Frequency and Clinical Implications of Referrals to Heart Failure Among Patients with Peripartum Cardiomyopathy.

Peripartum cardiomyopathy (PPCM) is a rare but significant cause of new-onset heart failure (HF) during the peri- and post-partum periods. Advances in GDMT for HF with reduced ventricular function have led to substantial improvements in survival and quality of life, yet few studies examine the longitudinal care received by patients with PPCM. The aim of this research is to address this gap by retrospectively characterizing patients with PPCM across a multihospital health system and investigating the frequency of cardiology and HF specialty referrals. Understanding whether surveillance and medical management differ among patients referred to HF will help to underscore the importance of referring patients with PPCM to HF specialists for optimal care.

The HFSA Advanced Heart Failure and Transplant Cardiology Fellowship Consensus Conference.

There is waning interest among cardiology trainees in pursuing an Advanced Heart Failure/Transplant Cardiology (AHFTC) fellowship as evidenced by fewer applicants in the National Resident Matching Program match to this specialty. This trend has generated considerable attention across the heart failure community. In response, the Heart Failure Society of America convened the AHFTC Fellowship Task Force with a charge to develop strategies to increase the value proposition of an AHFTC fellowship. Subsequently, the HFSA sponsored the AHFTC Fellowship Consensus Conference April 26-27, 2023. Before the conference, interviews of 44 expert stakeholders diverse across geography, site of practice (traditional academic medical center or other centers), specialty/area of expertise, sex, and stage of career were conducted virtually. Based on these interviews, potential solutions to address the declining interest in AHFTC fellowship were categorized into five themes: (1) alternative training pathways, (2) regulatory and compensation, (3) educational improvements, (4) exposure and marketing for pipeline development, and (5) quality of life and mental health. These themes provided structure to the deliberations of the AHFTC Fellowship Consensus Conference. The recommendations from the Consensus Conference were subsequently presented to the HFSA Board of Directors to inform strategic plans and interventions. The HFSA Board of Directors later reviewed and approved submission of this document. The purpose of this communication is to provide the HF community with an update summarizing the processes used and concepts that emerged from the work of the HFSA AHFTC Fellowship Task Force and Consensus Conference.

Association of Reduced Hospitalizations and Mortality Rates Among COVID-19-Vaccinated Patients With Heart Failure.

BACKGROUND: Patients with heart failure (HF) are at high risk for adverse outcomes when they have COVID-19. Reports of COVID-19 vaccine-related cardiac complications may contribute to vaccine hesitancy in patients with HF. METHODS: To analyze the impact of COVID-19 vaccine status on clinical outcomes in patients with HF, we conducted a retrospective cohort study of the association of COVID-19 vaccination status with hospitalizations, intensive care unit admission and mortality after adjustment for covariates. Inverse probability treatment-weighted models were used to adjust for potential confounding. RESULTS: Of 7094 patients with HF, 645 (9.1%) were partially vaccinated, 2200 (31.0%) were fully vaccinated, 1053 were vaccine-boosted (14.8%), and 3196 remained unvaccinated (45.1%) by January 2022. The mean age was 73.3 ± 14.5 years, and 48% were female. Lower mortality rates were observed in patients who were vaccine-boosted, followed by those who were fully vaccinated; they experienced lower mortality rates (HR 0.33; CI 0.23, 0.48) and 0.36 (CI 0.30, 0.43), respectively, compared to unvaccinated individuals (P< 0.001) over the mean follow-up time of 276.5 ± 104.9 days, whereas no difference was observed between those who were unvaccinated or only partially vaccinated. CONCLUSION: COVID-19 vaccination was associated with significant reduction in all-cause hospitalization rates and mortality rates, lending further evidence to support the importance of vaccination implementation in the high-risk population of patients living with HF.

Vaccines, Antibodies and Donors: Varying Attitudes and Policies Surrounding COVID-19 and Heart Transplantation.

