Empowering child health: Harnessing machine learning to predict acute respiratory infections in Ethiopian under-fives using demographic and health survey insights.

BACKGROUND: A dearth of studies showed that infectious diseases cause the majority of deaths among under-five children. Worldwide, Acute Respiratory Infection (ARI) continues to be the second most frequent cause of illness and mortality among children under the age of five. The paramount disease burden in developing nations, including Ethiopia, is still ARI. OBJECTIVE: This study aims to determine the magnitude and predictors of ARI among under-five children in Ethiopia using used state of the art machine learning algorithms. METHODS: Data for this study were derived from the 2016 Ethiopian Demographic and Health Survey. To predict the determinants of acute respiratory infections, we performed several experiments on ten machine learning algorithms (random forests, decision trees, support vector machines, Naïve Bayes, and K-nearest neighbors, Lasso regression, GBoost, XGboost), including one classic logistic regression model and an ensemble of the best performing models. The prediction ability of each machine-learning model was assessed using receiver operating characteristic curves, precision-recall curves, and classification metrics. RESULTS: The total ARI prevalence rate among 9501 under-five children in Ethiopia was 7.2%, according to the findings of the study. The overall performance of the ensemble model of SVM, GBoost, and XGBoost showed an improved performance in classifying ARI cases with an accuracy of 86%, a sensitivity of 84.6%, and an AUC-ROC of 0.87. The highest performing predictive model (the ensemble model) showed that the child's age, history of diarrhea, wealth index, type of toilet, mother's educational level, number of living children, mother's occupation, and type of fuel they used were an important predicting factor for acute respiratory infection among under-five children. CONCLUSION: The intricate web of factors contributing to ARI among under-five children was identified using an advanced machine learning algorithm. The child's age, history of diarrhea, wealth index, and type of toilet were among the top factors identified using the ensemble model that registered a performance of 86% accuracy. This study stands as a testament to the potential of advanced data-driven methodologies in unraveling the complexities of ARI in low-income settings.

Guidelines and Standard Frameworks for AI in Medicine: Protocol for a Systematic Literature Review.

BACKGROUND: Applications of artificial intelligence (AI) are pervasive in modern biomedical science. In fact, research results suggesting algorithms and AI models for different target diseases and conditions are continuously increasing. While this situation undoubtedly improves the outcome of AI models, health care providers are increasingly unsure which AI model to use due to multiple alternatives for a specific target and the "black box" nature of AI. Moreover, the fact that studies rarely use guidelines in developing and reporting AI models poses additional challenges in trusting and adapting models for practical implementation. OBJECTIVE: This review protocol describes the planned steps and methods for a review of the synthesized evidence regarding the quality of available guidelines and frameworks to facilitate AI applications in medicine. METHODS: We will commence a systematic literature search using medical subject headings terms for medicine, guidelines, and machine learning (ML). All available guidelines, standard frameworks, best practices, checklists, and recommendations will be included, irrespective of the study design. The search will be conducted on web-based repositories such as PubMed, Web of Science, and the EQUATOR (Enhancing the Quality and Transparency of Health Research) network. After removing duplicate results, a preliminary scan for titles will be done by 2 reviewers. After the first scan, the reviewers will rescan the selected literature for abstract review, and any incongruities about whether to include the article for full-text review or not will be resolved by the third and fourth reviewer based on the predefined criteria. A Google Scholar (Google LLC) search will also be performed to identify gray literature. The quality of identified guidelines will be evaluated using the Appraisal of Guidelines, Research, and Evaluation (AGREE II) tool. A descriptive summary and narrative synthesis will be carried out, and the details of critical appraisal and subgroup synthesis findings will be presented. RESULTS: The results will be reported using the PRISMA (Preferred Reporting Items for Systematic Review and Meta-Analyses) reporting guidelines. Data analysis is currently underway, and we anticipate finalizing the review by November 2023. CONCLUSIONS: Guidelines and recommended frameworks for developing, reporting, and implementing AI studies have been developed by different experts to facilitate the reliable assessment of validity and consistent interpretation of ML models for medical applications. We postulate that a guideline supports the assessment of an ML model only if the quality and reliability of the guideline are high. Assessing the quality and aspects of available guidelines, recommendations, checklists, and frameworks-as will be done in the proposed review-will provide comprehensive insights into current gaps and help to formulate future research directions. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): DERR1-10.2196/47105.

