by Halid Worku Jemil, Sonia Worku Semayneh, Altaseb Beyene Kassaw, Kassahun Dessie Gashu
IntroductionSevere stunting is one of the primary public health challenges in LMIC including Eastern African Countries, which affects millions of children. In addition, it was a major contributor for mortality and related complication of children aged under five. However, there is limited study conducted severe form of stunting by employing Machine learning (ML) in Eastern African Countries. Therefore, our study was demonstrated to predict and identify its major determinants using ML algorithms, furthermore, to improve model explainablity. Our study used Shapley Additive explanations (SHAP) and ARM to identify the determinants of severe stunting among under-five.
Methodscross-sectional study was conducted using DHS data from 2012–2022 in East Africa. 136,074 children were the source populations, and 76,019 children were the study population. Data were analyzed using Python version 3.7 and R version 4.3.3 for data preprocessing, modeling, and statistical analysis. Model performance was evaluated using accuracy and AUC. Furthermore, the SHAP analysis and ARM was used to further explain and interpret the determinants of severe stunting among children under five.
ResultsThe Random Forest performed the best in this analysis, with an accuracy of 87% and an AUC score of 0.83. The analysis indicated that women’s who do not practicing exclusive breastfeeding (SHAP value = +0.41), being from Burundi (SHAP value = +0.04), children being underweight (SHAP value = +0.25), lived in poor household (SHAP value = +0.40), child gender being male(SHAP value = +0.23), mothers height being short (SHAP value = +0.03), mothers being underweight (SHAP value = +0.18), child size at birth being small (SHAP value = +0.21), women’s being delivered in home(SHAP value = +0.07), mothers education being primary (SHAP value = +0.20), unimproved toilet (SHAP value = +0.06), distance to health facility being a big problem (SHAP value = +0.02), were associated with increase the risk of severe stunting among under five.
ConclusionThe Random Forest was the best-performing model for predicting severe stunting in Eastern African countries. To decrease the effects of severe stunting, integrated interventions should provide support for mothers with lower socioeconomic conditions, strengthen maternal education, empower women to practice exclusive breastfeeding, encourage facility deliveries, increase access for households to sanitary facilities, provide education on personal and environmental hygiene, provide mothers with information on the importance of complementary feeding for children as well as for the mothers, and provide near health facilities for mothers and essential care services.
This study investigated district-level variations in the impact of COVID-19 on maternal, neonatal and child healthcare (MNCH) service utilisation, delivery and health outcomes in Gauteng, one of the hardest-hit provinces in South Africa.
A cross-sectional quantitative study.
We collected District Health Information System data for MNCH services from all 493 public healthcare facilities across all five districts in Gauteng province. We applied simple linear regression to assess key performance indicators before (March 2019 to February 2020) and during (March 2020 to February 2021) the COVID-19 pandemic. A pooled multiple linear regression model compared the impact of the pandemic in each district with that of the Johannesburg reference district. The regression models followed the bootstrap approach. Analyses were performed in Stata V.17.0.
Regarding service utilisation, primary headcount under 5 years (n) significantly decreased in all five districts during COVID-19. The effect was greater in Johannesburg (–20 954.5, 95% CI –28 913.3 to –12 995.7; p
In Gauteng province, the COVID-19 pandemic caused a heterogeneous adverse impact on MNCH service utilisation, delivery and health outcomes across the districts. Recognising the geographical differences in the effects of outbreaks and pandemics is critically important for informed decision-making to support healthcare services recovery in affected areas and for planning against future crises.
Curable sexually transmitted infections (STIs) heavily rely on laboratory testing methods. Unfortunately, these diagnostic tools are infrequently used in certain regions of the country, which often results in suboptimal treatment for these infections. This study aimed to assess the prevalence of selected curable STIs among pregnant women.
Cross-sectional study.
The study was conducted in an antenatal care (ANC) unit in one of Southern Ethiopia’s general hospitals.
A total of 244 consecutive pregnant women attending the ANC follow-up were recruited for the study. All pregnant women have equal opportunity to participate; however, women who declined to be interviewed or give a sample were excluded. Pregnant women on antibiotic treatments were also excluded.
The study assessed the prevalence of selected curable STIs and associated factors. Neisseria gonorrhoea was diagnosed by culture, trichomoniasis by microscopic examination and syphilis by serological testing using a rapid diagnostic test cassette. Test results for trichomoniasis and gonorrhoea were obtained from vaginal and endocervical swabs. Factors associated with curable STIs were evaluated by bivariable and multivariate logistic regression.
The overall prevalence of curable STIs was 16.4% (40/244), with prevalence of 2% for gonorrhoea, 15.2% for trichomoniasis and 1.2% for syphilis. Alcohol intake (adjusted OR (AOR)=3.0; 95% CI 1.1 to 8.3; p=0.030), symptomatic treatment (AOR=3.6; 95% CI 1.4 to 8.6; p=0.004), residency (AOR=3.2; 95% CI 1.2 to 9.1; p=0.022) and pain while urinating (AOR=4.0; 95% CI 1.6 to 9.7; p=0.002) were all found to be associated with the existence of these STIs. This study has limitations; the cross-sectional study design and small sample size would provide limited information about factors associated with curable maternal STIs.
The study reveals a high prevalence of curable STIs among the participants, highlighting the need for further research aimed at improving their management during pregnancy.
This study aims to assess the burden and predictors of age-related macular degeneration (AMD) among older age patients with diabetes attending comprehensive specialised hospitals in Northwest Ethiopia.
A multicentre cross-sectional study was conducted among older patients with diabetes using a systematic random sampling technique.
The study was conducted at five comprehensive specialised hospitals in Northwest Ethiopia from 8 May to 8 June 2023.
