis was done through a decision curve and the model was internally validated using bootstrapping. Lastly, a risk score prediction measurement was established for simple application. RESULTS: Of 404, 108 (26.73%) (95% CI: 22.6-31.3) newborns were exposed to birth asphyxia during the follow-up time. Premature rupture of membrane, meconium aspiration syndrome, malpresentation, prolonged labor, Preterm, and tight nuchal was the significant prognostic predictors of birth asphyxia. The AUROC curve for birth asphyxia was 88.6% (95% CI: 84.6-92.2%), which indicated that the tool identified the newborns at risk for birth asphyxia very well. The AUROC of the simplified risk score algorithm, was 87.9 (95% CI, 84.0- 91.7%) and the risk score value of 2 was selected as the optimal cut-off value, with a sensitivity of 78.87%, a specificity of 83.26%, a positive predictive value of 63.23%, and a negative predictive value of 91.52%. CONCLUSIONS: We established birth asphyxia prediction tools by applying non-sophisticated maternal and neonatal characteristics for resource scares countries. The driven score has very good discriminative ability and prediction performance. This risk score tool would allow reducing neonatal morbidity and mortality related to birth asphyxia. Consequently, it will improve the overall neonatal health / under-five child health in low-income countries.

Time to death and its associated factors among infants in sub-Saharan Africa using the recent demographic and health surveys: shared frailty survival analysis.

BACKGROUND: Globally, approximately 4.1 million infants died, accounting for 75% of all under-five deaths. In sub-Saharan Africa (SSA), infant mortality was 52.7/1000 live births in 2018 This study aimed to assess the pooled estimate of infant mortality rate (IMR), time to death, and its associated factors in SSA using the recent demographic and health survey dataset between 2010 and 2018. METHODS: Data were retrieved from the standard demographic and health survey datasets among 33 SSA countries. A total of 93,765 samples were included. The data were cleaned using Microsoft Excel and STATA software. Data analysis was done using R and STATA software. Parametric shared frailty survival analysis was employed. Statistical significance was declared as a two-side P-value < 0.05. RESULTS: The pooled estimate of IMR in SSA was 51 per 1000 live births (95% Confidence Interval (CI): 46.65-55.21). The pooled estimate of the IMR was 53 in Central, 44 in Eastern, 44 in Southern, and 57 in Western Africa per 1000 live births. The cumulative survival probability at the end of 1 year was 56%. Multiple births (Adjusted Hazard ratio (AHR) = 2.68, 95% CI: 2.54-2.82), low birth weight infants (AHR = 1.28, 95% CI: 1.22-1.34), teenage pregnancy (AHR = 1.19, 95 CI: 1.10-1.29), preceding birth interval <  18 months (AHR = 3.27, 95% CI: 3.10-3.45), birth order ≥ four (AHR = 1.14, 95% CI:1.10-1.19), home delivery (AHR = 1.08, 95% CI: 1.04-1.13), and unimproved water source (AHR = 1.07, 95% CI: 1.01-1.13), female sex (AHR = 0.86, 95% CI: 0.83-0.89), immediately breastfeed (AHR = 0.24, 95% CI: 0.23-0.25), and educated mother (AHR = 0.88, 95% CI: 0.82-0. 95) and educated father (AHR = 0.90, 95% CI: 0.85-0.96) were statistically significant factors for infant mortality. CONCLUSION: Significant number of infants died in SSA. The most common cause of infant death is a preventable bio-demographic factor. To reduce infant mortality in the region, policymakers and other stakeholders should pay attention to preventable bio-demographic risk factors, enhance women education and improved water sources.

Time to death and its determinants among under-five children in Sub-Saharan Africa using the recent (2010-2018) demographic and health survey data: country-based shared frailty analyses.

