ABSTRACT
Aims: Deep learning methods have recently gained success in detecting left ventricular systolic dysfunction (LVSD) from electrocardiogram (ECG) waveforms. Despite their high level of accuracy, they are difficult to interpret and deploy broadly in the clinical setting. In this study, we set out to determine whether simpler models based on standard ECG measurements could detect LVSD with similar accuracy to that of deep learning models. Methods and results: Using an observational data set of 40 994 matched 12-lead ECGs and transthoracic echocardiograms, we trained a range of models with increasing complexity to detect LVSD based on ECG waveforms and derived measurements. The training data were acquired from the Stanford University Medical Center. External validation data were acquired from the Columbia Medical Center and the UK Biobank. The Stanford data set consisted of 40 994 matched ECGs and echocardiograms, of which 9.72% had LVSD. A random forest model using 555 discrete, automated measurements achieved an area under the receiver operator characteristic curve (AUC) of 0.92 (0.91-0.93), similar to a deep learning waveform model with an AUC of 0.94 (0.93-0.94). A logistic regression model based on five measurements achieved high performance [AUC of 0.86 (0.85-0.87)], close to a deep learning model and better than N-terminal prohormone brain natriuretic peptide (NT-proBNP). Finally, we found that simpler models were more portable across sites, with experiments at two independent, external sites. Conclusion: Our study demonstrates the value of simple electrocardiographic models that perform nearly as well as deep learning models, while being much easier to implement and interpret.
ABSTRACT
BACKGROUND: Artificial intelligence, particularly deep learning (DL), has immense potential to improve the interpretation of transthoracic echocardiography (TTE). Mitral regurgitation (MR) is the most common valvular heart disease and presents unique challenges for DL, including the integration of multiple video-level assessments into a final study-level classification. METHODS: A novel DL system was developed to intake complete TTEs, identify color MR Doppler videos, and determine MR severity on a 4-step ordinal scale (none/trace, mild, moderate, and severe) using the reading cardiologist as a reference standard. This DL system was tested in internal and external test sets with performance assessed by agreement with the reading cardiologist, weighted κ, and area under the receiver-operating characteristic curve for binary classification of both moderate or greater and severe MR. In addition to the primary 4-step model, a 6-step MR assessment model was studied with the addition of the intermediate MR classes of mild-moderate and moderate-severe with performance assessed by both exact agreement and ±1 step agreement with the clinical MR interpretation. RESULTS: A total of 61 689 TTEs were split into train (n=43 811), validation (n=8891), and internal test (n=8987) sets with an additional external test set of 8208 TTEs. The model had high performance in MR classification in internal (exact accuracy, 82%; κ=0.84; area under the receiver-operating characteristic curve, 0.98 for moderate/severe MR) and external test sets (exact accuracy, 79%; κ=0.80; area under the receiver-operating characteristic curve, 0.98 for moderate or greater MR). Most (63% internal and 66% external) misclassification disagreements were between none/trace and mild MR. MR classification accuracy was slightly higher using multiple TTE views (accuracy, 82%) than with only apical 4-chamber views (accuracy, 80%). In subset analyses, the model was accurate in the classification of both primary and secondary MR with slightly lower performance in cases of eccentric MR. In the analysis of the 6-step classification system, the exact accuracy was 80% and 76% with a ±1 step agreement of 99% and 98% in the internal and external test set, respectively. CONCLUSIONS: This end-to-end DL system can intake entire echocardiogram studies to accurately classify MR severity and may be useful in helping clinicians refine MR assessments.
ABSTRACT
Recent artificial intelligence (AI) advancements in cardiovascular care offer potential enhancements in effective diagnosis, treatment, and outcomes. More than 600 U.S. Food and Drug Administration-approved clinical AI algorithms now exist, with 10% focusing on cardiovascular applications, highlighting the growing opportunities for AI to augment care. This review discusses the latest advancements in the field of AI, with a particular focus on the utilization of multimodal inputs and the field of generative AI. Further discussions in this review involve an approach to understanding the larger context in which AI-augmented care may exist, and include a discussion of the need for rigorous evaluation, appropriate infrastructure for deployment, ethics and equity assessments, regulatory oversight, and viable business cases for deployment. Embracing this rapidly evolving technology while setting an appropriately high evaluation benchmark with careful and patient-centered implementation will be crucial for cardiology to leverage AI to enhance patient care and the provider experience.
