Impact of Pressure Recovery Adjustment on Aortic Valve Area Classification of Disease Severity in Transcatheter Aortic Valve Replacement Patients.

OBJECTIVES: To determine the impact of pressure recovery (PR) adjustment on disease severity grading in patients with severe aortic stenosis. The authors hypothesized that accounting for PR would result in echocardiographic reclassification of aortic stenosis severity in a significant number of patients. DESIGN: A retrospective observational study between October 2013 and February 2021. SETTING: A single-center, quaternary-care academic center. PARTICIPANTS: Adults (≥18 years old) who underwent transcatheter aortic valve implantation (TAVI). INTERVENTIONS: TAVI. MEASUREMENTS AND MAIN RESULTS: A total of 342 patients were evaluated in this study. Left ventricle mass index was significantly greater in patients who continued to be severe after PR (100.47 ± 28.77 v 90.15 ± 24.03, p = < 0.000001). Using PR-adjusted aortic valve area (AVA) resulted in the reclassification of 81 patients (24%) from severe to moderate aortic stenosis (AVA >1.0 cm2). Of the 81 patients who were reclassified, 23 patients (28%) had sinotubular junction (STJ) diameters >3.0 cm. CONCLUSION: Adjusting calculated AVA for PR resulted in a reclassification of a significant number of adult patients from severe to moderate aortic stenosis. PR was significantly larger in patients who reclassified from severe to moderate aortic stenosis after adjusting for PR. PR appeared to remain relevant in patients with STJ ≥3.0 cm. Clinicians need to be aware of PR and how to account for its effect when measuring pressure gradients with Doppler.

Automated Detection of Aortic Stenosis Using Machine Learning.

BACKGROUND: Aortic stenosis (AS) is a degenerative valve condition that is underdiagnosed and undertreated. Detection of AS using limited two-dimensional echocardiography could enable screening and improve appropriate referral and treatment of this condition. The aim of this study was to develop methods for automated detection of AS from limited imaging data sets. METHODS: Convolutional neural networks were trained, validated, and tested using limited two-dimensional transthoracic echocardiographic data sets. Networks were developed to accomplish two sequential tasks: (1) view identification and (2) study-level grade of AS. Balanced accuracy and area under the receiver operator curve (AUROC) were the performance metrics used. RESULTS: Annotated images from 577 patients were included. Neural networks were trained on data from 338 patients (average n = 10,253 labeled images), validated on 119 patients (average n = 3,505 labeled images), and performance was assessed on a test set of 120 patients (average n = 3,511 labeled images). Fully automated screening for AS was achieved with an AUROC of 0.96. Networks can distinguish no significant (no, mild, mild to moderate) AS from significant (moderate or severe) AS with an AUROC of 0.86 and between early (mild or mild to moderate AS) and significant (moderate or severe) AS with an AUROC of 0.75. External validation of these networks in a cohort of 8,502 outpatient transthoracic echocardiograms showed that screening for AS can be achieved using parasternal long-axis imaging only with an AUROC of 0.91. CONCLUSION: Fully automated detection of AS using limited two-dimensional data sets is achievable using modern neural networks. These methods lay the groundwork for a novel method for screening for AS.

Left ventricular stroke volume index following transcatheter aortic valve replacement is an early predictor of 1-year survival.

BACKGROUND: Adverse cardiac events are common following transcatheter aortic valve replacement (TAVR). Our aim was to investigate the low left ventricular stroke volume index (LVSVI) 30 days after TAVR as an early echocardiographic marker of survival. HYPOTHESIS: Steady-state (30-day) LVSVI after TAVR is associated with 1-year mortality. METHODS: A single-center retrospective analysis of all patients undergoing TAVR from 2017 to 2019. Baseline and 30-day post-TAVR echocardiographic LVSVI were calculated. Patients were stratified by pre-TAVR transaortic gradient, surgical risk, and change in transvalvular flow following TAVR. RESULTS: This analysis focuses on 238 patients treated with TAVR. The 1-year mortality rate was 9% and 124 (52%) patients had normal flow post-TAVR. Of those with pre-TAVR low flow, 67% of patients did not normalize LVSVI at 30 days. The 30-day normal flow was associated with lower 1-year mortality when compared to low flow (4% vs. 14%, p = .007). This association remained significant after adjusting for known predictors of risk (adjusted odds ratio [OR] of 3.45, 95% confidence interval: 1.02-11.63 [per 1 ml/m2 decrease], p = .046). Normalized transvalvular flow following TAVR was associated with reduced mortality (8%) when compared to those with persistent (15%) or new-onset low flow (12%) (p = .01). CONCLUSIONS: LVSVI at 30 days following TAVR is an early echocardiographic predictor of 1-year mortality and identifies patients with worse intermediate outcomes. More work is needed to understand if this short-term imaging marker might represent a novel therapeutic target.