BACKGROUND: Substantial global progress has been made in reducing under-five mortality since 1990, yet progress is insufficient to meet the sustainable development goal of 2030 which calls for ending preventable child deaths. There are disproportional survivals among children in the world. Therefore, the study aimed to assess the Survival status of under-five mortality and determinants in Sub-Saharan African Countries using the recent DHS data. METHODS: The data was retrieved from the birth record file from the standard Demographic and Health Survey dataset of Sub-Saharan Africa countries. Countries that have at least one survey between 2010 and 2018 were retrieved. Parametric shared frailty survival analysis was employed. RESULTS: A total of 27,221 (7.35%) children were died before celebrating their fifth birthday. Children at an early age were at higher risk of dying and then decrease proportionally with increased age. The risk of death among rich and middle family were lowered by 18 and 8% (AHR =0.82, 95% CI: 0.77-0.87) and (AHR = 0.92, 95% CI: 0.87-0.97) respectively, the hazard of death were 11, 19, 17, 90 and 55% (AHR = 1.06, 95% CI: 1.00-1.12), (AHR = 1.11,95%CI:1.04-1.19), (AHR = 1.17, 95% CI:1.12-1.23), (AHR = 1.90, 95%CI: 1.78-2.04) and (AHR = 1.55, 95% CI:1.47-1.63) higher than among children in rural, use unimproved water, delivered at home, born less than 18 months and between 18 and 23 months birth intervals respectively. The hazard of death was 7% among females and low birth weights (AHR = 0.93, 95%CI: 0.90 - 0.97) and (AHR = 0.93 95%CI: 0.89-0.97) respectively. There was also a significant association between multiple births and birth orders (AHR = 2.11, 95%CI: 2.51 - 2.90), (AHR = 3.01, 95%CI: 2.85-3.19) respectively. CONCLUSIONS: Death rate among under-five children was higher at an early age then decreases as age advanced. Wealth status, residence, water source, place of delivery, sex of the child, plurality, birth size, preceding birth interval, and birth order were the most predictor variables. The health care program should be designed to encourage a healthy family structure. The health care providers should intervene in the community to inspire maternal health services.

Spatial distribution of iron rich foods consumption and its associated factors among children aged 6-23 months in Ethiopia: spatial and multilevel analysis of 2016 Ethiopian demographic and health survey.

BACKGROUND: Micronutrient deficiencies are the most prevalent nutritional deficiencies that cause serious developmental problems in the globe. The aim of this study was to assess the spatial distribution of iron rich foods consumption and its associated factors among children aged 6-23 months in Ethiopia. METHODS: The data retrieved from the standard Ethiopian Demographic and Health Survey 2016 dataset with a total sample size of 3055 children aged 6-23 months. Spatial scan statistics done using Kuldorff's SaTScan version 9.6 software. ArcGIS version 10.7 software used to visualize spatial distribution for poor consumption of iron rich foods. Multilevel mixed-effects logistic regression analysis employed to identify the associated factors for good consumption of iron-rich foods. Level of statistical significance was declared at a two-sided P-value < 0.05. RESULTS: Overall, 21.41% (95% CI: 19.9-22.9) of children aged 6-23 months had good consumption of iron rich foods in Ethiopia. Poor consumption of iron rich foods highly clustered at Southern Afar, Southeastern Amhara and Tigray, and the Northern part of Somali Regional States of Ethiopia. In spatial scan statistics, children aged 6-23 months living in the most likely cluster were 21% more likely vulnerable to poor consumption of iron rich foods than those living outside the window (RR = 1.21, P-value < 0.001). Child aged 12-17 months (AOR = 1.90, 95% CI: 1.45-2.49) and 18-23 months (AOR = 2.05, 95% CI: 1.55-2.73), primary (AOR = 1.42, 95% CI:1.06-1.87) and secondary and above (AOR = 2.26, 95% CI: 1.47-3.46) mother's education level, rich (AOR = 1.49, 95% CI: 1.04-2.13) and middle (AOR = 1.83, 95% CI: 1.31-2.57) household wealth status, Amhara (AOR = 0.24, 95% CI: 0.09-0.60), Afar (AOR = 0.38, 95% CI: 0.17-0.84), and Harari (AOR = 2.11, 95% CI: 1.02-4.39) regional states of Ethiopia were statistically significant factors for good consumption of iron rich foods. CONCLUSION: Overall, the consumption of iron rich foods was low and spatially non-random in Ethiopia. Federal Ministry of Health and other stakeholders should give prior attention to the identified hot spot areas to enhance the consumption of iron rich foods among children aged 6-23 months.