INTRODUCTION: There are varied opinions in the United States regarding many aspects of care related to COVID-19. The purpose of this study was to examine the opinions of health care personnel and the policies of heart transplant centers concerning practices for the prevention and treatment of COVID-19 in donors and recipients of heart transplants. METHODS: Two anonymous, electronic web-based surveys were developed: 1 was administered to health care personnel through a mailing list maintained by the Heart Failure Society of America (HFSA); another was administered to U.S. medical adult and pediatric heart transplant (HT) program directors. Individual and group e-mails were sent with an embedded link to the respective surveys in February 2022. RESULTS: A total of 176 individuals (8.6%) responded to the survey administered through the HFSA. Of medical directors of transplant programs, 78 (54% response rate) completed a separate survey on their centers' policies. Although 95% (n = 167) of individuals indicated vaccination against COVID-19 should be required prior to HT, only 67% (n = 52) of centers mandated that practice. Similarly, 61% of individuals thought vaccination should be required prior to HT for caregivers, but only 13% of transplant centers mandated caregiver vaccination. Of the centers, 63% reported considering donors despite histories of recent COVID-19 infection (within 3 months), and 47% considered donors with current positive polymerase chain reaction tests. Regarding post-transplant care, only 22% of programs routinely measured antibodies to COVID-19, and 71% used tixagevimab/cilgavimab (Evusheld) for pre-exposure prophylaxis. CONCLUSIONS: There were significant differences between individual preferences and centers' practices with respect to COVID-19 management of candidates for and recipients of HT. Additionally, there was wide variation in policies among centers, reflecting the need for further study to inform consistent guidance and recommendations across centers to optimize equitable care for this high-risk patient population.

Sex Differences in Heart Failure.

Heart failure (HF) continues to be a major contributor of morbidity and mortality for men and women alike, yet how the predisposition for, course and management of HF differ between men and women remains underexplored. Sex differences in traditional risk factors as well as sex-specific risk factors influence the prevalence and manifestation of HF in unique ways. The pathophysiology of HF differs between men and women and may explain sex-specific differences in clinical presentation and diagnosis. This in turn, contributes to variation in response to both pharmacologic and device/surgical therapy. This review examines sex-specific differences in HF spanning prevalence, risk factors, pathophysiology, presentation, and therapies with a specific focus on highlighting gaps in knowledge with calls to action for future research efforts.

Frailty Measures of Patient-reported Activity and Fatigue May Predict 1-year Outcomes in Ambulatory Advanced Heart Failure: A Report From the REVIVAL Registry.

BACKGROUND: The Fried Frailty Phenotype predicts adverse outcomes in geriatric populations, but has not been well-studied in advanced heart failure (HF). The Registry Evaluation of Vital Information for Ventricular Assist Devices (VADs) in Ambulatory Life (REVIVAL) study prospectively collected frailty measures in patients with advanced HF to determine relevant assessments and their impact on clinical outcomes. METHODS AND RESULTS: HF-Fried Frailty was defined by 5 baseline components (1 point each): (1) weakness: hand grip strength less than 25% of body weight; (2) slowness based on time to walk 15 feet; (3) weight loss of more than 10 lbs in the past year; (4) inactivity; and (5) exhaustion, both assessed by the Kansas City Cardiomyopathy Questionnaire. A score of 0 or 1 was deemed nonfrail, 2 prefrail, and 3 or greater was considered frail. The primary composite outcome was durable mechanical circulatory support implantation, cardiac transplant or death at 1 year. Event-free survival for each group was determined by the Kaplan-Meier method and the hazard of prefrailty and frailty were compared with nonfrailty with proportional hazards modeling. Among 345 patients with all 5 frailty domains assessed, frailty was present in 17%, prefrailty in 40%, and 43% were nonfrail, with 67% (n = 232) meeting the criteria based on inactivity and 54% (n = 186) for exhaustion. Frail patients had an increased risk of the primary composite outcome (unadjusted hazard ratio [HR] 2.82, 95% confidence interval [CI] 1.52-5.24; adjusted HR 3.41, 95% CI 1.79-6.52), as did prefrail patients (unadjusted HR 1.97, 95% CI 1.14-3.41; adjusted HR 2.11, 95% CI 1.21-3.66) compared with nonfrail patients, however, the predictive value of HF-Fried Frailty criteria was modest (Harrel's C-statistic of 0.603, P = .004). CONCLUSIONS: The HF-Fried Frailty criteria had only modest predictive power in identifying ambulatory patients with advanced HF at high risk for durable mechanical circulatory support, transplant, or death within 1 year, driven primarily by assessments of inactivity and exhaustion. Focus on these patient-reported measures may better inform clinical trajectories in this population.

