Skip to content
DraconiticResearch
Search Sources API
182,544 studies31 sourcesUpdated 5 min ago
Peer-reviewedCC BY-NC-NDDiagnostic pathology

Development of a Bayesian subjective model for predicting the clinical diagnosis of Ebola in the Democratic Republic of the Congo.

  • John Kamwina Kebela1
  • Prince Kimpanga2
  • Jean Nyandwe2
  • Jack Kokolomami2
  • Steve Bwira3
  • Rostin Mabela4
  • Olivier Mangapi5
  • Berthe Barhayiga6
  • Godfroid Musema2
  • Bienvenu Kabasele7
  • Sylvain Munyanga3
7 affiliations
  1. 1Department of Epidemiology and Biostatistics, School of Public Health, Faculty of Medicine, University of Kinshasa, Kinshasa, BP 11850 Kin I, Democratic Republic of Congo. [email protected].
  2. 2Department of Epidemiology and Biostatistics, School of Public Health, Faculty of Medicine, University of Kinshasa, Kinshasa, BP 11850 Kin I, Democratic Republic of Congo.
  3. 3Department of Management, School of Public Health, University of Kinshasa, Kinshasa, Democratic Republic of Congo.
  4. 4Department of Mathematics and Computer Science, Faculty of Science, University of Kinshasa, Kinshasa, Democratic Republic of Congo.
  5. 5Educational and Vocational Guidance Department, Ilebo Higher Pedagogical Institute, Ilebo, Democratic Republic of Congo.
  6. 6Department of Anaesthesia and Intensive Care, University Clinics, University of Kinshasa, Kinshasa, Democratic Republic of Congo.
  7. 7Department of Public Health Emergencies and Operations Center (NIPH/PHEOC), National Institute of Public Health, Kinshasa, Democratic Republic of Congo.
Read full textPDFDOICite

Abstract

Background

Early clinical diagnosis of Ebola virus disease (EVD) is challenging in outbreak settings due to symptom overlap with malaria. We assessed whether a Bayesian subjective model (BSM) outperforms logistic regression model (LRM) in improving Ebola screening.Methods

The BSM, incorporating seven expert-identified predictive factors (hemorrhagic, neurological, digestive, pain, general symptoms, epidemiological link, respiratory), underwent internal and external validation using data from 450 patients triaged in the Democratic Republic of Congo. Diagnostic performance metrics (Sensitivity (Se), Specificity (Sp), Positive Predictive Value (PPV), Negative Predictive Value (NPV), overall accuracy) and likelihood ratios (LR) from BSM were compared to those of LRM. Significant odds ratios (OR) (95% CI excluding 1; p < 0.05) were identified via logistic regression analysis. Microsoft Excel and SPSS were utilized for statistical analyses.Results

Experts identified seven predictive factors for Ebola, with four (hemorrhagic, neurological, digestive signs, and epidemiological contact; L.R > 1.5) showing strong likelihood ratios (Table 3). Patients presenting both hemorrhagic signs and epidemiological link had a 60.5% probability of confirmed Ebola. The BSM exhibited higher sensitivity (84.2% vs. 71.7%) compared to LRM under real-world conditions. Significant predictors included hemorrhagic signs (OR 0.13, 95% CI 0.07-0.23, p < 0.001), pain signs (OR 2.06, 95% CI 1.19-3.58, p = 0.01), general signs (OR 3.92, 95% CI 1.76-8.73, p < 0.001), and epidemiological link (OR 0.54, 95% CI 0.30-0.96, p = 0.04); their confidence intervals excluded 1, confirming statistical significance.Conclusions

The BSM offers diagnostic superiority in triage of EVD cases, particularly in resource-limited outbreak settings. Further multicenter studies incorporating biological data are recommended to enhance its robustness and operational validity.

