Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
February 19, 2025Frontiers in PhysiologyOpen Access

The optimal random forest model achieved an area under the curve of 0.83 on the test set for predicting prolonged length of stay after lumbar spinal stenosis surgery, identifying intraoperative blood loss as the most significant contributor.

View Full Paper
Ask AI
Bookmark
Share

Why the study?

Lumbar spinal stenosis causes significant socioeconomic burdens and hospital length of stay evaluates postoperative complications and socioeconomic effects, motivating a machine learning-based tool to predict and interpret the risk of prolonged length of stay after surgery.

Can a machine learning model accurately predict prolonged length of stay after lumbar spinal stenosis surgery?

Population

540 patients registered from the spine surgery department in one hospital

Design

Retrospective cohort study

Key result

The optimal random forest model achieved an area under the curve of 0.83 on the test set for predicting prolonged length of stay after lumbar spinal stenosis surgery, identifying intraoperative blood loss as the most significant contributor.

Authors

PYPaierhati YashengXinjiang Medical UniversityAYAlimujiang YusufuXinjiang Medical UniversityHLHaopeng LuanShandong Provincial Hospital

Discussion

Loading...

Member takes

Overview

May support preoperative risk stratification after LSS surgery; hypothesis-generating and requires prospective validation.

Study Design

Type

Cohort (n=540)

Multicenter

No

Structured PICO

Can a machine learning model accurately predict prolonged length of stay after lumbar spinal stenosis surgery?

P
Population
540 adult patients with lumbar spinal stenosis who underwent open decompression and fusion surgery, evaluated retrospectively to predict prolonged postoperative hospital length of stay.
E
Exposure
Machine learning-based tool (Random Forest model) with SHAP and LIME for interpretable prediction of prolonged length of stay
O
Outcome
Prolonged length of stay (PLOS) after surgery, defined as hospital stay >= 75th percentile for LOS (8 days)

Main Result

Effect estimate: AUC 0.83

A random forest machine learning model combined with SHAP and LIME can accurately predict and interpret the risk of prolonged length of stay after lumbar spinal stenosis surgery.

Limitations

  • Retrospective study design
  • Single-center data collection
  • Potential for unmeasured confounding variables not included in the electronic health records

Cite This Study

Yasheng et al. (2025) conducted a cohort in Lumbar spinal stenosis (n=540). Random Forest machine learning model vs. Other machine learning models was evaluated on Prediction of prolonged length of stay (PLOS) ≥ 75th percentile (AUC 0.83). The optimal random forest model achieved an area under the curve of 0.83 on the test set for predicting prolonged length of stay after lumbar spinal stenosis surgery, identifying intraoperative blood loss as the most significant contributor.

synapsesocial.com/papers/6a86f2be2174986a53ca09bdhttps://doi.org/10.3389/fphys.2025.1542240
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. 1Risk score for postoperative complications in thoracic surgery2012 · 16 citations
  2. 2Study on Damage Accumulation and Life Prediction with Loads below Fatigue Limit Based on a Modified Nonlinear Model2018 · 28 citations
  3. 3Comparison of Support Vector Machine and Artificial Neural Network Systems for Drug/Nondrug Classification2003 · 565 citations
  4. 4An enhanced Genetic Folding algorithm for prostate and breast cancer detection2022 · 3 citations
  5. 5Common pitfalls in statistical analysis: Logistic regression2017 · 670 citations