Predicting Right Ventricular Failure After Left Ventricular Assist Device Implant: A Novel Approach
Bibliographic record
Abstract
AIMS: Right ventricular (RV) failure (RVF) after left ventricular assist device (LVAD) implant is an important cause of morbidity and mortality. Modern, data-driven approaches for defining and predicting RVF have been under-utilized. METHODS: Two hundred thirty-two patients were identified with a mean age of 55 years; 40 (17%) were women, 132 were (59%) Caucasian and 74 (32%) were Black. Patients were split between Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) Classes 1, 2 and 3 (25%, 38% and 34%, respectively). Within this group, 'provisional RVF' patients were identified, along with 'no RVF' patients. 'No RVF' patients were defined as patients who never demonstrated more than moderate RV dysfunction on a post-LVAD transthoracic echocardiogram (TTE) (ordinal RV function <3), never required an RV assist device (RVAD), were not discharged on sildenafil and were not on a pulmonary vasodilator or inotropic medication at 3 months after LVAD implant. In total, n = 67 patients were defined as 'no RVF'. The remaining patients represented the 'provisional RVF' population (n = 165). Extensive electronic health records queries yielded >1200 data points per patient. Using <1 and >1 month post-LVAD time windows motivated by established, expert-consensus definitions of 'early' and 'late' post-implant RVF, unbiased clustering analysis was performed to identify hidden patient 'phenogroups' within these two established RVF populations. Clusters were compared on post-implant clinical metrics and 1 year outcomes. Lastly, pre-implant metrics were used to generate models for predicting post-implant RVF phenogroup. RESULTS: Within the 'early RVF' time window, distinct 'well' and 'sick' patient phenogroup clusters were identified. These clusters had similar RV function and pulmonary vasodilator usage during the first month after LVAD but differed significantly in heart failure therapy tolerance, renal (P < 0.001) and hepatic (P = 0.013) function, RVAD usage (P = 0.001) and 1 year mortality (P = 0.047). Distinct 'well' and 'sick' phenogroups were also identified in the 'late RVF' time window. These clusters had similar RV function (P = 0.111) and RVAD proportions (P = 0.757) but differed significantly in heart failure medication tolerance, pulmonary vasodilator usage (P = 0.001) and 1 year mortality (P < 0.001). Prediction of phenogroup clusters from the 'early RVF' population achieved an area under the receiver operating characteristic curve (AUROC) of 0.84, with top predictors including renal function, liver function, heart rate and pre-LVAD RV function. CONCLUSIONS: Distinct, potentially predictable phenogroups of patients who have significantly different long-term outcomes exist within consensus-defined post-LVAD RVF populations.
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 | 0.003 | 0.008 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.004 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
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".