{
  "abstract": "ObjectiveTo combine performance-weighted ensemble SHAP aggregation with resampling-based assessment of feature-importance variability and patient-level attribution alignment for interpreting COVID-19 mortality predictions.\n\nMethodsWe analyzed a prospective cohort of 1,857 patients hospitalized with COVID-19 at two hospitals in Peru. Ten predictive algorithms were evaluated using five-fold cross-validation, and their mean AUROC values determined their attribution-aggregation weights. Within each model and training trial, signed SHAP values were normalized by the mean total absolute attribution across evaluation patients before performance-weighted aggregation. Feature importance was summarized across 30 resampled training trials using means and normal-approximation 95% confidence intervals. Patient-level attribution alignment was assessed using cosine similarity between signed feature-attribution vectors for corresponding patients, with whole-trial patient-correspondence randomization and Benjamini-Hochberg correction. Subgroup observations were reweighted toward the evaluated populations feature distributions for secondary comparisons of existing absolute SHAP values.\n\nResultsAmong the 1,857 included patients, 982 (52.9%) died during hospitalization. Random forest achieved the highest mean AUROC (0.925 {+/-} 0.010), followed by AdaBoost (0.919 {+/-} 0.017) and logistic regression (0.918 {+/-} 0.007), with variability reported as the SEM. Attribution profiles were more similar across repeated training trials of the same algorithm (mean Pearson r = 0.783; SEM, 0.004) than across algorithms within the same trial (mean Pearson r = 0.369; SEM, 0.006). The largest normalized attribution magnitudes were observed for dexamethasone use at home without oxygen support (6.186%; 95% CI, 5.795-6.577), PaO2/FiO2 ratio (3.014%; 95% CI, 2.787-3.242), shortness of breath (2.774%; 95% CI, 2.592-2.956), and FiO2 (2.560%; 95% CI, 2.284-2.836). Features differed in patient-level attribution alignment, indicating that importance magnitude and robustness provided complementary information. Among 1,763 eligible subgroup-feature comparisons, 64 had nominal one-sided p < 0.05, although none remained below 0.05 after Benjamini-Hochberg adjustment.\n\nConclusionsAlgorithms with similar predictive performance produced substantially different feature-attribution profiles. Normalized performance-weighted SHAP aggregation provided a representative relative-importance summary across algorithms, while resampling-based alignment assessment qualified the consistency of individual feature contributions. The resulting attribution patterns describe fitted-model behavior and should not be interpreted as causal clinical effects.",
  "abstract_shown": true,
  "affiliations": [
    "Global Health Institute, Michigan State University, MI, USA"
  ],
  "article_type": "PUBLISHAHEADOFPRINT",
  "arxiv": null,
  "author_total": 7,
  "authors": [
    {
      "name": "N. Dai"
    },
    {
      "affs": [
        0
      ],
      "name": "R. K. Briceno"
    },
    {
      "name": "A. N. Castaneda"
    },
    {
      "name": "M. A. Tresierra"
    },
    {
      "name": "M. L. Esteban"
    },
    {
      "name": "M. E. Rosas"
    },
    {
      "name": "R. C. Hinojosa"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/bt93fzr6k1m4h.bib",
  "cited_by": 0,
  "doi": "10.64898/2026.09.07.26362459",
  "fulltext_url": "https://www.medrxiv.org/content/10.64898/2026.09.07.26362459v1.full",
  "id": "bt93fzr6k1m4h",
  "issns": [],
  "landing_url": "https://www.medrxiv.org/content/10.64898/2026.09.07.26362459v1",
  "language": null,
  "licence": "CC-BY-NC-ND",
  "licence_label": "CC BY-NC-ND",
  "licence_source": "article",
  "licence_url": "https://creativecommons.org/licenses/by-nc-nd/4.0/",
  "metadata_licence": null,
  "native_subjects": [
    [
      "medrxiv",
      "public and global health"
    ],
    [
      "openalex",
      "1702|Explainable Artificial Intelligence (XAI)"
    ]
  ],
  "openalex": "W7212042145",
  "pdf_url": "https://www.medrxiv.org/content/10.64898/2026.09.07.26362459v1.full.pdf",
  "pmcid": null,
  "pmid": null,
  "published": "2026-09-07",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/bt93fzr6k1m4h.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:18:04Z",
      "native_id": "10.64898/2026.09.07.26362459",
      "source": "medrxiv",
      "url": "https://www.medrxiv.org/content/10.64898/2026.09.07.26362459v1"
    }
  ],
  "subjects": [
    {
      "field": "medicine",
      "origin": "medrxiv:public and global health",
      "subfield": "epidemiology-public-health"
    }
  ],
  "title": "Ensemble SHAP Aggregation and Attribution Variability in Clinical Machine Learning: A COVID-19 Mortality Study",
  "title_text": "Ensemble SHAP Aggregation and Attribution Variability in Clinical Machine Learning: A COVID-19 Mortality Study",
  "type": "preprint",
  "updated_at": "2026-10-30T12:18:04Z",
  "url": "https://research.draconitic.com/paper/bt93fzr6k1m4h",
  "venue": "medRxiv",
  "versions": []
}