Key result
Neural networks predicted chronicity (ICU length of stay > 7 days) more reliably than the statistical model, regardless of architecture.
Why the study?
Do connectionist models predict chronicity more reliably than statistical models in surgical ICU patients?
Population
491 patients in a surgical intensive care unit with length of stay 3 days who survived at least an…
Comparison
Connectionist models vs Statistical model (multiple logistic regression)
Design
Cohort
Authors
Loading...
May aid surgical ICU resource planning; leaves open prospective validation before any practice change.
Observational (n=491)
No
Do connectionist models predict chronicity more reliably than statistical models in surgical ICU patients?
Connectionist models (neural networks) may offer more reliable predictions of clinical trajectory and resource utilization in surgical ICUs compared to traditional statistical models.
Buchman et al. (1994) conducted an observational in Surgical intensive care unit patients (n=491). Neural networks (connectionist models) vs. Multiple logistic regression (statistical model) was evaluated on Chronicity (length of stay > 7 days). Neural networks predicted chronicity (ICU length of stay > 7 days) more reliably than the statistical model, regardless of architecture.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: