Mortality Prediction Model for Patients with Mental Disorders and COVID-19 in Henan, China

439 patient records from 10 hospitals in Henan Province, China, used to develop a machine learning model for predicting 31-day mortality. The dataset, shared by Yi Zhang on figshare, includes clinical and laboratory variables for hospitalized patients with mental disorders and confirmed COVID-19. The best-performing neural network model achieved an AUC of 0.911 on an independent test cohort of 131 patients.

Use Cases

• Benchmarking mortality prediction algorithms based on routine clinical and laboratory variables.
• Analyzing feature importance for COVID-19 outcomes in patients with mental disorders using SHAP analysis.
• Training models to identify high-risk patient groups for targeted clinical intervention.
• Studying the impact of specific predictors like concomitant hormone therapy, alkaline phosphatase, and lymphocyte count on survival.

Strengths

• Model validation on an independent test cohort of 131 patients.
• Data from a multicenter cohort of 439 patients across 10 hospitals.
• Use of multiple feature selection methods (LASSO, Boruta, random forest) and eight machine learning algorithms.
• Clear model performance metrics reported, including an AUC of 0.911 for the best model.

Limitations

• Row count for the underlying patient data is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Data reflects a specific regional and temporal context (hospitals in Henan Province, China), which may limit generalizability.

Provenance

Source

figshare

Collection Method

Data were retrospectively collected from hospital records.

Freshness

Last updated 2026-05-26 06:03:47; freshness should be verified.

Geography

Henan Province, China