Related

  1. In randomized controlled trials with nutrient supplementation, participants with subjective memory complaints should not be combined with those having a mild cognitive impairment diagnosis
    Mélanie Plourde, Michael Stiffel
    The American Journal of Clinical NutritionPeer-reviewedLicence not stated1 citedPsychiatry & mental healthHAL
  2. Child malnutrition in the Democratic Republic of the Congo: A Bayesian Gaussian regression analysis using the 2023–24 demographic and health survey
    Zacharie Tsala Dimbuene, Hugues Sampasa-Kanyinga
    PLOS OnePeer-reviewedCC BY 4.0Full textNutritionPLOS
  3. Relationship between subjective socioeconomic status and depression in Chinese rural adolescents: positive youth development as a moderator
    Gaoran Chen, Zheng Zhou, Wenqi Chen +1
    BMC Public HealthPeer-reviewedCC BY-NC-ND 4.0Psychiatry & mental healthCrossref
  4. Subjective word-finding complaints in older adults: Development and preliminary psychometric evaluation of a novel self-reported questionnaire
    F. El-Khaldi, A. Marier, P. Vitali +3
    medRxivPreprintAll rights reservedFull textPsychiatry & mental healthmedRxiv
  5. Development of a Qualitative Questionnaire for Indoor Air Quality Monitoring: A Scoping Review of Subjective Assessments in Studies Using Low-Cost Sensors
    Yong Yu, Alice Laghezza, Gaetano Settimo +2
    AtmospherePeer-reviewedCC BYFull textPharmacology & toxicologyDOAJ

Cite

BibTeXRIS
BibTeX
@article{kebela2026development,
  title = {{Development of a Bayesian subjective model for predicting the clinical diagnosis of Ebola in the Democratic Republic of the Congo.}},
  author = {John Kamwina Kebela and Prince Kimpanga and Jean Nyandwe and Jack Kokolomami and Steve Bwira and Rostin Mabela and Olivier Mangapi and Berthe Barhayiga and Godfroid Musema and Bienvenu Kabasele and Sylvain Munyanga},
  journal = {Diagnostic pathology},
  year = {2026},
  month = aug,
  doi = {10.1186/s13000-026-01825-4},
  url = {https://europepmc.org/article/MED/42972355},
  note = {CC BY-NC-ND},
}
RIS
TY  - JOUR
TI  - Development of a Bayesian subjective model for predicting the clinical diagnosis of Ebola in the Democratic Republic of the Congo.
AU  - John Kamwina Kebela
AU  - Prince Kimpanga
AU  - Jean Nyandwe
AU  - Jack Kokolomami
AU  - Steve Bwira
AU  - Rostin Mabela
AU  - Olivier Mangapi
AU  - Berthe Barhayiga
AU  - Godfroid Musema
AU  - Bienvenu Kabasele
AU  - Sylvain Munyanga
PY  - 2026
DA  - 2026/08/06
JO  - Diagnostic pathology
DO  - 10.1186/s13000-026-01825-4
UR  - https://europepmc.org/article/MED/42972355
AB  - Background  Early clinical diagnosis of Ebola virus disease (EVD) is challenging in outbreak settings due to symptom overlap with malaria. We assessed whether a Bayesian subjective model (BSM) outperforms logistic regression model (LRM) in improving Ebola screening.Methods  The BSM, incorporating seven expert-identified predictive factors (hemorrhagic, neurological, digestive, pain, general symptoms, epidemiological link, respiratory), underwent internal and external validation using data from 450 patients triaged in the Democratic Republic of Congo. Diagnostic performance metrics (Sensitivity (Se), Specificity (Sp), Positive Predictive Value (PPV), Negative Predictive Value (NPV), overall accuracy) and likelihood ratios (LR) from BSM were compared to those of LRM. Significant odds ratios (OR) (95% CI excluding 1; p < 0.05) were identified via logistic regression analysis. Microsoft Excel and SPSS were utilized for statistical analyses.Results  Experts identified seven predictive factors for Ebola, with four (hemorrhagic, neurological, digestive signs, and epidemiological contact; L.R > 1.5) showing strong likelihood ratios (Table 3). Patients presenting both hemorrhagic signs and epidemiological link had a 60.5% probability of confirmed Ebola. The BSM exhibited higher sensitivity (84.2% vs. 71.7%) compared to LRM under real-world conditions. Significant predictors included hemorrhagic signs (OR 0.13, 95% CI 0.07-0.23, p < 0.001), pain signs (OR 2.06, 95% CI 1.19-3.58, p = 0.01), general signs (OR 3.92, 95% CI 1.76-8.73, p < 0.001), and epidemiological link (OR 0.54, 95% CI 0.30-0.96, p = 0.04); their confidence intervals excluded 1, confirming statistical significance.Conclusions  The BSM offers diagnostic superiority in triage of EVD cases, particularly in resource-limited outbreak settings. Further multicenter studies incorporating biological data are recommended to enhance its robustness and operational validity.
SN  - 1746-1596
ER  -