Detecting Heart Disease from Multi-View Ultrasound Images via Supervised Attention Multiple Instance Learning.

Aortic stenosis (AS) is a degenerative valve condition that causes substantial morbidity and mortality. This condition is under-diagnosed and under-treated. In clinical practice, AS is diagnosed with expert review of transthoracic echocardiography, which produces dozens of ultrasound images of the heart. Only some of these views show the aortic valve. To automate screening for AS, deep networks must learn to mimic a human expert's ability to identify views of the aortic valve then aggregate across these relevant images to produce a study-level diagnosis. We find previous approaches to AS detection yield insufficient accuracy due to relying on inflexible averages across images. We further find that off-the-shelf attention-based multiple instance learning (MIL) performs poorly. We contribute a new end-to-end MIL approach with two key methodological innovations. First, a supervised attention technique guides the learned attention mechanism to favor relevant views. Second, a novel self-supervised pretraining strategy applies contrastive learning on the representation of the whole study instead of individual images as commonly done in prior literature. Experiments on an open-access dataset and a temporally-external heldout set show that our approach yields higher accuracy while reducing model size.

Generalizability of Cardiovascular Disease Clinical Prediction Models: 158 Independent External Validations of 104 Unique Models.

BACKGROUND: While clinical prediction models (CPMs) are used increasingly commonly to guide patient care, the performance and clinical utility of these CPMs in new patient cohorts is poorly understood. METHODS: We performed 158 external validations of 104 unique CPMs across 3 domains of cardiovascular disease (primary prevention, acute coronary syndrome, and heart failure). Validations were performed in publicly available clinical trial cohorts and model performance was assessed using measures of discrimination, calibration, and net benefit. To explore potential reasons for poor model performance, CPM-clinical trial cohort pairs were stratified based on relatedness, a domain-specific set of characteristics to qualitatively grade the similarity of derivation and validation patient populations. We also examined the model-based C-statistic to assess whether changes in discrimination were because of differences in case-mix between the derivation and validation samples. The impact of model updating on model performance was also assessed. RESULTS: Discrimination decreased significantly between model derivation (0.76 [interquartile range 0.73-0.78]) and validation (0.64 [interquartile range 0.60-0.67], P<0.001), but approximately half of this decrease was because of narrower case-mix in the validation samples. CPMs had better discrimination when tested in related compared with distantly related trial cohorts. Calibration slope was also significantly higher in related trial cohorts (0.77 [interquartile range, 0.59-0.90]) than distantly related cohorts (0.59 [interquartile range 0.43-0.73], P=0.001). When considering the full range of possible decision thresholds between half and twice the outcome incidence, 91% of models had a risk of harm (net benefit below default strategy) at some threshold; this risk could be reduced substantially via updating model intercept, calibration slope, or complete re-estimation. CONCLUSIONS: There are significant decreases in model performance when applying cardiovascular disease CPMs to new patient populations, resulting in substantial risk of harm. Model updating can mitigate these risks. Care should be taken when using CPMs to guide clinical decision-making.

External Validations of Cardiovascular Clinical Prediction Models: A Large-Scale Review of the Literature.

BACKGROUND: There are many clinical prediction models (CPMs) available to inform treatment decisions for patients with cardiovascular disease. However, the extent to which they have been externally tested, and how well they generally perform has not been broadly evaluated. METHODS: A SCOPUS citation search was run on March 22, 2017 to identify external validations of cardiovascular CPMs in the Tufts Predictive Analytics and Comparative Effectiveness CPM Registry. We assessed the extent of external validation, performance heterogeneity across databases, and explored factors associated with model performance, including a global assessment of the clinical relatedness between the derivation and validation data. RESULTS: We identified 2030 external validations of 1382 CPMs. Eight hundred seven (58%) of the CPMs in the Registry have never been externally validated. On average, there were 1.5 validations per CPM (range, 0-94). The median external validation area under the receiver operating characteristic curve was 0.73 (25th-75th percentile [interquartile range (IQR)], 0.66-0.79), representing a median percent decrease in discrimination of -11.1% (IQR, -32.4% to +2.7%) compared with performance on derivation data. 81% (n=1333) of validations reporting area under the receiver operating characteristic curve showed discrimination below that reported in the derivation dataset. 53% (n=983) of the validations report some measure of CPM calibration. For CPMs evaluated more than once, there was typically a large range of performance. Of 1702 validations classified by relatedness, the percent change in discrimination was -3.7% (IQR, -13.2 to 3.1) for closely related validations (n=123), -9.0 (IQR, -27.6 to 3.9) for related validations (n=862), and -17.2% (IQR, -42.3 to 0) for distantly related validations (n=717; P<0.001). CONCLUSIONS: Many published cardiovascular CPMs have never been externally validated, and for those that have, apparent performance during development is often overly optimistic. A single external validation appears insufficient to broadly understand the performance heterogeneity across different settings.

