Malnutrition Risk Factors in Chinese Schizophrenia Inpatients 2019

March to September 2019 data from 499 patients with schizophrenia at the Shanghai Third Mental Health Center. It contains clinical characteristics and laboratory parameters used to assess nutritional risk via the NRS2002 tool, with 22.44% of patients identified as at risk.

Use Cases

• Build a logistic regression model to predict NRS2002 nutritional risk category using age, BMI, and hemoglobin levels.
• Analyze correlations between length of hospital stay and laboratory indices like total protein and albumin.
• Train a classifier using BMI and age as features to identify patients at high nutritional risk.
• Conduct ROC curve analysis to evaluate the discriminative performance of BMI for nutritional risk prediction.

Strengths

• Data from 499 patients provides a substantive sample for a single-site clinical study.
• Includes both clinical characteristics and multiple laboratory parameters (hemoglobin, total protein, albumin).
• Clear temporal coverage from March to September 2019.

Limitations

• Single-site data from one center in Shanghai limits geographic generalizability across China.
• 42.5 KB file size indicates a small, aggregated summary dataset, not raw patient-level records.
• Unknown row and column structure prevents assessment of feature granularity.

Provenance

Source

figshare, author Shijing Yu.

Collection Method

Clinical study enrolling hospitalized patients, collecting clinical and lab data, using NRS2002 for nutritional risk assessment.

Time Range

March to September 2019.

Freshness

Last updated 2026-03-25, though underlying data is from 2019.

Geography

Shanghai Third Mental Health Center, China.