Spatio-temporal distribution and associated factors of home delivery in Ethiopia. Further multilevel and spatial analysis of Ethiopian demographic and health surveys 2005-2016.

BACKGROUND: Globally, between 2012 and 2017, 80% of live births occurred at health facilities assisted by skilled health personnel. In Ethiopia, in 2016 only 26% of live births attended by skilled health personal. This study aimed to assess the spatial patterns and associated factors of home delivery in Ethiopia using 2005, 2011, and 2016 Ethiopian Demographic and Health Surveys. METHODS: A total of 33,482 women who gave live birth in the 5 years preceding each survey were included for this study. ArcGIS version 10.7 software was used to visualize the spatial distribution of home delivery. The Bernoulli model was applied using Kilduff SaTScan version 9.6 software to identify significant purely spatial clusters for home delivery in Ethiopia. A multilevel logistic regression model was fitted to identify factors associated with home delivery. A p-value < 0.05 was taken to declare statistically significant predictors. RESULTS: Home delivery was declined from 94.78% in 2005, 90.05% in 2011, and 73.44% in 2016 in Ethiopia. Among the three surveys, consistently high clustering of home delivery was observed in Amhara and Southern Nations Nationalities and People's Regions (SNNPR) of Ethiopia. In spatial scan statistics analysis, a total of 128 clusters (RR = 1.04, P-value < 0.001) in 2005, and 90 clusters (RR = 1.11, P-value < 0.001) in 2011, and 55 clusters (RR = 1.29, P-value < 0.001) in 2016 significant primary clusters were identified. Educational status of women and husband, religion, distance to the health facility, mobile access, antenatal care visit, birth order, parity, wealth index, residence, and Region were statistically significant predictors of home delivery. CONCLUSION: The spatial distribution of home delivery among the three consecutive surveys were non-random in Ethiopia. Educational status of women and husband, religion, distance to the health facility, wealth index, residence, parity, mobile access, Region, and antenatal care visit were statistically significant predictors of home delivery in Ethiopia. Therefore, an intervention needs to improve the coverage of antenatal care visit, and health care facilities. Ministry of health and other stakeholders should give special attention to women living in Amhara and SNNPR states of Ethiopia.

Determinants of person-centered maternity care at the selected health facilities of Dessie town, Northeastern, Ethiopia: community-based cross-sectional study.

BACKGROUND: Person-centered maternity care is providing care that is respectful and responsive to individual women's preferences, needs, and values and ensuring that their values guide all clinical decisions during childbirth. Although person-centered health care is one of the factors that increase client satisfaction and increased health service utilization in Ethiopia, little is known about predictors of person-centered maternity care. Therefore, the aim of this study was to identify the determinant factors of person-centered maternity care among mothers who gave birth in selected health facilities in Dessie town, Northeastern, Ethiopia. METHODS: A community-based cross-sectional study was conducted with a total of 317 study participants at Dessie town selected by a simple random sampling technique. The data was coded and entered Epi-data version 4.4 and analyzed using SPSS version 23. Descriptive statistics was presented using tables and figures. Multivariable linear regression analysis was used to identify factors associated with Person-Centered Maternity Care. Two sides P-value < 0.05 was taken to declare statistically significant. RESULTS: Overall, 310 study participants participated with a response rate of 97.8%. In multivariable linear regression, rural residence (ß = -4.12; 95% CI: -7.60, -0.67), family average monthly income ≤ 3000 birr (ß = -6.20, 95% CI: -9.40, -3.04), night time delivery(ß = -2.98, 95%CI: -5.90, -0.06), dead fetus outcome during delivery (ß = -12.7; 95% CI: -21.80, -3.50), and 2-7 days health facility length of stay (ß = -5.07, 95% CI: -9.20, -0.92) were significantly decreased Person Center Maternity Care score, whereas private health institution delivery (ß = 14.13, 95% CI: 7.70, 20.60) is significantly increased Person centered maternity care score. CONCLUSIONS: This study revealed that most of the factors that affect person-centered maternity care are modifiable factors. Therefore, Primary attention should be given to improve the quality of care through effective communication between clients and providers at each level of the health care delivery system to increase the uptake of high-quality facility-based births.