Large-scale validation of the prediction model risk of bias assessment Tool (PROBAST) using a short form: high risk of bias models show poorer discrimination.

OBJECTIVE: To assess whether the Prediction model Risk Of Bias ASsessment Tool (PROBAST) and a shorter version of this tool can identify clinical prediction models (CPMs) that perform poorly at external validation. STUDY DESIGN AND SETTING: We evaluated risk of bias (ROB) on 102 CPMs from the Tufts CPM Registry, comparing PROBAST to a short form consisting of six PROBAST items anticipated to best identify high ROB. We then applied the short form to all CPMs in the Registry with at least 1 validation (n=556) and assessed the change in discrimination (dAUC) in external validation cohorts (n=1,147). RESULTS: PROBAST classified 98/102 CPMS as high ROB. The short form identified 96 of these 98 as high ROB (98% sensitivity), with perfect specificity. In the full CPM registry, 527 of 556 CPMs (95%) were classified as high ROB, 20 (3.6%) low ROB, and 9 (1.6%) unclear ROB. Only one model with unclear ROB was reclassified to high ROB after full PROBAST assessment of all low and unclear ROB models. Median change in discrimination was significantly smaller in low ROB models (dAUC -0.9%, IQR -6.2-4.2%) compared to high ROB models (dAUC -11.7%, IQR -33.3-2.6%; P<0.001). CONCLUSION: High ROB is pervasive among published CPMs. It is associated with poor discriminative performance at validation, supporting the application of PROBAST or a shorter version in CPM reviews.

Comorbidity Burden and Adverse Outcomes After Transcatheter Aortic Valve Replacement.

Background Transcatheter aortic valve replacement (TAVR) has become the preferred treatment for symptomatic patients with aortic stenosis and elevated procedural risk. Many deaths following TAVR are because of noncardiac causes and comorbid disease burden may be a major determinant of postprocedure outcomes. The prevalence of comorbid conditions and associations with outcomes after TAVR has not been studied. Methods and Results This was a retrospective single-center study of patients treated with TAVR from January 2015 to October 2018. The association between 21 chronic conditions and short- and medium-term outcomes was assessed. A total of 341 patients underwent TAVR and had 1-year follow-up. The mean age was 81.4 (SD 8.0) years with a mean Society of Thoracic Surgeons predicted risk of mortality score of 6.7% (SD 4.8). Two hundred twenty (65%) patients had ≥4 chronic conditions present at the time of TAVR. There was modest correlation between Society of Thoracic Surgeons predicted risk of mortality and comorbid disease burden (r=0.32, P<0.001). After adjusting for Society of Thoracic Surgeons predicted risk of mortality, age, and vascular access, each additional comorbid condition was associated with increased rates of 30-day rehospitalizations (odds ratio, 1.21; 95% CI, 1.02-1.44), a composite of 30-day rehospitalization and 30-day mortality (odds ratio, 1.20; 95% CI, 1.02-1.42), and 1-year mortality (odds ratio, 1.29; 95% CI, 1.05-1.59). Conclusions Comorbid disease burden is associated with worse clinical outcomes in high-risk patients treated with TAVR. The risks associated with comorbid disease burden are not adequately captured by standard risk assessment. A systematic assessment of comorbid conditions may improve risk stratification efforts.

What is Quality End-of-Life Care for Patients With Heart Failure? A Qualitative Study With Physicians.

Background Advanced heart failure (AHF) carries a morbidity and mortality that are similar or worse than many advanced cancers. Despite this, there are no accepted quality metrics for end-of-life (EOL) care for patients with AHF. Methods and Results As a first step toward identifying quality measures, we performed a qualitative study with 23 physicians who care for patients with AHF. Individual, in-depth, semistructured interviews explored physicians' perceptions of characteristics of high-quality EOL care and the barriers encountered. Interviews were analyzed using software-assisted line-by-line coding in order to identify emergent themes. Although some elements and barriers of high-quality EOL care for AHF were similar to those described for other diseases, we identified several unique features. We found a competing desire to avoid overly aggressive care at EOL alongside a need to ensure that life-prolonging interventions were exhausted. We also identified several barriers related to identifying EOL including greater prognostic uncertainty, inadequate recognition of AHF as a terminal disease and dependence of symptom control on disease-modifying therapies. Conclusions Our findings support quality metrics that prioritize receipt of goal-concordant care over utilization measures as well as a need for more inclusive payment models that appropriately reflect the dual nature of many AHF therapies.