The study included 832 diabetic individuals aged 40 years and above.
Data were collected using a pretested structured questionnaire and physical examinations.
In this study, a total of 832 participants were involved, with a response rate of 96.85%. The burden of AMD was 15.4% (95% CI 13.0% to 18.0%). Male sex (adjusted OR (AOR) 2.04, 95% CI 1.17 to 3.56), older age (AOR 6.91, 95% CI 3.17 to 15.08), diabetes duration of 10 and more years (AOR 3.00, 95% CI 1.91 to 4.69), higher body mass index (AOR 2.56, 95% CI 1.15 to 5.71), presence of hypertension (AOR 2.45, 95% CI 1.56 to 3.85) and family history of diabetes mellitus (DM) (AOR 2.29, 95% CI 1.40 to 3.76) were positively associated with AMD.
This study found that the prevalence of AMD among patients with diabetes was 15.4%. Older age, male sex, longer DM duration, higher body mass index, presence of hypertension and family history of DM were significantly associated with AMD. Targeted screening of at-risk individuals for AMD, public health awareness campaigns focusing on these factors and further research to understand the burden and underlying mechanisms of these associations with AMD are recommended.
To compare the quality and time efficiency of physician-written summaries with customised large language model (LLM)-generated medical summaries integrated into the electronic health record (EHR) in a non-English clinical environment.
Cross-sectional non-inferiority validation study.
Tertiary academic hospital.
52 physicians from 8 specialties at a large Dutch academic hospital participated, either in writing summaries (n=42) or evaluating them (n=10).
Physician writers wrote summaries of 50 patient records. LLM-generated summaries were created for the same records using an EHR-integrated LLM. An independent, blinded panel of physician evaluators compared physician-written summaries to LLM-generated summaries.
Primary outcome measures were completeness, correctness and conciseness (on a 5-point Likert scale). Secondary outcomes were preference and trust, and time to generate either the physician-written or LLM-generated summary.
The completeness and correctness of LLM-generated summaries did not differ significantly from physician-written summaries. However, LLM summaries were less concise (3.0 vs 3.5, p=0.001). Overall evaluation scores were similar (3.4 vs 3.3, p=0.373), with 57% of evaluators preferring LLM-generated summaries. Trust in both summary types was comparable, and interobserver variability showed excellent reliability (intraclass correlation coefficient 0.975). Physicians took an average of 7 min per summary, while LLMs completed the same task in just 15.7 s.
LLM-generated summaries are comparable to physician-written summaries in completeness and correctness, although slightly less concise. With a clear time-saving benefit, LLMs could help reduce clinicians’ administrative burden without compromising summary quality.
This study aimed to explore adverse childhood experiences (ACEs) and their association with common mental disorders (CMD) among college students in Ethiopia.
Cross-sectional study.
Addis Ababa University, College of Health Sciences.
A total of 345 participants completed the whole questionnaire.
The study used a stratified random sampling technique. Data were gathered through self-administered questionnaires. The instruments used included adapted sociodemographic questions, the ACEs International Questionnaire, the Patient Health Questionnaire, the Generalized Anxiety Disorder Scale and a brief tool for assessing substance use. To examine the relationship between ACEs and various independent variables, both binary and multivariate logistic regression analyses were employed.
In the total sample (n=345), the participant’s mean age was 22.2 (± 2.03), with the majority being females (58%). About 16% of the participants reported depression symptoms and 14.2% had anxiety. The majority of the participants (80%) had at least one ACE and one quarter (25.2%) of the participants had experienced four or more ACEs. The most prevalent type of ACE was community violence (35.4%). One fifth (20%) of the participants had reported having experienced childhood sexual abuse. After controlling for confounding variables, those with four or more ACEs were 6.17 times (adjusted OR (aOR) 6.17; 2.51, 15.18) and 6.0 times (aOR 6.0; 2.25, 16.02) more likely to have depression and anxiety, respectively.
There was a dose-response relationship between ACEs and both anxiety and depression. Identifying and preventing ACEs at an early stage could contribute to reduce depression and anxiety among young people. Efforts to prevent ACEs should target not only individuals but also extend to households and communities.
This study aimed to assess functional disability and associated factors among people with severe mental illness attending public hospitals in Harar town, Eastern Ethiopia.
An institution-based cross-sectional study was conducted in public hospitals found in Harar town, Eastern Ethiopia, from 1 April to 10 May 2023.
A total of 342 patients with severe mental illness attending public hospitals in Harar town, Eastern Ethiopia, were included.
The main outcome of this study was functional disability, which was measured using a 12-item version of the WHO Disability Assessment Schedule.
In this study, the mean (±) score of functional disability among patients with severe mental illness was 31.8 (±9.6) (95% CI: 30.7 to 32.8). The multiple linear regression model revealed that unemployment (β: 1.93; 95% CI: 0.28 to 3.59), total illness duration (5–10 years and above 10 years) (β: 3.71; 95% CI: 1.95 to 5.48 and β: 4.51; 95% CI: 2.41 to 6.62, respectively), age at illness onset (β: –0.37; 95% CI: –0.45 to –0.28), having drug side effects (β: 3.05; 95% CI: 1.46 to 4.63), medication non-adherence (β: 6.15; 95% CI: 4.56 to 7.73) and having high perceived stigma (β: 2.27; 95% CI: 0.72 to 3.82) were significantly associated with functional disability.
Patients with severe mental illness had a high mean functional disability score. Unemployment, age at onset of mental illness, total duration of illness, medication side effects, medication non-adherence and higher perceived stigma were factors that were associated with functional disability. Given the mean functional disability score observed among respondents, a multipronged intervention strategy that prioritises early detection and management of severe mental illness, particularly targeting individuals at risk, is recommended to mitigate disability and enhance quality of life.