Prediction of pre-eclampsia with machine learning approaches: Leveraging important information from routinely collected data.

BACKGROUND: Globally, pre-eclampsia (PE) is a leading cause of maternal and perinatal morbidity and mortality. PE prediction using routinely collected data has the advantage of being widely applicable, particularly in low-resource settings. Early intervention for high-risk women might reduce PE incidence and related complications. We aimed to replicate our machine learning (ML) published work predicting another maternal condition (gestational diabetes) to (1) predict PE using routine health data, (2) identify the optimal ML model, and (3) compare it with logistic regression approach. METHODS: Data were from a large health service network with 48,250 singleton pregnancies between January 2016 and June 2021. Supervised ML models were employed. Maternal clinical and medical characteristics were the feature variables (predictors), and a 70/30 data split was used for training and testing the model. Predictive performance was assessed using area under the curve (AUC) and calibration plots. Shapley value analysis assessed the contribution of feature variables. RESULTS: The random forest approach provided excellent discrimination with an AUC of 0.84 (95% CI: 0.82-0.86) and highest prediction accuracy (0.79); however, the calibration curve (slope of 1.21, 95% CI 1.13-1.30) was acceptable only for a threshold of 0.3 or less. The next best approach was extreme gradient boosting, which provided an AUC of 0.77 (95% CI: 0.76-0.79) and well-calibrated (slope of 0.93, 95% CI 0.85-1.01). Logistic regression provided good discrimination performance with an AUC of 0.75 (95% CI: 0.74-0.76) and perfect calibration. Nulliparous, pre-pregnancy body mass index, previous pregnancy with prior PE, maternal age, family history of hypertension, and pre-existing hypertension and diabetes were the top-ranked features in Shapley value analysis. CONCLUSION: Two ML models created the highest-performing prediction using routinely collected data to identify women at high risk of PE, with acceptable discrimination. However, to confirm this result and also examine model generalisability, external validation studies are needed in other settings, utilising standardised prognostic factors.

Micronutrient intake status and associated factors in children aged 6-23 months in sub-Saharan Africa.