Can Clinical Predictive Models Identify Patients Who Should Not Receive TAVR? A Systematic Review.

BACKGROUND: One third of high- and prohibitive-risk TAVR patients remain severely symptomatic or die 1 year after treatment. There is interest in identifying individuals for whom this procedure is futile and should not be offered. METHODS: We performed a systematic review of the highest reported stratum of risk in TAVR clinical predictive models (CPMs). We explore whether currently available predictive models can identify patients for whom TAVR is futile, based on a quantitative futility definition and the observed and predicted outcomes for patients in the highest stratum of risk. RESULTS: 17 TAVR CPMs representing 69,191 treated patients were published from 2013 to 2018. When reported, the median number of patients in the highest stratum of risk was 569 (range 1 to 1759). Observed mortality for this risk stratum ranged from 9% at 30 days to 59% at 1 year after TAVR. Statistical confidence in these observed event rates was low. The highest predicted event rates ranged from 11.0% for in-hospital mortality to 75.1% for the composite of mortality or high symptom burden 1 year after TAVR. CONCLUSION: No high-risk TAVR group in currently available TAVR CPMs had an appropriate event rate and adequate statistical power to meet a quantitative definition of futility.

Clinical Predictive Models of Sudden Cardiac Arrest: A Survey of the Current Science and Analysis of Model Performances.

Background More than 500 000 sudden cardiac arrests (SCAs) occur annually in the United States. Clinical predictive models (CPMs) may be helpful tools to differentiate between patients who are likely to survive or have good neurologic recovery and those who are not. However, which CPMs are most reliable for discriminating between outcomes in SCA is not known. Methods and Results We performed a systematic review of the literature using the Tufts PACE (Predictive Analytics and Comparative Effectiveness) CPM Registry through February 1, 2020, and identified 81 unique CPMs of SCA and 62 subsequent external validation studies. Initial cardiac rhythm, age, and duration of cardiopulmonary resuscitation were the 3 most commonly used predictive variables. Only 33 of the 81 novel SCA CPMs (41%) were validated at least once. Of 81 novel SCA CPMs, 56 (69%) and 61 of 62 validation studies (98%) reported discrimination, with median c-statistics of 0.84 and 0.81, respectively. Calibration was reported in only 29 of 62 validation studies (41.9%). For those novel models that both reported discrimination and were validated (26 models), the median percentage change in discrimination was -1.6%. We identified 3 CPMs that had undergone at least 3 external validation studies: the out-of-hospital cardiac arrest score (9 validations; median c-statistic, 0.79), the cardiac arrest hospital prognosis score (6 validations; median c-statistic, 0.83), and the good outcome following attempted resuscitation score (6 validations; median c-statistic, 0.76). Conclusions Although only a small number of SCA CPMs have been rigorously validated, the ones that have been demonstrate good discrimination.

Clinical Prediction Models for Valvular Heart Disease.

Background While many clinical prediction models (CPMs) exist to guide valvular heart disease treatment decisions, the relative performance of these CPMs is largely unknown. We systematically describe the CPMs available for patients with valvular heart disease with specific attention to performance in external validations. Methods and Results A systematic review identified 49 CPMs for patients with valvular heart disease treated with surgery (n=34), percutaneous interventions (n=12), or no intervention (n=3). There were 204 external validations of these CPMs. Only 35 (71%) CPMs have been externally validated. Sixty-five percent (n=133) of the external validations were performed on distantly related populations. There was substantial heterogeneity in model performance and a median percentage change in discrimination of -27.1% (interquartile range, -49.4%--5.7%). Nearly two-thirds of validations (n=129) demonstrate at least a 10% relative decline in discrimination. Discriminatory performance of EuroSCORE II and Society of Thoracic Surgeons (2009) models (accounting for 73% of external validations) varied widely: EuroSCORE II validation c-statistic range 0.50 to 0.95; Society of Thoracic Surgeons (2009) Models validation c-statistic range 0.50 to 0.86. These models performed well when tested on related populations (median related validation c-statistics: EuroSCORE II, 0.82 [0.76, 0.85]; Society of Thoracic Surgeons [2009], 0.72 [0.67, 0.79]). There remain few (n=9) external validations of transcatheter aortic valve replacement CPMs. Conclusions Many CPMs for patients with valvular heart disease have never been externally validated and isolated external validations appear insufficient to assess the trustworthiness of predictions. For surgical valve interventions, there are existing predictive models that perform reasonably well on related populations. For transcatheter aortic valve replacement (CPMs additional external validations are needed to broadly understand the trustworthiness of predictions.