Assessing the FAIRness of Deep Learning Models in Cardiovascular Disease Using Computed Tomography Images: Data and Code Perspective.

The interest in the application of AI in medicine has intensely increased over the past decade with most of the changes in the past five years. Most recently, the application of deep learning algorithms in prediction and classification of cardiovascular diseases (CVD) using computed tomography (CT) images showed promising results. The notable and exciting advancement in this area of study is, however, associated with different challenges related to the findability (F), accessibility(A), interoperability(I), reusability(R) of both data and source code. The aim of this work is to identify reoccurring missing FAIR-related features and to assess the level of FAIRness of data and models used to predict/diagnose cardiovascular diseases from CT images. We evaluated the FAIRness of data and models in published studies using the RDA (Research Data Alliance) FAIR Data maturity model and FAIRshake toolkit. The finding showed that although AI is anticipated to bring ground breaking solutions for complex medical problems, the findability, accessibility, interoperability and reusability of data/metadata/code is still a prominent challenge.

Navigating the AI frontiers in cardiovascular research: a bibliometric exploration and topic modeling.

Artificial intelligence (AI) has emerged as a promising field in cardiovascular disease (CVD) research, offering innovative approaches to enhance diagnosis, treatment, and patient outcomes. In this study, we conducted bibliometric analysis combined with topic modeling to provide a comprehensive overview of the AI research landscape in CVD. Our analysis included 23,846 studies from Web of Science and PubMed, capturing the latest advancements and trends in this rapidly evolving field. By employing LDA (Latent Dirichlet Allocation) we identified key research themes, trends, and collaborations within the AI-CVD domain. The findings revealed the exponential growth of AI-related research in CVD, underscoring its immense potential to revolutionize cardiovascular healthcare. The annual scientific publication of machine learning papers in CVD increases continuously and significantly since 2016, with an overall annual growth rate of 22.8%. Almost half (46.2%) of the growth happened in the last 5 years. USA, China, India, UK and Korea were the top five productive countries in number of publications. UK, Germany and Australia were the most collaborative countries with a multiple country publication (MCP) value of 42.8%, 40.3% and 40.0% respectively. We observed the emergence of twenty-two distinct research topics, including "stroke and robotic rehabilitation therapy," "robotic-assisted cardiac surgery," and "cardiac image analysis," which persisted as major topics throughout the years. Other topics, such as "retinal image analysis and CVD" and "biomarker and wearable signal analyses," have recently emerged as dominant areas of research in cardiovascular medicine. Convolutional neural network appears to be the most mentioned algorithm followed by LSTM (Long Short-Term Memory) and KNN (K-Nearest Neighbours). This indicates that the future direction of AI cardiovascular research is predominantly directing toward neural networks and image analysis. As AI continues to shape the landscape of CVD research, our study serves as a comprehensive guide for researchers, practitioners, and policymakers, providing valuable insights into the current state of AI in CVD research. This study offers a deep understanding of research trends and paves the way for future directions to maximiz the potential of AI to effectively combat cardiovascular diseases.

Disparities in Regional Publication Trends on the Topic of Artificial Intelligence in Biomedical Science Over the Last Five Years: A Bibliometric Analysis.

Bibliometric analysis is a scientific method that allows researchers to explore the current trend in a certain research area using citation information. This study aims to provide a meta-view of artificial intelligence studies focused on biomedicine in the last five years, which will provide an insight into current trends and future research directions. Besides the observation of increased publication rates in the area of AI in biomedicine, the results indicate a lower contribution from and a sparser network connectivity of countries with limited resources. Thus, working toward collaboration in terms of infrastructure and implementing alternative solutions such as FAIR (Findable, Accessible, Interoperable and Reproducible) and open access platforms could improve the collaborative nature of international health projects.

Treatment Satisfaction and Its Associated Factors of Dolutegravir Based Regimen in a Resource Limited Setting.