Micronutrient deficiency has serious consequences across all ages worldwide, particularly in sub-Saharan Africa. Poor micronutrient (MN) consumption among children remains a major public health concern in many countries. Low literacy, poor diets, a lack of access to healthcare, and insufficient agricultural productivity made proper micronutrient consumption difficult, despite numerous interventions. Thus, this research aimed to determine the adequate intake of MNs among 6-23 months of age children in sub-Saharan Africa. Initially, a two-stage stratified sampling technique was applied for the selected recent demographic and health survey data. The data source was the (2015-2020) standard demographic and health surveys (DHS) among 20 Sub-Saharan African countries. Micronutrient intake status (the outcome variable) was determined using either food rich in Vitamin A or iron consumed within the previous 24 hr or multiple micronutrient powder or iron supplementation within the previous 07 days or vitamin A supplementation or deworming within 06 months. A generalized linear mixed model based on Modified Poisson regression and robust error variance was used to identify factors associated with children's adequate micronutrient intake. An adjusted Prevalence Ratio (aPR) with a 95% confidence interval was used to identify factors, strength and direction of the association. The total samples of 65,187 children aged 6-23 months were included. Of all, 73.99% (95% CI: 73.65, 74.32) of children had adequate intake of micronutrients. Maternal education (primary (aPR = 1.04, 95% CI: 1.02, 1.06) and secondary (aPR = 1.07, 95% CI: 1.04, 1.09)), father's education (primary (aPR = 1.03, 95% CI: 1.01, 1.05) and secondary (aPR = 1.04, 95% CI: 1.02, 1.06)), occupational status of the mother (aPR = 1.04, 95% CI: 1.02, 1.06), pregnancy interval (aPR = 0.97, 95% CI: 0.95, 0.99), exclusive breastfeeding status (aPR = 0.83, 95% CI: 0.82, 0.85), birthweight (average (aPR = 1.03, 95% CI: 1.01, 1.05) and larger than average (aPR = 1.04, 95% CI: 1.02, 1.06)), multiple/twin at birth (aPR = 0.94, 95% CI: 0.91, 0.98), child age (aPR = 1.22, 95% CI: 1.19, 1.25), number of children in home (aPR = 1.02, 95% CI: 1.01, 1.03), ANC utilization (aPR = 1.20, 95% CI: 1.15, 1.27), place of birth (AOR = 0.93, 95% CI: 0.91, 0.95), rich households (aPR = 1.03, 95% CI: 1.01, 1.05), and countries from Central (aPR = 1.07, 95% CI: 1.04, 1.09), South Africa (aPR = 1.07, 95% CI: 1.03, 1.11), and West African (aPR = 0.95, 95% CI: 0.92, 0.99) were associated with level of micronutrients intake status. The prevalence of adequate intake of MN was considerable. Variables at the child, family and community levels were associated with adequate intake of micronutrients. Consequently, stakeholders' involvement is required in healthcare and community settings.

Magnitude and Factors Associated with Gestational Weight Gain Adequacy among Pregnant Women in South Gondar Zone, Northwest Ethiopia.

Background: Weight gain during pregnancy depends on the maternal prepregnancy weight and height. Inappropriate weight gain has negative consequences, including the health care system and society because of its adverse birth outcomes. Objective: This study aimed to assess the magnitude and factors associated with gestational weight gain in Northwest Ethiopia. Methods: From September 2018 to June 2019, a community-based prospective follow-up study was conducted in Northwest Ethiopia. A total of 422 pregnant women were followed from conception to delivery and the data were collected using a multistage sampling technique. Stata 14 standard edition (SE) software was used for data analysis. Multinomial logistic regression was used to determine the relationship between dependent and independent variables. P value of ≤0.05 was used to determine statistical significance. Results: Majority of the participants had normal weight gain [65.12%, 95% confidence interval (CI): 60.08, 69.85]. Besides this, the rate of inadequate and overadequate weight gain was 21.53% (95% CI: 17.60, 26.05) and 13.35% (95% CI: 10.22, 17.25), respectively. Inadequate weight gain was linked to meal frequency [adjusted odd ratio (AOR): 0.52, 95% CI: 0.28, 0.97], targeted supplementary feeding program (TSFP) enrollment (AOR: 2.47; 95% CI: 1.35, 4.50), parity (AOR: 0.18; 95% CI: 0.05, 0.62), and alcohol consumption history (AOR: 0.47; 95% CI: 0.25, 0.88), whereas overadequate weight gain was associated with residency (AOR: 5.22; 95% CI: 2.43, 11.22) and TSFP status (AOR: 2.22; 95% CI: 1.08, 4.57). Conclusions: This study revealed a notable magnitude of both inadequate and overadequate weight gain. It found that good meal frequency, alcohol consumption, parity, and TSFP enrollment were associated with a reduced risk of inadequate weight gain during pregnancy. In addition, the study identified residence and TSFP enrollment as factors linked to overadequate weight gain during pregnancy.

Neonatal outcomes of preterm neonates and its predictors in Ethiopian public hospitals: Multicenter prospective follow-up study.