AKI in Hospitalized Patients with COVID-19.

BACKGROUND: Early reports indicate that AKI is common among patients with coronavirus disease 2019 (COVID-19) and associated with worse outcomes. However, AKI among hospitalized patients with COVID-19 in the United States is not well described. METHODS: This retrospective, observational study involved a review of data from electronic health records of patients aged ≥18 years with laboratory-confirmed COVID-19 admitted to the Mount Sinai Health System from February 27 to May 30, 2020. We describe the frequency of AKI and dialysis requirement, AKI recovery, and adjusted odds ratios (aORs) with mortality. RESULTS: Of 3993 hospitalized patients with COVID-19, AKI occurred in 1835 (46%) patients; 347 (19%) of the patients with AKI required dialysis. The proportions with stages 1, 2, or 3 AKI were 39%, 19%, and 42%, respectively. A total of 976 (24%) patients were admitted to intensive care, and 745 (76%) experienced AKI. Of the 435 patients with AKI and urine studies, 84% had proteinuria, 81% had hematuria, and 60% had leukocyturia. Independent predictors of severe AKI were CKD, men, and higher serum potassium at admission. In-hospital mortality was 50% among patients with AKI versus 8% among those without AKI (aOR, 9.2; 95% confidence interval, 7.5 to 11.3). Of survivors with AKI who were discharged, 35% had not recovered to baseline kidney function by the time of discharge. An additional 28 of 77 (36%) patients who had not recovered kidney function at discharge did so on posthospital follow-up. CONCLUSIONS: AKI is common among patients hospitalized with COVID-19 and is associated with high mortality. Of all patients with AKI, only 30% survived with recovery of kidney function by the time of discharge.

Coronavirus Disease 2019 in Solid Organ Transplant: A Multicenter Cohort Study.

BACKGROUND: The coronavirus disease 2019 (COVID-19) pandemic has led to significant reductions in transplantation, motivated in part by concerns of disproportionately more severe disease among solid organ transplant (SOT) recipients. However, clinical features, outcomes, and predictors of mortality in SOT recipients are not well described. METHODS: We performed a multicenter cohort study of SOT recipients with laboratory-confirmed COVID-19. Data were collected using standardized intake and 28-day follow-up electronic case report forms. Multivariable logistic regression was used to identify risk factors for the primary endpoint, 28-day mortality, among hospitalized patients. RESULTS: Four hundred eighty-two SOT recipients from >50 transplant centers were included: 318 (66%) kidney or kidney/pancreas, 73 (15.1%) liver, 57 (11.8%) heart, and 30 (6.2%) lung. Median age was 58 (interquartile range [IQR] 46-57), median time post-transplant was 5 years (IQR 2-10), 61% were male, and 92% had ≥1 underlying comorbidity. Among those hospitalized (376 [78%]), 117 (31%) required mechanical ventilation, and 77 (20.5%) died by 28 days after diagnosis. Specific underlying comorbidities (age >65 [adjusted odds ratio [aOR] 3.0, 95% confidence interval [CI] 1.7-5.5, P < .001], congestive heart failure [aOR 3.2, 95% CI 1.4-7.0, P = .004], chronic lung disease [aOR 2.5, 95% CI 1.2-5.2, P = .018], obesity [aOR 1.9, 95% CI 1.0-3.4, P = .039]) and presenting findings (lymphopenia [aOR 1.9, 95% CI 1.1-3.5, P = .033], abnormal chest imaging [aOR 2.9, 95% CI 1.1-7.5, P = .027]) were independently associated with mortality. Multiple measures of immunosuppression intensity were not associated with mortality. CONCLUSIONS: Mortality among SOT recipients hospitalized for COVID-19 was 20.5%. Age and underlying comorbidities rather than immunosuppression intensity-related measures were major drivers of mortality.

