Mortality Prediction Model for Patients with Mental Disorders and COVID-19

A dataset of 439 hospitalized patients with mental disorders and laboratory-confirmed COVID-19 was used to develop machine-learning models for predicting 31-day mortality. Data were collected retrospectively from 10 hospitals in Henan Province, China, and the study was authored by Yi Zhang. The dataset was last updated on 2026-05-26.

Use Cases

• Developing mortality risk prediction models based on routine clinical and laboratory variables.
• Comparing the performance of different machine-learning algorithms for clinical outcome prediction.
• Identifying key predictors of mortality using SHAP analysis on features like concomitant hormone therapy and lymphocyte count.
• Conducting survival analysis between model-defined risk groups using Kaplan–Meier methods.

Strengths

• Dataset includes 439 patient records from a multicenter cohort of 10 hospitals.
• Model development used multiple feature selection methods (LASSO, Boruta, random forest) and eight machine-learning algorithms.
• The best-performing model achieved an AUC of 0.911 on an independent test cohort of 131 patients.

Limitations

• Row count for the underlying data is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The data reflects a specific regional cohort from Henan Province, China, which may limit generalizability.

Provenance

Source

Yi Zhang via figshare.

Collection Method

Data were retrospectively collected from hospitalized patients.

Freshness

Last updated 2026-05-26 06:03:48; freshness should be verified.

Geography

Henan Province, China.