INTRODUCTION: Treatment satisfaction is patient's evaluation of the process of taking the medication and its use. Currently dolutegravir based regimen is first-line agent for treatment of human immunodeficiency virus. But evidence is scarce regarding treatment satisfaction. Thus, the aim of the current study was to assess treatment satisfaction and associated factors of dolutegravir based regimen among adult human immunodeficiency virus patients attending at Debre Markos referral 2020. METHODS: Institutional-based cross-sectional study was conducted. A systematic random sampling technique was used to collect data from June 25 to August 25, 2020 at Debre Markos referral hospital. It was entered into Epi Info and exported to SPSS version 23 for analysis. Bivariable and multivariable logistic regression was used to identify factors. Variables with p<0.05 were considered as statistically significant. RESULTS: From a total of 359, 349 participants (97.2%) responded to the study. In this study, 70.5% of participants reported higher treatment satisfaction. Monthly average income of ≥3500 birr (AOR: 2.88; 95% CI: 1.26, 6.58), 1600-2500 birr (AOR: 2.47; 95% CI: 1.11, 5.48), 800-1600 birr (AOR: 3.11; 95% CI: 1.31, 7.37), positive belief about medications (AOR: 3.05; 1.76, 5.28), having a discussion with health care providers (AOR: 3.05, 95% CI: 1.58, 5.88), patients without concurrent medication (AOR: 7.72, 95% CI: 3.29, 18.07), and being male (AOR: 2.10, 95% CI: 1.14, 3.87) were associated with treatment satisfaction. CONCLUSION: Overall, dolutegravir based regimen showed high treatment satisfaction. Monthly income, positive beliefs about medications, discussing about treatment options, sex and concurrent medications were associated with treatment satisfaction. Thus, it is crucial to improve treatment satisfaction by promoting positive belief towards medication and also by engaging patients in treatment decisions.

Healthcare providers' acceptance of telemedicine and preference of modalities during COVID-19 pandemics in a low-resource setting: An extended UTAUT model.

BACKGROUND: In almost all lower and lower middle-income countries, the healthcare system is structured in the customary model of in-person or face to face model of care. With the current global COVID-19 pandemics, the usual health care service has been significantly altered in many aspects. Given the fragile health system and high number of immunocompromised populations in lower and lower-middle income countries, the economic impacts of COVID-19 are anticipated to be worse. In such scenarios, technological solutions like, Telemedicine which is defined as the delivery of healthcare service remotely using telecommunication technologies for exchange of medical information, diagnosis, consultation and treatment is critical. The aim of this study was to assess healthcare providers' acceptance and preferred modality of telemedicine and factors thereof among health professionals working in Ethiopia. METHODS: A multi-centric online survey was conducted via social media platforms such as telegram channels, Facebook groups/pages and email during Jul 1- Sep 21, 2020. The questionnaire was adopted from previously validated model in low income setting. Internal consistency of items was assessed using Cronbach alpha (α), composite reliability (CR) and average variance extracted (AVE) to evaluate both discriminant and convergent validity of constructs. The extent of relationship among variables were evaluated by Structural equation modeling (SEM) using SPSS Amos version 23. RESULTS: From the expected 423 responses, 319 (75.4%) participants responded to the survey questionnaire during the data collection period. The majority of participants were male (78.1%), age <30 (76.8%) and had less than five years of work experience (78.1%). The structural model result confirmed the hypothesis "self-efficacy has a significant positive effect on effort expectancy" with a standardized coefficient estimate (ß) of 0.76 and p-value <0.001. The result also indicated that self-efficacy, effort expectancy, performance expectancy, facilitating conditions and social influence have a significant direct effect on user's attitude toward using telemedicine. User's behavioral intention to use telemedicine was also influenced by effort expectancy and attitude. The model also ruled out that performance expectancy, facilitating conditions and social influence does not directly influence user's intention to use telemedicine. The squared multiple correlations (r2) value indicated that 57.1% of the variance in attitude toward using telemedicine and 63.6% of the variance in behavioral intention to use telemedicine is explained by the current structural model. CONCLUSION: This study found that effort expectancy and attitude were significantly predictors of healthcare professionals' acceptance of telemedicine. Attitude toward using telemedicine systems was also highly influenced by performance expectancy, self-efficacy and facilitating conditions. effort expectancy and attitude were also significant mediators in predicting users' acceptance of telemedicine. In addition, mHealth approach was the most preferred modality of telemedicine and this opens an opportunity to integrate telemedicine systems in the health system during and post pandemic health services in low-income countries.

