Explore the vast range of titles available.

Background Despite the adverse health outcomes associated with longer duration diarrhea (LDD), there are currently no clinical decision tools for timely identification and better management of children with increased risk. This study utilizes machine learning (ML) to derive and validate a predictive model for LDD among children presenting with diarrhea to health facilities. Methods LDD was defined as a diarrhea episode lasting ≥ 7 days. We used 7 ML algorithms to build prognostic models for the prediction of LDD among children < 5 years using de-identified data from Vaccine Impact on Diarrhea in Africa study ( N  = 1,482) in model development and data from Enterics for Global Health Shigella study ( N  = 682) in temporal validation of the champion model. Features included demographic, medical history and clinical examination data collected at enrolment in both studies. We conducted split-sampling and employed K-fold cross-validation with over-sampling technique in the model development. Moreover, critical predictors of LDD and their impact on prediction were obtained using an explainable model agnostic approach. The champion model was determined based on the area under the curve (AUC) metric. Model calibrations were assessed using Brier, Spiegelhalter’s z -test and its accompanying p -value. Results There was a significant difference in prevalence of LDD between the development and temporal validation cohorts (478 [32.3%] vs 69 [10.1%]; p  < 0.001). The following variables were associated with LDD in decreasing order: pre-enrolment diarrhea days (55.1%), modified Vesikari score(18.2%), age group (10.7%), vomit days (8.8%), respiratory rate (6.5%), vomiting (6.4%), vomit frequency (6.2%), rotavirus vaccination (6.1%), skin pinch (2.4%) and stool frequency (2.4%). While all models showed good prediction capability, the random forest model achieved the best performance (AUC [95% Confidence Interval]: 83.0 [78.6–87.5] and 71.0 [62.5–79.4]) on the development and temporal validation datasets, respectively. While the random forest model showed slight deviations from perfect calibration, these deviations were not statistically significant (Brier score = 0.17, Spiegelhalter p -value = 0.219). Conclusions Our study suggests ML derived algorithms could be used to rapidly identify children at increased risk of LDD. Integrating ML derived models into clinical decision-making may allow clinicians to target these children with closer observation and enhanced management.

Introduction Stunting affects one-fifth of children globally with diarrhea accounting for an estimated 13.5% of stunting. Identifying risk factors for its precursor, linear growth faltering (LGF), is critical to designing interventions. Moreover, developing new predictive models for LGF using more recent data offers opportunity to enhance model accuracy, interpretability and capture new insights. We employed machine learning (ML) to derive and validate a predictive model for LGF among children enrolled with diarrhea in the Vaccine Impact on Diarrhea in Africa (VIDA) study and the Enterics for Global Heath (EFGH) ― Shigella study in rural western Kenya. Methods We used 7 diverse ML algorithms to retrospectively build prognostic models for the prediction of LGF (≥ 0.5 decrease in height/length for age z-score [HAZ]) among children 6–35 months. We used de-identified data from the VIDA study ( n  = 1,106) combined with synthetic data ( n  = 8,894) in model development, which entailed split-sampling and K-fold cross-validation with over-sampling technique, and data from EFGH-Shigella study ( n  = 655) for temporal validation. Potential predictors ( n  = 65) included demographic, household-level characteristics, illness history, anthropometric and clinical data were identified using boruta feature selection with an explanatory model analysis used to enhance interpretability. Results The prevalence of LGF in the development and temporal validation cohorts was 187 (16.9%) and 147 (22.4%), respectively. Feature selection identified the following 6 variables used in model development, ranked by importance: age (16.6%), temperature (6.0%), respiratory rate (4.1%), SAM (3.4%), rotavirus vaccination (3.3%), and skin turgor (2.1%). While all models showed good prediction capability, the gradient boosting model achieved the best performance (area under the curve % [95% Confidence Interval]: 83.5 [81.6–85.4] and 65.6 [60.8–70.4]) on the development and temporal validation datasets, respectively. Conclusion Our findings accentuate the enduring relevance of established predictors of LGF whilst demonstrating the practical utility of ML algorithms for rapid identification of at-risk children.

Cryptosporidium is a leading cause of moderate-to-severe diarrhea (MSD) in young children in Africa. We examined factors associated with Cryptosporidium infection in MSD cases enrolled at the rural western Kenya Global Enteric Multicenter Study (GEMS) site from 2008-2012. At health facility enrollment, stool samples were tested for enteric pathogens and data on clinical, environmental, and behavioral characteristics collected. Each child's health status was recorded at 60-day follow-up. Data were analyzed using logistic regression. Of the 1,778 children with MSD enrolled as cases in the GEMS-Kenya case-control study, 11% had Cryptosporidium detected in stool by enzyme immunoassay; in a genotyped subset, 81% were C. hominis. Among MSD cases, being an infant, having mucus in stool, and having prolonged/persistent duration diarrhea were associated with being Cryptosporidium-positive. Both boiling drinking water and using rainwater as the main drinking water source were protective factors for being Cryptosporidium-positive. At follow-up, Cryptosporidium-positive cases had increased odds of being stunted (adjusted odds ratio [aOR] = 1.65, 95% CI: 1.06-2.57), underweight (aOR = 2.08, 95% CI: 1.34-3.22), or wasted (aOR = 2.04, 95% CI: 1.21-3.43), and had significantly larger negative changes in height- and weight-for-age z-scores from enrollment. Cryptosporidium contributes significantly to diarrheal illness in young children in western Kenya. Advances in point of care detection, prevention/control approaches, effective water treatment technologies, and clinical management options for children with cryptosporidiosis are needed.