Subject(s)
Artificial Intelligence , Cardiovascular Diseases , Humans , Cardiovascular Diseases/therapy , Cardiovascular Diseases/diagnosis , CardiologyABSTRACT
Recent artificial intelligence (AI) advancements in cardiovascular care offer potential enhancements in diagnosis, treatment, and outcomes. Innovations to date focus on automating measurements, enhancing image quality, and detecting diseases using novel methods. Applications span wearables, electrocardiograms, echocardiography, angiography, genetics, and more. AI models detect diseases from electrocardiograms at accuracy not previously achieved by technology or human experts, including reduced ejection fraction, valvular heart disease, and other cardiomyopathies. However, AI's unique characteristics necessitate rigorous validation by addressing training methods, real-world efficacy, equity concerns, and long-term reliability. Despite an exponentially growing number of studies in cardiovascular AI, trials showing improvement in outcomes remain lacking. A number are currently underway. Embracing this rapidly evolving technology while setting a high evaluation benchmark will be crucial for cardiology to leverage AI to enhance patient care and the provider experience.
Subject(s)
Artificial Intelligence , Cardiovascular Diseases , Humans , Cardiovascular Diseases/therapy , Cardiovascular Diseases/diagnosis , Cardiology/methodsABSTRACT
BACKGROUND AND AIMS: Early identification of cardiac structural abnormalities indicative of heart failure is crucial to improving patient outcomes. Chest X-rays (CXRs) are routinely conducted on a broad population of patients, presenting an opportunity to build scalable screening tools for structural abnormalities indicative of Stage B or worse heart failure with deep learning methods. In this study, a model was developed to identify severe left ventricular hypertrophy (SLVH) and dilated left ventricle (DLV) using CXRs. METHODS: A total of 71 589 unique CXRs from 24 689 different patients completed within 1 year of echocardiograms were identified. Labels for SLVH, DLV, and a composite label indicating the presence of either were extracted from echocardiograms. A deep learning model was developed and evaluated using area under the receiver operating characteristic curve (AUROC). Performance was additionally validated on 8003 CXRs from an external site and compared against visual assessment by 15 board-certified radiologists. RESULTS: The model yielded an AUROC of 0.79 (0.76-0.81) for SLVH, 0.80 (0.77-0.84) for DLV, and 0.80 (0.78-0.83) for the composite label, with similar performance on an external data set. The model outperformed all 15 individual radiologists for predicting the composite label and achieved a sensitivity of 71% vs. 66% against the consensus vote across all radiologists at a fixed specificity of 73%. CONCLUSIONS: Deep learning analysis of CXRs can accurately detect the presence of certain structural abnormalities and may be useful in early identification of patients with LV hypertrophy and dilation. As a resource to promote further innovation, 71 589 CXRs with adjoining echocardiographic labels have been made publicly available.
Subject(s)
Deep Learning , Hypertrophy, Left Ventricular , Radiography, Thoracic , Humans , Hypertrophy, Left Ventricular/diagnostic imaging , Radiography, Thoracic/methods , Female , Male , Middle Aged , Echocardiography/methods , Aged , Heart Failure/diagnostic imaging , Heart Ventricles/diagnostic imaging , ROC CurveABSTRACT
AIMS: Variation in diagnostic performance of single-photon emission computed tomography (SPECT) myocardial perfusion imaging (MPI) has been observed, yet the impact of cardiac size has not been well characterized. We assessed whether low left ventricular volume influences SPECT MPI's ability to detect obstructive coronary artery disease (CAD) and its interaction with age and sex. METHODS AND RESULTS: A total of 2066 patients without known CAD (67% male, 64.7 ± 11.2 years) across nine institutions underwent SPECT MPI with solid-state scanners followed by coronary angiography as part of the REgistry of Fast Myocardial Perfusion Imaging with NExt Generation SPECT. Area under receiver-operating characteristic curve (AUC) analyses evaluated the performance of quantitative and visual assessments according to cardiac size [end-diastolic volume (EDV); <20th vs. ≥20th population or sex-specific percentiles], age (<75 vs. ≥75 years), and sex. Significantly decreased performance was observed in patients with low EDV compared with those without (AUC: population 0.72 vs. 0.78, P = 0.03; sex-specific 0.72 vs. 0.79, P = 0.01) and elderly patients compared with younger patients (AUC 0.72 vs. 0.78, P = 0.03), whereas males and females demonstrated similar AUC (0.77 vs. 0.76, P = 0.67). The reduction in accuracy attributed to lower volumes was primarily observed in males (sex-specific threshold: EDV 0.69 vs. 0.79, P = 0.01). Accordingly, a significant decrease in AUC, sensitivity, specificity, and negative predictive value for quantitative and visual assessments was noted in patients with at least two characteristics of low EDV, elderly age, or male sex. CONCLUSION: Detection of CAD with SPECT MPI is negatively impacted by small cardiac size, most notably in elderly and male patients.
