Predictive Nomogram for ICU Pressure Injury Deterioration from a Retrospective Study

A retrospective cohort study of 278 ICU patients with Stage 1 pressure injuries, aiming to predict deterioration. The dataset, authored by Chenyan Zhang and last updated in May 2026, was used to develop a nomogram based on three admission factors. The model demonstrated an AUC of 0.800 in the training set and 0.785 in the validation set.

Use Cases

• Developing predictive models for pressure injury progression based on patient admission data.
• Validating clinical risk scores using discrimination metrics like AUC and calibration.
• Building tools for early ICU patient risk stratification based on diabetes status, norepinephrine dose, and albumin levels.

Strengths

• Model performance is quantified with specific metrics, including an AUC of 0.800 in training and 0.785 in validation.
• The study clearly defines a cohort of 278 patients split into training (n=195) and validation (n=83) sets.
• Predictors are identified with specific odds ratios and confidence intervals (e.g., diabetes OR: 2.451; 95% CI: 1.139–5.274).

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the underlying data is unknown, which may limit suitability assessment.
• The dataset is very small at 17.2 KB, indicating limited scope.

Provenance

Source

figshare

Collection Method

Retrospective cohort study of ICU patients.

Freshness

Last updated 2026-05-18 05:47:44; freshness should be verified.