Abstract Background Although Shigella is an important cause of diarrhea in Kenyan children, robust research platforms capable of conducting incidence-based Shigella estimates and eventual Shigella-targeted clinical trials are needed to improve Shigella-related outcomes in children. Here, we describe characteristics of a disease surveillance platform whose goal is to support incidence and consequences of Shigella diarrhea as part of multicounty surveillance aimed at preparing sites and assembling expertise for future Shigella vaccine trials. Methods We mobilized our preexisting expertise in shigellosis, vaccinology, and diarrheal disease epidemiology, which we combined with our experience conducting population-based sampling, clinical trials with high (97%–98%) retention rates, and healthcare utilization surveys. We leveraged our established demographic surveillance system (DSS), our network of healthcare centers serving the DSS, and our laboratory facilities with staff experienced in performing microbiologic and molecular diagnostics to identify enteric infections. We joined these resources with an international network of sites with similar capabilities and infrastructure to form a cohesive scientific network, designated Enterics for Global Health (EFGH), with the aim of expanding and updating our knowledge of the epidemiology and adverse consequences of shigellosis and enriching local research and career development priorities. Conclusions Shigella surveillance data from this platform could help inform Shigella vaccine trials. A diarrhea-specific research platform able to perform population-based enumeration and hospital-based surveillance plus support mentorship of young scientists has been established in western Kenya to support Enterics for Global Health–Shigella surveillance and future vaccine trials.

Abstract Background Accurate estimation of diarrhea incidence from facility-based surveillance requires estimating the population at risk and accounting for case patients who do not seek care. The Enterics for Global Health (EFGH) Shigella surveillance study will characterize population denominators and healthcare-seeking behavior proportions to calculate incidence rates of Shigella diarrhea in children aged 6–35 months across 7 sites in Africa, Asia, and Latin America. Methods The Enterics for Global Health (EFGH) Shigella surveillance study will use a hybrid surveillance design, supplementing facility-based surveillance with population-based surveys to estimate population size and the proportion of children with diarrhea brought for care at EFGH health facilities. Continuous data collection over a 24 month period captures seasonality and ensures representative sampling of the population at risk during the period of facility-based enrollments. Study catchment areas are broken into randomized clusters, each sized to be feasibly enumerated by individual field teams. Conclusions The methods presented herein aim to minimize the challenges associated with hybrid surveillance, such as poor parity between survey area coverage and facility coverage, population fluctuations, seasonal variability, and adjustments to care-seeking behavior. The Enterics for Global Health (EFGH): Shigella Surveillance Study will utilize a hybrid surveillance design to characterize population denominators and healthcare-seeking behavior to calculate Shigella diarrhea incidence rates in children aged 6 to 35 months across sites in Africa, Asia, and Latin America.

Abstract Background Rigorous data management systems and planning are essential to successful research projects, especially for large, multicountry consortium studies involving partnerships across multiple institutions. Here we describe the development and implementation of data management systems and procedures for the Enterics For Global Health (EFGH) Shigella surveillance study—a 7-country diarrhea surveillance study that will conduct facility-based surveillance concurrent with population-based enumeration and a health care utilization survey to estimate the incidence of Shigella­-associated diarrhea in children 6 to 35 months old. Methods The goals of EFGH data management are to utilize the knowledge and experience of consortium members to collect high-quality data and ensure equity in access and decision-making. During the planning phase before study initiation, a working group of representatives from each EFGH country site, the coordination team, and other partners met regularly to develop the data management systems for the study. Results This resulted in the Data Management Plan, which included selecting REDCap and SurveyCTO as the primary database systems. Consequently, we laid out procedures for data processing and storage, study monitoring and reporting, data quality control and assurance activities, and data access. The data management system and associated real-time visualizations allow for rapid data cleaning activities and progress monitoring and will enable quicker time to analysis. Conclusions Experiences from this study will contribute toward enriching the sparse landscape of data management methods publications and serve as a case study for future studies seeking to collect and manage data consistently and rigorously while maintaining equitable access to and control of data. This paper describes the data management systems and processes employed by the Enterics for Global Health: Shigella Surveillance Study (EFGH). Systems and procedures were designed to promote equity in access and decision-making while maintaining high-quality data standards.