Subject(s)
Coronary Artery Disease , Myocardial Perfusion Imaging , Registries , Tomography, Emission-Computed, Single-Photon , Humans , Male , Female , Middle Aged , Myocardial Perfusion Imaging/methods , Aged , Tomography, Emission-Computed, Single-Photon/methods , Coronary Artery Disease/diagnostic imaging , Organ Size , Sex Factors , Coronary Angiography/methods , ROC Curve , Age Factors , Sensitivity and SpecificityABSTRACT
BACKGROUND: Preoperative risk assessments used in clinical practice are insufficient in their ability to identify risk for postoperative mortality. Deep-learning analysis of electrocardiography can identify hidden risk markers that can help to prognosticate postoperative mortality. We aimed to develop a prognostic model that accurately predicts postoperative mortality in patients undergoing medical procedures and who had received preoperative electrocardiographic diagnostic testing. METHODS: In a derivation cohort of preoperative patients with available electrocardiograms (ECGs) from Cedars-Sinai Medical Center (Los Angeles, CA, USA) between Jan 1, 2015 and Dec 31, 2019, a deep-learning algorithm was developed to leverage waveform signals to discriminate postoperative mortality. We randomly split patients (8:1:1) into subsets for training, internal validation, and final algorithm test analyses. Model performance was assessed using area under the receiver operating characteristic curve (AUC) values in the hold-out test dataset and in two external hospital cohorts and compared with the established Revised Cardiac Risk Index (RCRI) score. The primary outcome was post-procedural mortality across three health-care systems. FINDINGS: 45 969 patients had a complete ECG waveform image available for at least one 12-lead ECG performed within the 30 days before the procedure date (59 975 inpatient procedures and 112 794 ECGs): 36 839 patients in the training dataset, 4549 in the internal validation dataset, and 4581 in the internal test dataset. In the held-out internal test cohort, the algorithm discriminates mortality with an AUC value of 0·83 (95% CI 0·79-0·87), surpassing the discrimination of the RCRI score with an AUC of 0·67 (0·61-0·72). The algorithm similarly discriminated risk for mortality in two independent US health-care systems, with AUCs of 0·79 (0·75-0·83) and 0·75 (0·74-0·76), respectively. Patients determined to be high risk by the deep-learning model had an unadjusted odds ratio (OR) of 8·83 (5·57-13·20) for postoperative mortality compared with an unadjusted OR of 2·08 (0·77-3·50) for postoperative mortality for RCRI scores of more than 2. The deep-learning algorithm performed similarly for patients undergoing cardiac surgery (AUC 0·85 [0·77-0·92]), non-cardiac surgery (AUC 0·83 [0·79-0·88]), and catheterisation or endoscopy suite procedures (AUC 0·76 [0·72-0·81]). INTERPRETATION: A deep-learning algorithm interpreting preoperative ECGs can improve discrimination of postoperative mortality. The deep-learning algorithm worked equally well for risk stratification of cardiac surgeries, non-cardiac surgeries, and catheterisation laboratory procedures, and was validated in three independent health-care systems. This algorithm can provide additional information to clinicians making the decision to perform medical procedures and stratify the risk of future complications. FUNDING: National Heart, Lung, and Blood Institute.
Subject(s)
Deep Learning , Humans , Risk Assessment/methods , Algorithms , Prognosis , ElectrocardiographyABSTRACT
Artificial intelligence, computational simulations, and extended reality, among other 21st century computational technologies, are changing the health care system. To collectively highlight the most recent advances and benefits of artificial intelligence, computational simulations, and extended reality in cardiovascular therapies, we coined the abbreviation AISER. The review particularly focuses on the following applications of AISER: 1) preprocedural planning and clinical decision making; 2) virtual clinical trials, and cardiovascular device research, development, and regulatory approval; and 3) education and training of interventional health care professionals and medical technology innovators. We also discuss the obstacles and constraints associated with the application of AISER technologies, as well as the proposed solutions. Interventional health care professionals, computer scientists, biomedical engineers, experts in bioinformatics and visualization, the device industry, ethics committees, and regulatory agencies are expected to streamline the use of AISER technologies in cardiovascular interventions and medicine in general.
