Machine-learning informed subtypes of heart failure and their palliative care trajectories: a population-based cohort study
Bibliographic record
Abstract
Abstract Introduction Heart failure (HF) significantly contributes to global morbidity and mortality, yet the provision of palliative care (PC) remains suboptimal in this population. Despite its anticipated benefits in symptom management and quality of life, the patterns of PC engagement across diverse HF subgroups are not well-defined. This study aims to identify HF subtypes and evaluate their PC trajectories to inform targeted strategies for improving PC access and delivery. Methods Using linked electronic health records from the Clinical Practice Research Datalink (CPRD) in England, we identified individuals aged ≥18 years with incident HF who died during follow-up. A previously developed machine learning-informed analytical framework was applied to define distinct HF subgroups based on demographic characteristics, biomarkers, hospitalizations, and comorbidities. Kaplan-Meier (KM) estimates were used to assess survival and PC trajectories across these subgroups. Results Among 51,332 individuals with incident HF, 15,376 (30%) received PC. Five distinct HF subgroups were identified: Cancer-Associated (9.3%), Cardiometabolic (12.7%), Metabolic (42.9%), AF-Associated (5.9%), and Late-Onset (29.2%). The median time to PC initiation (in those receiving it) following the diagnosis of HF was 4.3 years (IQR: 1.8–8.5). The Cancer-Associated subtype received PC earliest (median 2.6 years post-HF), while the Late-Onset subtype had the latest initiation (2.9 years) and the shortest median PC duration before death (0.35 years). Approximately 30% of patients had limited PC access. KM analysis demonstrated significant differences in PC receipt across these subtypes (p < 0.0001). Interpretation: Despite the high symptom burden and end-of-life needs in HF, receipt of PC remains limited, particularly among ‘late-onset’ HF patients. These findings underscore the inequity of PC provision even for high-risk groups in whom the early integration of wholistic care might benefit quality of life.Heart failure/palliative care subtypes Percentage of subtypes of HF/PC
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 | 0.002 | 0.004 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 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".