Practices of Referring Patients to Advanced Heart Failure Centers.

BACKGROUND: Therapies for advanced heart failure (AHF) improve the likelihood of survival in a growing population of patients with stage D heart failure (HF). Successful implementation of these therapies is dependent upon timely and appropriate referrals to AHF centers. METHODS: We performed a retrospective analysis of patients referred to 9 AHF centers for evaluation for AHF therapies. Patients' demographics, referring providers' characteristics, referral circumstances, and evaluation outcomes were collected. RESULTS: The majority of referrals (n = 515) were male (73.4%), and a majority of those were in the advanced state of the disease: very low left ventricular ejection fraction (<20% in 51.5%); 59.4% inpatient; and high risk Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) profiles (74.5% profile 1-3). HF cardiologists (49.1%) were the most common originating referral source; the least common (4.9%) were electrophysiologists. Common clinical triggers for referral included worsening HF (30.0%), inotrope dependence (19.6%), hospitalization (19.4%), and cardiogenic shock (17.8%). Most commonly, AHF therapies were not offered because patients were too sick (38.0%-45.1%) or for psychosocial reasons (20.3%-28.6%). Compared to non-HF cardiologists, patients referred by HF cardiologists were offered an AHF therapy more often (66.8% vs 58.4%, P = 0.0489). Of those not offered any AHF therapy, 28.4% received home inotropic therapy, and 14.5% were referred to hospice. CONCLUSIONS: In this multicenter review of AHF referrals, HF cardiologists referred the most patients despite being a relatively small proportion of the overall clinician population. Late referral was prevalent in this high-risk patient population and correlates with worsened outcomes, suggesting a significant need for broad clinician education regarding the benefits, triggers and appropriate timing of referral to AHF centers for optimal patient outcomes.

Right Ventricular Dysfunction Is Common and Identifies Patients at Risk of Dying in Cardiogenic Shock.

BACKGROUND: Understanding the prognostic impact of right ventricular dysfunction (RVD) in cardiogenic shock (CS) is a key step toward rational diagnostic and treatment algorithms and improved outcomes. Using a large multicenter registry, we assessed (1) the association between hemodynamic markers of RVD and in-hospital mortality, (2) the predictive value of invasive hemodynamic assessment incorporating RV evaluation, and (3) the impact of RVD severity on survival in CS. METHODS AND RESULTS: Inpatients with CS owing to acute myocardial infarction (AMI) or heart failure (HF) between 2016 and 2019 were included. RV parameters (right atrial pressure, right atrial/pulmonary capillary wedge pressure [RA/PCWP], pulmonary artery pulsatility index [PAPI], and right ventricular stroke work index [RVSWI]) were assessed between survivors and nonsurvivors, and between etiology and SCAI stage subcohorts. Multivariable logistic regression analysis determined hemodynamic predictors of in-hospital mortality; the resulting models were compared with SCAI staging alone. Nonsurvivors had a significantly higher right atrial pressure and RA/PCWP and lower PAPI and RVSWI than survivors, consistent with more severe RVD. Compared with AMI, patients with HF had a significantly lower RA/PCWP (0.58 vs 0.66, P = .001) and a higher PAPI (2.71 vs 1.78, P < .001) and RVSWI (5.70 g-m/m2 vs 4.66 g-m/m2, P < .001), reflecting relatively preserved RV function. Paradoxically, multiple RVD parameters (PAPI, RVSWI) were associated with mortality in the HF but not the AMI cohort. RVD was more severe with advanced SCAI stage, although its prognostic value was progressively diluted in stages D and E. Multivariable modelling incorporating the RA/PCWP improved the predictive value of SCAI staging (area under the curve [AUC] 0.78 vs 0.73, P < .001), largely driven by patients with HF (AUC 0.82 vs 0.71, P < .001). CONCLUSIONS: RVD is associated with poor outcomes in CS, with key differences across etiology and shock severity. Further studies are needed to assess the usefulness of RVD assessment in guiding therapy.