Subject(s)
Artificial Intelligence , Humans , Treatment OutcomeABSTRACT
The electrocardiogram (ECG) is the most frequently performed cardiovascular diagnostic test, but it is unclear how much information resting ECGs contain about long term cardiovascular risk. Here we report that a deep convolutional neural network can accurately predict the long-term risk of cardiovascular mortality and disease based on a resting ECG alone. Using a large dataset of resting 12-lead ECGs collected at Stanford University Medical Center, we developed SEER, the Stanford Estimator of Electrocardiogram Risk. SEER predicts 5-year cardiovascular mortality with an area under the receiver operator characteristic curve (AUC) of 0.83 in a held-out test set at Stanford, and with AUCs of 0.78 and 0.83 respectively when independently evaluated at Cedars-Sinai Medical Center and Columbia University Irving Medical Center. SEER predicts 5-year atherosclerotic disease (ASCVD) with an AUC of 0.67, similar to the Pooled Cohort Equations for ASCVD Risk, while being only modestly correlated. When used in conjunction with the Pooled Cohort Equations, SEER accurately reclassified 16% of patients from low to moderate risk, uncovering a group with an actual average 9.9% 10-year ASCVD risk who would not have otherwise been indicated for statin therapy. SEER can also predict several other cardiovascular conditions such as heart failure and atrial fibrillation. Using only lead I of the ECG it predicts 5-year cardiovascular mortality with an AUC of 0.80. SEER, used alongside the Pooled Cohort Equations and other risk tools, can substantially improve cardiovascular risk stratification and aid in medical decision making.
ABSTRACT
Recent advances in large language models (LLMs) have demonstrated remarkable successes in zero- and few-shot performance on various downstream tasks, paving the way for applications in high-stakes domains. In this study, we systematically examine the capabilities and limitations of LLMs, specifically GPT-3.5 and ChatGPT, in performing zero-shot medical evidence summarization across six clinical domains. We conduct both automatic and human evaluations, covering several dimensions of summary quality. Our study demonstrates that automatic metrics often do not strongly correlate with the quality of summaries. Furthermore, informed by our human evaluations, we define a terminology of error types for medical evidence summarization. Our findings reveal that LLMs could be susceptible to generating factually inconsistent summaries and making overly convincing or uncertain statements, leading to potential harm due to misinformation. Moreover, we find that models struggle to identify the salient information and are more error-prone when summarizing over longer textual contexts.
ABSTRACT
We evaluated the association between left cardiac 3-dimensional echocardiographic parameters and brain injury in a single-center prospective study of neonates with neonatal encephalopathy. On day 2 of life, neonates with brain injury had greater left ventricle end-diastolic and stroke volume but also greater peak global circumferential strain detected by 3-dimensional echocardiogram.
Subject(s)
Brain Injuries , Echocardiography, Three-Dimensional , Ventricular Dysfunction, Left , Infant, Newborn , Humans , Infant , Ventricular Function, Left , Prospective Studies , Echocardiography/methods , Echocardiography, Three-Dimensional/methods , Stroke Volume , Heart Ventricles/diagnostic imaging , Brain Injuries/complications , Brain Injuries/diagnostic imaging , Ventricular Dysfunction, Left/diagnostic imagingABSTRACT
Recent advances in large language models (LLMs) have demonstrated remarkable successes in zero- and few-shot performance on various downstream tasks, paving the way for applications in high-stakes domains. In this study, we systematically examine the capabilities and limitations of LLMs, specifically GPT-3.5 and ChatGPT, in performing zero-shot medical evidence summarization across six clinical domains. We conduct both automatic and human evaluations, covering several dimensions of summary quality. Our study has demonstrated that automatic metrics often do not strongly correlate with the quality of summaries. Furthermore, informed by our human evaluations, we define a terminology of error types for medical evidence summarization. Our findings reveal that LLMs could be susceptible to generating factually inconsistent summaries and making overly convincing or uncertain statements, leading to potential harm due to misinformation. Moreover, we find that models struggle to identify the salient information and are more error-prone when summarizing over longer textual contexts.