Healthcare providers' digital competency: a cross-sectional survey in a low-income country setting.

BACKGROUND: Healthcare providers across all clinical practice settings are progressively relying and adapting information communication technologies to perform their professional activities. In this era of technology, healthcare providers especially in lower income countries should have at least basic digital competency if a successful application of technology is to be achieved. The aim of this study was to assess digital competency of healthcare providers among seven public health centers in North-West Ethiopia. METHODS: A cross-sectional study design was applied to assess the basic digital competency of healthcare providers working in seven public health centers in North-west Amhara regional state, Ethiopia. Self-administered questionnaire adopted from the European commission's digital competency framework for assessing digital competency were used. A multivariable logistic regression was performed to identify factors associated with basic digital competency with p-value< 0.05 as a rule out for statistical significance. The strength of association was explained in terms of coefficient estimate, adjusted odds ratio and a 95% confidence interval (CI). RESULT: From the total of 193 healthcare providers included in the study, 167 of them responded which is a response rate of 86.5%. The majority of respondents 88 (52.7%) were males and the mean age was 28.2 years with a standard deviation of 5.5 years. The result indicated that all items demonstrated an adequate level of internal consistency with Cronbach alpha > 0 .7. Healthcare providers in those public health centers reported that problem solving, safety and communication are the most common challenges encountered. The multivariable logistic regression model indicated that factors such as sex, educational status, profession type, monthly income and years of experience are statistically significant predictors. CONCLUSION: Basic digital competency level of healthcare providers working in public health centers in this setting is relatively low. The results highlight the need to improve digital competency among healthcare providers focusing on the identified skill gaps.

E-health literacy and associated factors among chronic patients in a low-income country: a cross-sectional survey.

BACKGROUND: Chronic patients persistently seek for health information on the internet for medication information seeking, nutrition, disease management, information regarding disease preventive actions and so on. Consumers ability to search, find, appraise and use health information from the internet is known as eHealth literacy skill. eHealth literacy is a congregate set of six basic skills (traditional literacy, health literacy, information literacy, scientific literacy, media literacy and computer literacy). The aim of this study was to assess eHealth literacy level and associated factors among internet user chronic patients in North-west Ethiopia. METHODS: Institutional based cross-sectional study design was conducted. Stratified sampling technique was used to select 423 study participants among chronic patients. The eHealth literacy scale (eHEALS) was used for data collection. The eHEALS is a validated eight-item Likert scaled questionnaire used to asses self-reported capability of eHealth consumers to find, appraise, and use health related information from the internet to solve health problems. Statistical Package for Social science version 20 was used for data entry and further analysis. Multivariable logistic regression was used to examine the association between the eHealth literacy skill and associated factors. Significance was obtained at 95% CI and p < 0.05. RESULT: In total, 423 study subjects were approached and included in the study from February to May, 2019. The response rate to the survey was 95.3%. The majority of respondents 268 (66.3%) were males and mean age was 35.58 ± 14.8 years. The multivariable logistic regression model indicated that participants with higher education (at least having the diploma) are more likely to possess high eHealth literacy skill with Adjusted Odds Ratio (AOR): 3.48, 95% CI (1.54, 7.87). similarly, being government employee AOR: 1.71, 95% CI (1.11, 2.68), being urban resident AOR: 1.37, 95% CI (0.54, 3.49), perceived good health status AOR: 3.97, 95% CI (1.38, 11.38), having higher income AOR: 4.44, 95% CI (1.32, 14.86), Daily internet use AOR: 2.96, 95% CI (1.08, 6.76), having good knowledge about the availability and importance of online resources AOR: 3.12, 95% CI (1.61, 5.3), having positive attitude toward online resources AOR: 2.94, 95% CI (1.07, 3.52) and higher level of computer literacy AOR: 3.81, 95% CI (2.19, 6.61) were the predictors positively associated with higher eHealth literacy level. CONCLUSION: Besides the mounting indication of efficacy, the present data confirm that internet use and eHealth literacy level of chronic patients in this setting is relatively low which clearly implicate that there is a need to fill the skill gap in eHealth literacy among chronic patients which might help them in finding and evaluating relevant online sources for their health-related decisions.