PredischaRge initiation of Ivabradine in the ManagEment of Heart Failure: Results of the PRIME-HF Trial.

BACKGROUND: Ivabradine is guideline-recommended to reduce heart failure (HF) hospitalization in patients with stable chronic HF with reduced ejection fraction (EF). Ivabradine initiation following acute HF has had limited evaluation, and there are few randomized data in US patients. The PredischaRge initiation of Ivabradine in the ManagEment of Heart Failure (PRIME-HF) study was conducted to address predischarge ivabradine initiation in stabilized acute HF patients. METHODS: PRIME-HF was an investigator-initiated, randomized, open-label study of predischarge initiation of ivabradine versus usual care. Eligible patients were hospitalized for acute HF but stabilized, with EF ≤35%, on maximally tolerated ß-blocker and in sinus rhythm with heart rate ≥70 beats/min. Ivabradine was acquired per routine care. The primary end point was the proportion of patients on ivabradine at 180 days. Additional end points included heart rate change, patient-reported outcomes, ß-blocker use/dose, and safety events (symptomatic bradycardia and hypotension). RESULTS: Overall, 104 patients (36% women, 64% African American) were randomized, and the study was terminated early because of funding limitations. At 180 days, 21 of 52 (40.4%) of patients randomized to predischarge initiation were treated with ivabradine compared with 6 of 52 (11.5%) randomized to usual care (odds ratio 5.19, 95% CI 1.88-14.33, P = .002). The predischarge initiation group experienced greater reduction in heart rate through 180 days (mean -10.0 beats/min, 95% CI -15.7 to -4.3 vs 0.7 beats/min, 95% CI -5.4 to 6.7, P = .011). Patient-reported outcomes, ß-blocker use/dose, and safety events were similar (all P > .05). CONCLUSIONS: Ivabradine initiation prior to discharge among stabilized HF patients increased ivabradine use at 180 days and lowered heart rates without reducing ß-blockers or increasing adverse events. As the trial did not achieve the planned enrollment, additional studies are needed.

PROVIDE-HF primary results: Patient-Reported Outcomes inVestigation following Initiation of Drug therapy with Entresto (sacubitril/valsartan) in heart failure.

BACKGROUND: In PARADIGM-HF, sacubitril/valsartan improved quality of life (QOL) versus enalapril in heart failure with reduced ejection fraction (HFrEF), yet limited data are available regarding QOL changes after sacubitril/valsartan initiation in routine practice. METHODS: PROVIDE-HF was a prospective study within a national research network (Patient-Centered Outcomes Research Network) of HFrEF outpatients recently initiated on sacubitril/valsartan versus controls with recent angiotensin-converting enzyme inhibitor/angiotensin receptor blocker initiation/dose change. The primary end point was mean Kansas City Cardiomyopathy Questionnaire (KCCQ) change through 12 weeks. Other end points included responder analyses: ≥5-point and ≥20-point KCCQ increase. Adjusted QOL change was estimated after propensity score weighting. RESULTS: Overall, 270 patients had both baseline and 12-week KCCQ data (151 sacubitril/valsartan; 119 control). The groups had similar demographics and HF details: median EF 28% and N-terminal pro-brain natriuretic peptide 1083 pg/mL. Sacubitril/valsartan patients had larger improvements in KCCQ (mean difference +4.76; P = .027) and were more likely to have a ≥5-point and ≥20-point response (all P < .05). Adjusted comparisons demonstrated similar numerical improvements in the change in KCCQ (+4.55; 95% CI -0.89 to 9.99; P = .101) and likelihood of ≥5-point increase (odds ratio 1.55; 95% CI: 0.84-2.86; P = .16); ≥20-point increase remained statistically significant (odds ratio 3.79; 95% CI 1.47-9.73; P = .006). CONCLUSIONS: In this prospective HFrEF study of sacubitril/valsartan initiation compared with recent angiotensin-converting enzyme inhibitor/angiotensin receptor blocker initiation/dose change, the between-group difference in the primary end point, mean KCCQ change at 12 weeks was not statistically significant following adjustment, but sacubitril/valsartan initiation was associated with early improvements in QOL and a higher likelihood of ≥20-point improvement in KCCQ at 12 weeks. These data add additional real-world evidence related to patient-reported outcomes following the initiation of sacubitril/valsartan in routine clinical practice.