ABSTRACT
BACKGROUND: Valvular heart disease is an important contributor to cardiovascular morbidity and mortality and remains underdiagnosed. Deep learning analysis of electrocardiography (ECG) may be useful in detecting aortic stenosis (AS), aortic regurgitation (AR), and mitral regurgitation (MR). OBJECTIVES: This study aimed to develop ECG deep learning algorithms to identify moderate or severe AS, AR, and MR alone and in combination. METHODS: A total of 77,163 patients undergoing ECG within 1 year before echocardiography from 2005-2021 were identified and split into train (n = 43,165), validation (n = 12,950), and test sets (n = 21,048; 7.8% with any of AS, AR, or MR). Model performance was assessed using area under the receiver-operating characteristic (AU-ROC) and precision-recall curves. Outside validation was conducted on an independent data set. Test accuracy was modeled using different disease prevalence levels to simulate screening efficacy using the deep learning model. RESULTS: The deep learning algorithm model accuracy was as follows: AS (AU-ROC: 0.88), AR (AU-ROC: 0.77), MR (AU-ROC: 0.83), and any of AS, AR, or MR (AU-ROC: 0.84; sensitivity 78%, specificity 73%) with similar accuracy in external validation. In screening program modeling, test characteristics were dependent on underlying prevalence and selected sensitivity levels. At a prevalence of 7.8%, the positive and negative predictive values were 20% and 97.6%, respectively. CONCLUSIONS: Deep learning analysis of the ECG can accurately detect AS, AR, and MR in this multicenter cohort and may serve as the basis for the development of a valvular heart disease screening program.
Subject(s)
Aortic Valve Insufficiency , Aortic Valve Stenosis , Deep Learning , Heart Valve Diseases , Mitral Valve Insufficiency , Aortic Valve Insufficiency/diagnosis , Aortic Valve Stenosis/diagnosis , Electrocardiography , Heart Valve Diseases/diagnosis , Heart Valve Diseases/epidemiology , Humans , Mitral Valve Insufficiency/diagnosis , Mitral Valve Insufficiency/epidemiologyABSTRACT
Heart failure with preserved ejection fraction is a heterogeneous clinical syndrome that includes distinct subtypes with different pathophysiologies, genetics, and treatment. Distinguishing heart failure with preserved ejection fraction caused by transthyretin cardiac amyloidosis (ATTR-CA) is critical given its specific treatment. We analyzed a single-center retrospective cohort to determine the association of body mass index (BMI) with a composite of either ATTR-CA or the valine-to-isoleucine substitution (Val122Ile) variant genotype (ATTR-CA+Val122Ile). These BMI differences were prospectively evaluated in the multicenter Screening for Cardiac Amyloidosis using nuclear imaging for Minority Populations (SCAN-MP) study of Black and Hispanic patients with heart failure. The association of BMI with ATTR-CA+Val122Ile was compared by Wilcoxon rank sum analysis and combined with age, gender, and maximum left ventricle wall thickness in multivariable logistic regression. In the retrospective analysis (n = 469), ATTR-CA+Val122Ile was identified in n = 198 (40%), who had a lower median BMI (25.8 kg/m2, interquartile range [IQR] 23.4 to 28.9) than other patients (27.1 kg/m2, IQR 23.9 to 32.0) (p <0.001). In multivariable logistic regression, BMI <30 kg/m2 (odds ratio 2.6, 95% confidence interval 1.5 to 4.5) remained independently associated with ATTR-CA+Val122Ile with a greater association in Black and Hispanic patients (odds ratio 5.8, 95% confidence interval 1.7 to 19.6). In SCAN-MP (n = 201), 17 (8%) had either ATTR-CA (n = 10) or were Val122Ile carriers (n = 7) with negative pyrophosphate scans. BMI was lower (25.4 kg/m2 [IQR 24.3 to 28.2]) in ATTR-CA+Val122Ile patients than in non-amyloid patients (32.7 kg/m2 [28.3 to 38.6]) (p <0.001), a finding that persisted in multivariable analysis (p = 0.002). In conclusion, lower BMI is associated with ATTR-CA+Val122Ile in heart failure with increased left ventricle wall thickness, particularly in Black and Hispanic patients, and may aid in the identification of those benefiting from ATTR-CA evaluation.