Introduction: Preterm birth remains the most significant clinical and public health encounter. Preterm infant outcomes pose key evidence for clinicians and policymakers and are extensively used to set clinical and policy verdicts to improve services. It is necessary to conduct the outcomes of neonates frequently, as it varies from place to place and even from time to time in a similar place. There is limited literature in Ethiopia about preterm neonates' outcomes and their predictors. Objective: This study aimed to assess the neonatal outcomes of preterm neonates and their predictors in South Gondar zone public hospitals, Northwest Ethiopia, 2021. Methods: A prospective observational study was employed on 462 preterm neonates in South Gondar Zone Public Hospitals. The data were entered into Epidata 4.6 and analyzed using STATA version 16/MP software. A parametric log-normal survival model was used to identify possible predictors for preterm neonate death. Statistical significance was declared at a P-value less than 0.05. Result: The overall preterm survival rate was 71.1% (95% CI: 66.7, 75.1). Thirty-six percent of preterm neonates were diagnosed with sepsis. One-fourth of the neonates had respiratory distress syndrome. Gestational age greater than 34 weeks (ß = 1.04; 95% CI: 0.53, 1.56), respiratory distress syndrome (ß = 0.85; 95% CI: 0.49, 1.22), body mass index (ß = -1.34; 95% CI: -1.87, -0.80), non-union marital status (ß = -0.71; 95% CI: -1.34, -0.09), multiple pregnancies (ß = -0.66; 95% CI: -0.99-0.32), multiparous (ß = 0.35; 95% CI: 0.01, 0.69), hypothermia (ß = -1.19; 95% CI: -1.76, -0.62), Kangaroo Mother Care (ß = -1.9; 95% CI: -2.34, -1.41) and non-cephalic presentation (ß = -1.23; 95% CI: -1.99,-0.46) were significant predictors. Conclusion: In this study, the preterm survival rate was low. Gestational age greater than 34 weeks, no respiratory distress syndrome, and multiparous mothers were positively associated with the survival of preterm neonates. Though, high pre-pregnancy maternal body mass index, non-union marital status of mothers, multiple pregnancies, hypothermia, Kangaroo mother care is not given, and non-cephalic presentation were negatively associated. A significant focus should be given to implementing WHO recommendations on preventing and caring for preterm births.

Development of a Nomogram for Clinical Risk Prediction of Preterm Neonate Death in Ethiopia.

Introduction: In 2020, over 6,500 newborn deaths occured every day, resulting in 2.4 million children dying in their 1st month of life. Ethiopia is one of the countries that will need to step up their efforts and expedite progress to meet the 2030 sustainable development goal. Developing prediction models to forecast the mortality of preterm neonates could be valuable in low-resource settings with limited amenities, such as Ethiopia. Therefore, the study aims to develop a nomogram for clinical risk prediction of preterm neonate death in Ethiopia in 2021. Methods: A prospective follow-up study design was employed. The data were used to analyze using R-programming version 4.0.3 software. The least absolute shrinkage and selection operator (LASSO) regression is used for variable selection to be retained in the multivariable model. The model discrimination probability was checked using the ROC (AUROC) curve area. The model's clinical and public health impact was assessed using decision curve analysis (DCA). A nomogram graphical presentation created an individualized prediction of preterm neonate risk of mortality. Results: The area under the receiver operating curve (AUROC) discerning power for five sets of prognostic determinants (gestational age, respiratory distress syndrome, multiple neonates, low birth weight, and kangaroo mother care) is 92.7% (95% CI: 89.9-95.4%). This prediction model was particular (specificity = 95%) in predicting preterm death, with a true positive rate (sensitivity) of 77%. The best cut point value for predicting a high or low risk of preterm death (Youden index) was 0.3 (30%). Positive and negative predictive values at the Youden index threshold value were 85.4 percent and 93.3 percent, respectively. Conclusion: This risk prediction model provides a straightforward nomogram tool for predicting the death of preterm newborns. Following the preterm neonates critically based on the model has the highest cost-benefit ratio.