Validation of the Ethiopian Version of eHealth Literacy Scale (ET-eHEALS) in a Population with Chronic Disease.

BACKGROUND: Although the measurement scale developed by Norman and Skinner is the widely used scale to assess consumers' eHealth literacy, translating and validating the scale for the language of the target population under consideration is necessary. Amharic is the official national language of Ethiopia, with 29.3% of native speakers. METHODS: The total sample size calculated was 187 with 6% non-response rate. The internal consistency of the ET-eHEALS was measured using Cronbach's alpha coefficient. Test-retest reliability was assessed by re-administering the ET-eHEALS questionnaire to 74 patients which is 40% of the total sample size. Construct validity was evaluated using exploratory factor analysis. The Kaiser-Meyer-Olkin (KMO) statistic and Bartlett's test of sphericity were used to check the suitability of performing the factor analysis. RESULTS: Of the respondents, 63.1% (n=118) were males and 55.1% (n=103) were aged between 18 and 35 years, with 57.2% (n=107) of the participants being educated to high school diploma level or higher. Cronbach's alpha coefficient for the translated ET-eHEALS total score was 0.94. Test-retest reliability of ET-eHEALS total score was acceptable for both agreements and consistent with ICC (interclass correlation coefficient) of 0.92. The KMO ratio of sampling appropriateness was acceptable (0.91), and Bartlett's test of sphericity was significant with p < 0.001. The EFA (exploratory factor analysis) extracted two factors based on an extraction principle of a minimum eigenvalue of one. The extracted factor explained 80.2% of the common variance which is 51.8% for factor 1 and 28.4% for factor 2. Except for item, item fit for both infit and outfit mean squares were within the adequate range (0.5-1.5). CONCLUSION: This study depicted that ET-eHEALS is a consistent and valid instrument to evaluate Amharic-speaking chronic patients' eHealth literacy level. Since there is no prior validation of eHEALS in low-income country, this finding may indicate important directions for further improvement in eHEALS item performance in resource-limited settings.

Internet use and eHealth literacy among health-care professionals in a resource limited setting: a cross-sectional survey.

BACKGROUND: Health-care professionals should be able to identify and use reputable health care-information sources from the Internet and other relevant sources of information, in order to make good medical decisions. The level in health professional eHealth literacy and the extent of Internet use in a resource-constrained setting is not well documented. The aim of this study was to assess the extent of Internet use and eHealth literacy among a cross section of health-care professionals at the University of Gondar Comprehensive Specialized Hospital, northwest Ethiopia. METHODS: An institution-based cross-sectional study was conducted to assess Internet use and eHealth literacy among health professionals working at the hospital from November 20 to January 17, 2018. Descriptive analysis was used to describe Internet use and eHealth literacy. Multivariable logistic regression was done to identify which factors were associated with the eHealth literacy of participants. RESULTS: In total, 291 study subjects were approached and included in the study, with a response rate of 98.6%. The majority of respondents were female (53.7%) and the mean age was 30.09±5.025 years. Only 47.4% of survey respondents said that they used the Internet regularly for professional/medical updates. The mean eHealth literacy was 27.840±5.691. The majority of participants with high eHealth literacy were aged 21-29 years. and females were slightly more literate regarding eHealth than males (33.1%). Age, type of profession, salary, and years of experience were significantly associated with eHealth literacy. CONCLUSION: The present data confirm that Internet use and eHealth literacy of health professionals is noticeably good, which clearly suggests that there is an opportunity for eHealth to be integrated in the health-care system in tertiary-health facilities in northern Ethiopia if appropriate training and education is provided.