A Machine Learning-derived Clinical Score Predicts Passive Ventilation 48 Hours After Initiation of Invasive Mechanical Ventilation
Bibliographic record
Abstract
Abstract Introduction: Neurostimulation of the diaphragm through an indwelling catheter placed in the internal jugular vein and stimulating the diaphragm via the bilateral phrenic nerves can help to ameliorate these negative effects of inactivity of the diaphragm during invasive mechanical ventilation (IMV). There is considerable interest in identifying which patients are at risk for ongoing passive ventilation without diaphragm activity. The objective of this study was to use a machine learning model to develop a clinical score that predicts risk of passive ventilation at 48 hours after initiation of IMV for patients with acute hypoxemic respiratory failure (AHRF). Methods: A database of patients who underwent IMV was retrospectively queried to identify subjects with recorded occluded inspiratory airway pressure (Pocc), a measure of passive versus spontaneous breathing during IMV. Passive ventilation at 48 hours was defined as Pocc = 0 at 24 and 48 hours; patients with Pocc < 0 at 24 or 48 hours were deemed to not be breathing passively. A k-folds cross-validated best subsets supervised machine learning approach was used to identify clinical covariates best predicting passive ventilation at 48 hours. These covariates were used to create a clinical score to predict passive ventilation at 48 hours; the performance of the clinical score was assessed with a leave-one-out cross validation (LOOCV) area under the curve (AUC) approach. All analyses were performed in Rstudio Statistical Software. Results: There were 1186 patients in the database. 539 patients had Pocc value < 0 at 24 hours or 48 hours, while 549 patients had Pocc value = 0 at 24 and 48 hours. The k-folds cross-validated best subsets supervised machine learning model identified baseline values of prone positioning status at baseline, positive end-expiratory pressure (PEEP) setting, airway occlusion pressure (P0.1), and fraction of inspiratory oxygen (FiO2) as the covariates that predicted passive ventilation status at 48 hours after initiation of IMV. A clinical prediction score including these four covariates had an AUC of 0.836 for predicting the outcome of passive ventilation at 48 hours after initiation of IMV. Conclusion: In patients with AHRF undergoing IMV, a machine learning model identified baseline prone positioning, PEEP, P0.1, and FiO2 as best able to predict passive ventilation at 48 hours after initiation of IMV. A clinical prediction score using these 4 variables had AUC of 0.832 for predicting the outcome of interest. This prediction score may help identify which patients might benefit from diaphragmatic neurostimulation.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.005 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".