Impact of Socioeconomic Factors on Patient Desire for Early LVAD Therapy Prior to Inotrope Dependence.

BACKGROUND: Worsening heart failure (HF) and health-related quality of life (HRQOL) have been shown to impact the decision to proceed with left ventricular assist device (LVAD) implantation, but little is known about how socioeconomic factors influence expressed patient preference for LVAD. METHODS AND RESULTS: Ambulatory patients with advanced systolic HF (n=353) reviewed written information about LVAD therapy and completed a brief survey to indicate whether they would want an LVAD to treat their current level of HF. Ordinal logistic regression analyses identified clinical and demographic predictors of LVAD preference. Higher New York Heart Association (NYHA) class, worse HRQOL measured by Kansas City Cardiomyopathy Questionnaire, lower education level, and lower income were significant univariable predictors of patients wanting an LVAD. In the multivariable model, higher NYHA class (OR [odds ratio]: 1.43, CI [confidence interval]: 1.08-1.90, P = .013) and lower income level (OR: 2.10, CI: 1.18 - 3.76, P = .012 for <$40,000 vs >$80,000) remained significantly associated with wanting an LVAD. CONCLUSION: Among ambulatory patients with advanced systolic HF, treatment preference for LVAD was influenced by level of income independent of HF severity. Understanding the impact of socioeconomic factors on willingness to consider LVAD therapy may help tailor counseling towards individual needs.

A manifesto of collaborative longitudinal cardiovascular care in heart failure.

In this document, we outline the challenges faced by patients and clinicians in heart failure, specifically centered around the needed coordination of care among the various subspecialties within cardiovascular medicine. We call for a more organized and collaborative effort among clinicians in primary care, general cardiology, electrophysiology, interventional cardiology, cardiothoracic surgery, cardiac imaging, and heart failure-all caring for mutual patients. Care is contextualized within the framework of two phases: a cardiomyopathy phase and an advanced heart failure phase, each of which lends to different considerations in therapy. Ultimately multidisciplinary coordinated care within cardiovascular medicine may lead to greater patient and clinician satisfaction as well as improved outcomes, but this remains to be investigated.

Characterization of Patients Who Return to Hospital Following Discharge from Hospitalization for COVID-19.

