Development and validation of a risk model for the prediction of cardiovascular hospital admission using CMR-based phenotype in patients with known or suspected cardiovascular disease
Bibliographic record
Abstract
Abstract Background Cardiovascular diseases remain the leading cause of morbidity worldwide and impose the highest economic burden among noncommunicable diseases. Much of these costs are related to hospitalizations for adverse cardiovascular events, which may be reduced by targeted management of high-risk patients. Cardiac markers derived from CMR imaging have been shown to be strong independent predictors of prognosis within specific cohorts. However, its capacity to broadly contribute to risk models aimed at predicting incident cardiac hospitalization has not been demonstrated. Purpose Using a large clinical outcomes registry of patients clinically referred for CMR, develop and validate a nomogram for prediction of cardiovascular hospital admission. Methods A total of 7127 consecutive patients were prospectively recruited between 02/2015 and 07/2019. All patients completed standardized health questionnaires and CMR imaging protocols. A nomogram was developed for prediction of cardiovascular hospitalization, inclusive of admission for heart failure, MI, cardiac arrest, heart transplant, LVAD implantation, or stroke. The risk model was derived from 80% (n=5702) of the cohort using Cox modelling that included CMR, medication, laboratory, and patient-reported health variables. Model validation was assessed by discrimination and calibration procedures applied to the remaining 20% of patients (n=1425). A minimum follow-up of six months was mandated. Results The derivation cohort was comprised of 38% females with a median age of 56 (IQR 44–65) years. During a median follow-up of 934 days, 514 (9.0%) events occurred. The validation cohort was similarly comprised of 37% females with a median age of 57 (IQR 44–66) years. During a median follow-up of 970 days, 142 (10.0%) events occurred. Numerous CMR parameters were significantly different between those experiencing versus not experiencing the primary composite outcome, including: LVEF (44% vs 59%, p<0.0001), RVEF (52% vs 55%, p<0.0001), LV mass (65g/m2 vs 56g/m2, p<0.0001), and LA volume (43mL/m2 vs 34mL/m2, p<0.0001). These and other CMR-derived characteristics were independently predictive of the composite outcome by univariate modelling (Figure 1A). An eight-variable nomogram (Figure 1B) was developed using a stepwise multivariate model that exhibited high discrimination in both the derivation and validation cohorts (C-index 0.81 and 0.83, respectively). Continuous model calibration curves indicated satisfactory external performance. The model was able to discriminate risk of hospitalization at 1-year with a dynamic range of 20–99%. Conclusion Using data available at time of CMR imaging, we derived and validated a Cox-based nomogram that offers robust prediction of future cardiovascular admissions. This tool may provide value for the identification of patients who may benefit from targeted surveillance and management strategies, and may offer a foundation for improved patient-specific cost modelling. Figure 1 Funding Acknowledgement Type of funding source: None
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 | 0.008 | 0.011 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".