Suboptimal blood pressure control and its associated factors among people living with diabetes mellitus in sub-Saharan Africa: a systematic review and meta-analysis.

BACKGROUND: Suboptimal blood pressure control among people living with diabetes mellitus (DM) is one of the primary causes of cardiovascular complications and death in sub-Saharan Africa (SSA). However, there is a paucity of evidence on the prevalence and associated factors of suboptimal blood pressure control in SSA. Therefore, this review aimed to estimate its pooled prevalence and associated factors among people living with DM in SSA.  METHODS: We systematically searched PubMed, African Journals OnLine, HINARI, ScienceDirect, Google Scholar, and direct Google to access observational studies conducted in SSA. Microsoft Excel spreadsheet was used to extract the data, which was exported into STATA/MP version 16.0 for further analyses. Heterogeneity across studies was checked using Cochran's Q test statistics and I2 test, and small study effect was checked using Funnel plot symmetry and Egger's statistical test at a 5% significant level. A random-effects model was used to estimate the pooled prevalence and associated factors of suboptimal blood pressure control at a 95% confidence interval (CI) and significance level of p < 0.05. RESULTS: Of the 7329 articles retrieved, 21 articles were eligible for the meta-analysis. After performing random-effects model, the pooled prevalence of suboptimal blood pressure control was 69.8% (95% CI: 63.43, 76.25%). Poor adherence to antihypertensive treatment (OR = 1.7; 95% CI: 1.03-2.80, I2 = 0.0%, p = 0.531) and overweight (OR = 2.4, 95% CI: 1.57-3.68, I2 = 0.00%, p = 0.47) were significantly associated with suboptimal blood pressure control. CONCLUSIONS: The prevalence of suboptimal blood pressure control among diabetic patients in SSA was high, and poor adherence to antihypertensive treatment and overweight were significantly associated with suboptimal blood pressure control. Hence, there is an urgent need for initiatives to improve and control hypertension, and preventive measures should concentrate on modifiable risk factors. SYSTEMATIC REVIEW REGISTRATION: PROSPERO CRD42020187901.

The pooled estimate of the total fertility rate in sub-Saharan Africa using recent (2010-2018) Demographic and Health Survey data.

Background: Even though the total fertility rate (TFR) has decreased significantly over the past decades in many countries, it has remained stable in sub-Saharan African (SSA) countries. However, there is variation among the sub-regions and inhabitants of SSA. Therefore, this study aimed to conduct a meta-analysis of demographic and health surveys (DHS) to estimate the pooled level of TFR in SSA and to depict sub-regional and inhabitant differences. Methods: The data source for this study was the standard Demographic and Health Survey datasets of 33 sub-Saharan African countries, accessed through www.meaasuredhs.com between 2010 and 2018. The point estimate of the total fertility rate with its corresponding standard error in each sub-Saharan African country was estimated using the DHS.rates R package. Using the point estimate of the TFR with the standard error of each country, the pooled estimate of the TFR was generated by the metan STATA command. Results: The study comprised 1,324,466 live births in total. The pooled estimate of sub-Saharan Africa's overall fertility rate was five children per woman (95% CI: 4.63-5.37). Consequently, the pooled estimate of total fertility for people living in urban and rural areas was 3.90 (95% CI: 3.60-4.21) and 5.82 (95% CI: 5.43-6.21) children per woman, respectively. In sub-group analysis, the pooled estimates of the TFR for the East African, Central African, Southern African, and West African regions, respectively, were 4.74, 5.59, 3.18, and 5.38 children per woman. Total fertility rates were greater in low-income nations (5.45), lower-middle-income countries (4.70), and high-middle-income countries (3.80). Conclusions: SSA has a relatively high total fertility rate. The regions of West and Central Africa have the highest overall fertility rate. The fertility rate is higher in countries with a large rural population and low income. Strategies should be developed to address this public health concern, especially in rural Central and Western Africa.

Internet Addiction and Its Associated Factors Among African High School and University Students: Systematic Review and Meta-Analysis.

