Predictive Nomogram for ICU Pressure Injury Deterioration from a 278-Patient Study

278 ICU patients with Stage 1 pressure injuries were used to develop a predictive nomogram. The model, based on three admission factors, achieved an AUC of 0.800 in training and 0.785 in validation. Author Chenyan Zhang published this retrospective study on figshare under a CC-BY-4.0 license.

Use Cases

• Developing predictive models for pressure injury progression based on admission variables like diabetes status and albumin levels.
• Validating clinical risk scores using discrimination metrics (AUC) and calibration techniques mentioned in the description.
• Creating clinical decision support tools based on the described nomogram for early intervention in ICU patients.

Strengths

• Model performance is quantified with specific metrics, including an AUC of 0.800 in the training set.
• The study details its cohort size (278 patients) and validation split (70%/30%).
• Key predictors (diabetes, norepinephrine dose, albumin) are identified with odds ratios and confidence intervals.

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 21.6 KB, indicating limited scope.

Provenance

Source

figshare

Collection Method

Retrospective cohort study of eligible ICU patients, with data split into training and validation sets.

Freshness

Last updated 2026-05-18 05:47:43; freshness should be verified.