Implementation and impact of mhealth in the management of diabetes mellitus in Africa: A systematic review and meta-analysis DOI Creative Commons
Franklin O. Dike, Jean Claude Mutabazi, Ezekiel Musa

et al.

PLOS Digital Health, Journal Year: 2025, Volume and Issue: 4(4), P. e0000776 - e0000776

Published: April 8, 2025

Background The World Health Organization (WHO) has proposed the concept of mobile health to support healthcare systems delivery worldwide. Mobile (mHealth) involves using Information and Communication Technology (ICT) for care provision or services. In context Africa, a region that witnessed significant increase in phone availability usage last decade corresponding rise incidence prevalence diabetes mellitus, this study global implications. We conducted systematic review on extent mHealth implementation managing mellitus Africa. estimated its impact achieving desired glycemic targets, sustained control, preventing complications past decade. Methods analysis studies assessing utilization patients with Africa were considered based PICO method: Population, Intervention, Comparator, Outcomes. MEDLINE, PubMed, SCOPUS, Pan African Clinical Trials Registry searched. Two authors, independent each other, screened titles abstracts retrieved search strategy, full-text articles, assessed them eligibility, extracting data after that. A third reviewer was brought resolve disagreements between two authors by discussion. revised Cochrane Collaboration Risk Bias Tool used assess quality included studies. narrative synthesis extracted done due paucity eligible studies, results summarized meta-analysis. Results None six measured mean FPG percentage changes as primary outcomes. Five change HbA1c from baseline end study. ranged 3.6% 20.53%, significance three meta-analysis overall WMD (95% CI) 0.992 (0.48, 1.50). This, combination high z score 3.822, p <0.001 suggests statistically effect is not likely chance. However, considerable heterogeneity (I 2 = 63.9%, 0.026) present implying observed may be generalizable all differences characteristics case most sample size duration addressed secondary outcomes measuring direct relationships these interventions prevention early detection complications. Conclusion Overall, there reduction levels among individuals living type following interventions. Few heterogeneity. Therefore, we recommend more well-designed randomized controlled trials investigate efficacy management Systematic registration PROSPERO CRD42021218674

Language: Английский

Accurate disaster entity recognition based on contextual embeddings in self-attentive BiLSTM-CRF DOI Creative Commons
Noor E. Hafsa, Hadeel Alzoubi,

Atikah Saeed Almutlq

et al.

PLoS ONE, Journal Year: 2025, Volume and Issue: 20(3), P. e0318262 - e0318262

Published: March 26, 2025

Automated extraction of disaster-related named entities is crucial for gathering pertinent information during natural or human-made crises. Timely and reliable data vital effective disaster management, benefiting humanitarian response authorities, law enforcement agencies, other concerned organizations. Online news media plays a pivotal role in disseminating crisis-related emergencies facilitating post-hazard operations. To extract relevant entities, contextual embedding features prove instrumental. In this study, we investigate the automatic from an annotated dataset 1000 online articles. These articles are carefully with 14 crisis-specific obtained ontologies. generate vector representations words, construct novel word model inspired by Word2vec. features, combined lexicon encoded using contextualized deep Bi-directional LSTM network augmented self-attention conditional random field (CRF) layers. We compare performance our proposed existing approaches. Through extensive evaluation on independent test set 200 that includes more than 80,000 tokens, context-sensitive optimized NER achieves impressive results at sentence level. With Precision 92%, Recall 91%, Accuracy 87%, F1-score outperforms those utilizing general non-contextual embeddings, including fine-tuned BERT models, showcasing its superior performance.

Language: Английский

Citations

0
Implementation and impact of mhealth in the management of diabetes mellitus in Africa: A systematic review and meta-analysis DOI Creative Commons
Franklin O. Dike, Jean Claude Mutabazi, Ezekiel Musa

et al.

PLOS Digital Health, Journal Year: 2025, Volume and Issue: 4(4), P. e0000776 - e0000776

Published: April 8, 2025

Background The World Health Organization (WHO) has proposed the concept of mobile health to support healthcare systems delivery worldwide. Mobile (mHealth) involves using Information and Communication Technology (ICT) for care provision or services. In context Africa, a region that witnessed significant increase in phone availability usage last decade corresponding rise incidence prevalence diabetes mellitus, this study global implications. We conducted systematic review on extent mHealth implementation managing mellitus Africa. estimated its impact achieving desired glycemic targets, sustained control, preventing complications past decade. Methods analysis studies assessing utilization patients with Africa were considered based PICO method: Population, Intervention, Comparator, Outcomes. MEDLINE, PubMed, SCOPUS, Pan African Clinical Trials Registry searched. Two authors, independent each other, screened titles abstracts retrieved search strategy, full-text articles, assessed them eligibility, extracting data after that. A third reviewer was brought resolve disagreements between two authors by discussion. revised Cochrane Collaboration Risk Bias Tool used assess quality included studies. narrative synthesis extracted done due paucity eligible studies, results summarized meta-analysis. Results None six measured mean FPG percentage changes as primary outcomes. Five change HbA1c from baseline end study. ranged 3.6% 20.53%, significance three meta-analysis overall WMD (95% CI) 0.992 (0.48, 1.50). This, combination high z score 3.822, p <0.001 suggests statistically effect is not likely chance. However, considerable heterogeneity (I 2 = 63.9%, 0.026) present implying observed may be generalizable all differences characteristics case most sample size duration addressed secondary outcomes measuring direct relationships these interventions prevention early detection complications. Conclusion Overall, there reduction levels among individuals living type following interventions. Few heterogeneity. Therefore, we recommend more well-designed randomized controlled trials investigate efficacy management Systematic registration PROSPERO CRD42021218674

Language: Английский

Citations

0