Introduction: Internet addiction is characterized by excessive and uncontrolled use of the internet affecting everyday life. Adolescents are the primary risk group for internet addiction. Data on internet addiction is lacking in Africa. Thus, this review aimed to determine the pooled prevalence of internet addiction and its associated factors among high school and university students in Africa. Methods: A comprehensive literature search was conducted using electronic databases (PubMed/MEDLINE, Web of science, Hinari, and Google scholar) to locate potential studies. Heterogeneity between studies was checked using Cochrane Q test statistics and I2 test statistics and small-study effects were checked using Egger's statistical test at a 5% significance level. A sensitivity analysis was performed. A random-effects model was employed to estimate the pooled prevalence and associated factors of internet addiction among students. The primary outcome of measure of this review was the prevalence of internet addiction and the secondary outcome of measures are the factors associated with internet addiction. Results: A total of 5,562 studies were identified among the five databases. Of these, 28 studies from 10 countries with 14,946 high school and university students were included in this review. The overall pooled prevalence of internet addiction among the students was 34.53% [95% Confidence Interval (CI): 26.83, 42.23, I2 = 99.20%]. Male sex [Pooled Odds Ratio (POR) = 1.92, 95% CI:1.43, 2.57 I2 = 0.00], urban residence (POR = 2.32, 95% CI:1.19, 4.53, I2 = 59.39%), and duration of daily internet use for more than 4 h (POR = 2.25, 95% CI:1.20, 4.21, I2 = 0.00%, were significantly associated with internet addiction among adolescents. Conclusion: Almost one-third of university and high school students in Africa are addicted to the internet. Male students, those from urban areas, and those who use the internet for more than 4 h per day have higher odds of internet addiction. Thus, we recommend that health planners and policymakers pay attention to the use of the internet and internet addiction in Africa.

The effect of age on the incidence of COVID-19 complications: a systematic review and meta-analysis.

BACKGROUND: The coronavirus (COVID-19) pandemic was reported from Wuhan, China, on December 31, 2019, and the pandemic was spread to more than 212 countries in the globe. This meta-analysis aimed to assess the pooled incidence of COVID-19 complications and to identify the association between the incidence of complications and age. METHODS: Comprehensive databases, PubMed, Hinari, and Google Scholar, were used to locate potential articles for this review. Data were extracted using Microsoft Excel and imported to the STATA/MP version 16.0 software for analysis. Heterogeneity between studies was assessed using the Cochrane Q test statistics and I2 test, and small study effect was checked using Egger's statistical test at 5% significant level. Sensitivity analysis was checked. A random-effects model was conducted to estimate the pooled incidence of COVID-19 complications. Univariate meta-regression was conducted to identify the association between the mean ages with each complication. RESULTS: From the total of 1237 studies, 12 studies were included with a total of 3064 COVID-19 patients. The most complications were acute respiratory distress syndrome (30.93%, 95%CI 21.3-40.6%) followed by acute liver injury (22.8%, 95%CI 14-31.5%), shock (10.9%, 95%CI 7.4-14.4%), acute kidney injury (7%, 95%CI 3.8-10.4%), and acute cardiac injury (6.4%, 95%CI 2.8-15.6%). Univariate meta-regression revealed that as the mean age increased by 1 year, the incidence of acute respiratory distress syndrome, acute kidney injury, acute cardiac injury, and shock increased by a factor of 2.9 (ß = 2.9, 95%CI 2.4-3.4, adjusted R2 = 88), 0.4 (ß = 0.4, 95%CI 0.04-0.72, adjusted R2 = 54), 1.6 (ß = 1.6, 95%CI 1.1-2.1, adjusted R2 = 85), and 1.1 (ß = 1.1, 95%CI 0.8-1.5, adjusted R2 = 26) times respectively. CONCLUSION: Significant complications of COVID-19 viral infections were reported. Older populations were a high-risk group of developing adverse complications as compared to their counterparts. Health care professionals should give primary attention to those risk group individuals.