BACKGROUND: Data on patients with coronavirus disease 2019 (COVID-19) who return to hospital after discharge are scarce. Characterization of these patients may inform post-hospitalization care. OBJECTIVE: To describe clinical characteristics of patients with COVID-19 who returned to the emergency department (ED) or required readmission within 14 days of discharge. DESIGN: Retrospective cohort study of SARS-COV-2-positive patients with index hospitalization between February 27 and April 12, 2020, with ≥ 14-day follow-up. Significance was defined as P < 0.05 after multiplying P by 125 study-wide comparisons. PARTICIPANTS: Hospitalized patients with confirmed SARS-CoV-2 discharged alive from five New York City hospitals. MAIN MEASURES: Readmission or return to ED following discharge. RESULTS: Of 2864 discharged patients, 103 (3.6%) returned for emergency care after a median of 4.5 days, with 56 requiring inpatient readmission. The most common reason for return was respiratory distress (50%). Compared with patients who did not return, there were higher proportions of COPD (6.8% vs 2.9%) and hypertension (36% vs 22.1%) among those who returned. Patients who returned also had a shorter median length of stay (LOS) during index hospitalization (4.5 [2.9,9.1] vs 6.7 [3.5, 11.5] days; Padjusted = 0.006), and were less likely to have required intensive care on index hospitalization (5.8% vs 19%; Padjusted = 0.001). A trend towards association between absence of in-hospital treatment-dose anticoagulation on index admission and return to hospital was also observed (20.9% vs 30.9%, Padjusted = 0.06). On readmission, rates of intensive care and death were 5.8% and 3.6%, respectively. CONCLUSIONS: Return to hospital after admission for COVID-19 was infrequent within 14 days of discharge. The most common cause for return was respiratory distress. Patients who returned more likely had COPD and hypertension, shorter LOS on index-hospitalization, and lower rates of in-hospital treatment-dose anticoagulation. Future studies should focus on whether these comorbid conditions, longer LOS, and anticoagulation are associated with reduced readmissions.

Machine Learning to Predict Mortality and Critical Events in a Cohort of Patients With COVID-19 in New York City: Model Development and Validation.

BACKGROUND: COVID-19 has infected millions of people worldwide and is responsible for several hundred thousand fatalities. The COVID-19 pandemic has necessitated thoughtful resource allocation and early identification of high-risk patients. However, effective methods to meet these needs are lacking. OBJECTIVE: The aims of this study were to analyze the electronic health records (EHRs) of patients who tested positive for COVID-19 and were admitted to hospitals in the Mount Sinai Health System in New York City; to develop machine learning models for making predictions about the hospital course of the patients over clinically meaningful time horizons based on patient characteristics at admission; and to assess the performance of these models at multiple hospitals and time points. METHODS: We used Extreme Gradient Boosting (XGBoost) and baseline comparator models to predict in-hospital mortality and critical events at time windows of 3, 5, 7, and 10 days from admission. Our study population included harmonized EHR data from five hospitals in New York City for 4098 COVID-19-positive patients admitted from March 15 to May 22, 2020. The models were first trained on patients from a single hospital (n=1514) before or on May 1, externally validated on patients from four other hospitals (n=2201) before or on May 1, and prospectively validated on all patients after May 1 (n=383). Finally, we established model interpretability to identify and rank variables that drive model predictions. RESULTS: Upon cross-validation, the XGBoost classifier outperformed baseline models, with an area under the receiver operating characteristic curve (AUC-ROC) for mortality of 0.89 at 3 days, 0.85 at 5 and 7 days, and 0.84 at 10 days. XGBoost also performed well for critical event prediction, with an AUC-ROC of 0.80 at 3 days, 0.79 at 5 days, 0.80 at 7 days, and 0.81 at 10 days. In external validation, XGBoost achieved an AUC-ROC of 0.88 at 3 days, 0.86 at 5 days, 0.86 at 7 days, and 0.84 at 10 days for mortality prediction. Similarly, the unimputed XGBoost model achieved an AUC-ROC of 0.78 at 3 days, 0.79 at 5 days, 0.80 at 7 days, and 0.81 at 10 days. Trends in performance on prospective validation sets were similar. At 7 days, acute kidney injury on admission, elevated LDH, tachypnea, and hyperglycemia were the strongest drivers of critical event prediction, while higher age, anion gap, and C-reactive protein were the strongest drivers of mortality prediction. CONCLUSIONS: We externally and prospectively trained and validated machine learning models for mortality and critical events for patients with COVID-19 at different time horizons. These models identified at-risk patients and uncovered underlying relationships that predicted outcomes.