Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
January 1, 1998Medical Decision Making

The prognostic accuracy of neural networks was similar to logistic regression (ROC areas 76.0% vs 75.8%), but calibration was better (Hosmer-Lemeshow chi-square 18.6 vs 45.0).

View Full Paper
Ask AI
Bookmark
Share

Why the study?

Does a neural network model improve the prediction of perioperative cardiac complications compared to logistic regression in vascular surgery patients?

Population

1,081 vascular surgery patients (567 in derivation set, 514 in validation set)

Comparison

Neural network risk prediction model based on… vs Logistic regression risk prediction model

Design

Cohort

Follow-up

perioperative

Authors

PLPablo LapuertaGLGilbert L’ItalienSPSumita D. Paul

Discussion

Loading...

Member takes

Overview

May improve calibration of perioperative cardiac risk estimates; extends modeling comparisons but should not yet change practice.

Structured PICO

Does a neural network model improve the prediction of perioperative cardiac complications compared to logistic regression in vascular surgery patients?

P
Population
1,081 vascular surgery patients (567 in derivation set, 514 in validation set)
I
Intervention
Neural network risk prediction model based on cardiac risk factors and dipyridamole thallium results
C
Comparator
Logistic regression risk prediction model
O
Outcome
Perioperative cardiac complications

Neural networks provide better calibration than logistic regression for predicting perioperative cardiac risk in vascular surgery patients, avoiding the overestimation of risk in high-risk groups.

Cite This Study

Lapuerta et al. (1998) studied this question.

synapsesocial.com/papers/6a12d67f92637892a9a7594ehttps://doi.org/10.1177/0272989x9801800114
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Use of a Probabilistic Neural Network to Estimate the Risk of Mortality after Cardiac Surgery1997 · 68 citations
  2. 2Artificial intelligence versus logistic regression statistical modelling to predict cardiac complications after noncardiac surgery1994 · 36 citations
  3. 3Predicting adverse outcomes of cardiac surgery with the application of artificial neural networks2008 · 21 citations
  4. 4Development and validation of a bayesian model for perioperative cardiac risk assessment in a cohort of 1,081 vascular surgical candidates1996 · 188 citations
  5. 5Machine learning vs. traditional methods for predicting postoperative cardiac complications after non‐cardiac surgery: a systematic review and Bayesian network meta‐analysis2026 · 2 citations