CURB-65 + PLR: 30-Day Mortality Predictions for Severe Pneumonia Patients

Chengtian Lv published a dataset on 2026-06-02 containing results from a retrospective study of 214 adult patients with severe community-acquired pneumonia (SCAP). The data likely contains variables used to evaluate the predictive value of combining the CURB-65 score with the platelet-to-lymphocyte ratio (PLR) for 30-day all-cause mortality. The study reported a 30-day mortality rate of 25.7% and compared the predictive performance of CURB-65 alone (AUC=0.702) versus CURB-65 combined with PLR (AUC=0.772).

Use Cases

• Training a binary classification model to predict 30-day mortality based on clinical scores and blood count ratios.
• Benchmarking the performance of the CURB-65 score against enhanced models that incorporate hematological markers like PLR.
• Conducting survival analysis to stratify patient risk groups (low, medium, high) based on CURB-65 scores.
• Evaluating model calibration and discrimination using metrics like AUC, sensitivity, and specificity reported in the study.

Strengths

• Dataset is based on a study of 214 adult patients, providing a specific clinical cohort size.
• Results include statistically significant performance metrics, such as an AUC of 0.772 for the combined model.
• Clear temporal outcome defined as 30-day all-cause mortality, with a reported rate of 25.7%.

Limitations

• Row count and column-level documentation are absent; field semantics must be inferred after download.
• The dataset is very small at 70.5 KB, indicating limited scope and sample size.
• Data may reflect temporal and selection bias inherent to a single retrospective study from an unspecified emergency department.

Provenance

Source

figshare, author Chengtian Lv

Collection Method

Retrospective study of emergency department patients.

Freshness

Last updated 2026-06-02 04:21:27; freshness should be verified.