Subject(s)
Amyloid Neuropathies, Familial , Cardiomyopathies , Heart Diseases , Heart Failure , Amyloid Neuropathies, Familial/diagnostic imaging , Amyloid Neuropathies, Familial/genetics , Body Mass Index , Hispanic or Latino , Humans , Prealbumin/genetics , Retrospective StudiesABSTRACT
BACKGROUND: Prospective studies demonstrate that aggressive pharmacological therapy combined with pump speed optimization may result in myocardial recovery in larger numbers of patients supported with left ventricular assist device (LVAD). This study sought to determine whether the use of machine learning (ML) based models predict LVAD patients with myocardial recovery resulting in pump explant. METHODS: A total of 20 270 adult patients with a durable continuous-flow LVAD in the INTERMACS registry (Interagency Registry for Mechanically Assisted Circulatory Support) were included in the study. Ninety-eight raw clinical variables were screened using the least absolute shrinkage and selection operator for selection of features associated with LVAD-induced myocardial recovery. ML models were developed in the training data set (70%) and were assessed in the validation data set (30%) by receiver operating curve and Kaplan-Meier analysis. RESULTS: Least absolute shrinkage and selection operator identified 28 unique clinical features associated with LVAD-induced myocardial recovery, including age, cause of heart failure, psychosocial risk factors, laboratory values, cardiac rate and rhythm, and echocardiographic indices. ML models achieved area under the receiver operating curve of 0.813 to 0.824 in the validation data set outperforming logistic regression-based new INTERMACS recovery risk score (area under the receiver operating curve of 0.796) and previously established LVAD recovery risk scores (INTERMACS Cardiac Recovery Score and INTERMACS Recovery Score by Topkara et al) with area under the receiver operating curve of 0.744 and 0.748 (P<0.05). Patients who were predicted to recover by ML models demonstrated a significantly higher incidence of myocardial recovery resulting in LVAD explant in the validation cohort compared with those who were not predicted to recover (18.8% versus 2.6% at 4 years of pump support). CONCLUSIONS: ML can be a valuable tool to identify subsets of LVAD patients who may be more likely to respond to myocardial recovery protocols.
Subject(s)
Heart Failure/epidemiology , Heart Ventricles/physiopathology , Heart-Assist Devices , Machine Learning , Adolescent , Adult , Cohort Studies , Heart Failure/physiopathology , Heart-Assist Devices/adverse effects , Humans , Incidence , Middle Aged , Myocardium/pathology , Prospective Studies , Registries/statistics & numerical data , Young AdultABSTRACT
Importance: Critical illness, a marked inflammatory response, and viruses such as SARS-CoV-2 may prolong corrected QT interval (QTc). Objective: To evaluate baseline QTc interval on 12-lead electrocardiograms (ECGs) and ensuing changes among patients with and without COVID-19. Design, Setting, and Participants: This cohort study included 3050 patients aged 18 years and older who underwent SARS-CoV-2 testing and had ECGs at Columbia University Irving Medical Center from March 1 through May 1, 2020. Patients were analyzed by treatment group over 5 days, as follows: hydroxychloroquine with azithromycin, hydroxychloroquine alone, azithromycin alone, and neither hydroxychloroquine nor azithromycin. ECGs were manually analyzed by electrophysiologists masked to COVID-19 status. Multivariable modeling evaluated clinical associations with QTc prolongation from baseline. Exposures: COVID-19, hydroxychloroquine, azithromycin. Main Outcomes and Measures: Mean QTc prolongation, percentage of patients with QTc of 500 milliseconds or greater. Results: A total of 965 patients had more than 2 ECGs and were included in the study, with 561 (58.1%) men, 198 (26.2%) Black patients, and 191 (19.8%) aged 80 years and older. There were 733 patients (76.0%) with COVID-19 and 232 patients (24.0%) without COVID-19. COVID-19 infection was associated with significant mean QTc prolongation from baseline by both 5-day and 2-day multivariable models (5-day, patients with COVID-19: 20.81 [95% CI, 15.29 to 26.33] milliseconds; P < .001; patients without COVID-19: -2.01 [95% CI, -17.31 to 21.32] milliseconds; P = .93; 2-day, patients with COVID-19: 17.40 [95% CI, 12.65 to 22.16] milliseconds; P < .001; patients without COVID-19: 0.11 [95% CI, -12.60 to 12.81] milliseconds; P = .99). COVID-19 infection was independently associated with a modeled mean 27.32 (95% CI, 4.63-43.21) millisecond increase in QTc at 5 days compared with COVID-19-negative status (mean QTc, with COVID-19: 450.45 [95% CI, 441.6 to 459.3] milliseconds; without COVID-19: 423.13 [95% CI, 403.25 to 443.01] milliseconds; P = .01). More patients with COVID-19 not receiving hydroxychloroquine and azithromycin had QTc of 500 milliseconds or greater compared with patients without COVID-19 (34 of 136 [25.0%] vs 17 of 158 [10.8%], P = .002). Multivariable analysis revealed that age 80 years and older compared with those younger than 50 years (mean difference in QTc, 11.91 [SE, 4.69; 95% CI, 2.73 to 21.09]; P = .01), severe chronic kidney disease compared with no chronic kidney disease (mean difference in QTc, 12.20 [SE, 5.26; 95% CI, 1.89 to 22.51; P = .02]), elevated high-sensitivity troponin levels (mean difference in QTc, 5.05 [SE, 1.19; 95% CI, 2.72 to 7.38]; P < .001), and elevated lactate dehydrogenase levels (mean difference in QTc, 5.31 [SE, 2.68; 95% CI, 0.06 to 10.57]; P = .04) were associated with QTc prolongation. Torsades de pointes occurred in 1 patient (0.1%) with COVID-19. Conclusions and Relevance: In this cohort study, COVID-19 infection was independently associated with significant mean QTc prolongation at days 5 and 2 of hospitalization compared with day 0. More patients with COVID-19 had QTc of 500 milliseconds or greater compared with patients without COVID-19.
Subject(s)
Azithromycin , COVID-19 Drug Treatment , COVID-19 , Electrocardiography , Hydroxychloroquine , Long QT Syndrome , Aged, 80 and over , Anti-Infective Agents/administration & dosage , Anti-Infective Agents/adverse effects , Azithromycin/administration & dosage , Azithromycin/adverse effects , COVID-19/diagnosis , COVID-19/epidemiology , COVID-19 Testing/methods , Drug Therapy, Combination/methods , Drug Therapy, Combination/statistics & numerical data , Electrocardiography/methods , Electrocardiography/statistics & numerical data , Female , Hospitalization/statistics & numerical data , Humans , Hydroxychloroquine/administration & dosage , Hydroxychloroquine/adverse effects , Long QT Syndrome/chemically induced , Long QT Syndrome/diagnosis , Long QT Syndrome/epidemiology , Long QT Syndrome/virology , Male , Middle Aged , New York/epidemiology , Outcome and Process Assessment, Health Care , Risk Factors , SARS-CoV-2 , Time FactorsABSTRACT
OBJECTIVE: Coronavirus disease 2019 (COVID-19) patients are at risk for resource-intensive outcomes including mechanical ventilation (MV), renal replacement therapy (RRT), and readmission. Accurate outcome prognostication could facilitate hospital resource allocation. We develop and validate predictive models for each outcome using retrospective electronic health record data for COVID-19 patients treated between March 2 and May 6, 2020. MATERIALS AND METHODS: For each outcome, we trained 3 classes of prediction models using clinical data for a cohort of SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2)-positive patients (n = 2256). Cross-validation was used to select the best-performing models per the areas under the receiver-operating characteristic and precision-recall curves. Models were validated using a held-out cohort (n = 855). We measured each model's calibration and evaluated feature importances to interpret model output. RESULTS: The predictive performance for our selected models on the held-out cohort was as follows: area under the receiver-operating characteristic curve-MV 0.743 (95% CI, 0.682-0.812), RRT 0.847 (95% CI, 0.772-0.936), readmission 0.871 (95% CI, 0.830-0.917); area under the precision-recall curve-MV 0.137 (95% CI, 0.047-0.175), RRT 0.325 (95% CI, 0.117-0.497), readmission 0.504 (95% CI, 0.388-0.604). Predictions were well calibrated, and the most important features within each model were consistent with clinical intuition. DISCUSSION: Our models produce performant, well-calibrated, and interpretable predictions for COVID-19 patients at risk for the target outcomes. They demonstrate the potential to accurately estimate outcome prognosis in resource-constrained care sites managing COVID-19 patients. CONCLUSIONS: We develop and validate prognostic models targeting MV, RRT, and readmission for hospitalized COVID-19 patients which produce accurate, interpretable predictions. Additional external validation studies are needed to further verify the generalizability of our results.