Featured Quality Research

August 2023:

Asishana Avo Osho, MD, MPH
Assistant Professor of Surgery, Division of Cardiac Surgery
Massachusetts General Hospital, Harvard Medical School

Data Registry: Get With The Guidelines^® - Coronary Artery Disease

Published Study: Race-Based Differences in ST-Segment–Elevation Myocardial Infarction Process Metrics and Mortality From 2015 Through 2021: An Analysis of 178 062 Patients From the American Heart Association Get With The Guidelines–Coronary Artery Disease Registry. Circulation.

When you started this study, what were you aiming to learn?
“We were hoping to shed light on any potential race/ethnicity-based inequities in STEMI quality metrics nationally.”

Who was your target patient population and how did you study them?
“The database made this easy and we are very grateful for the GWTG-CAD program and team. We hoped to include all patients who presented with a STEMI and were captured in the database. It was important to include patients who had complete race/ethnicity information and good records about how long it took to receive care within different phases.”

Did the study results turn out as you had expected or were you surprised with what you found?
“We suspected there might be some disparities in care based on race-ethnicity but didn't anticipate that it would apply across the board with all of the quality metrics. We also didn't expect that there would be resultant mortality differences in some populations.”

How do you see this study impacting the future of medicine and patient care?
“We hope that this study will stimulate modifications in practice and implementation of systems of care to address the identified inequities. Future program improvements will need to factor our findings as they are designed. Additionally, some hypothesis related to language will need to be studied.”

July 2023:

Jamie Diamond, MD, MPH
Heart Failure & Transplant Cardiology Fellow
Duke University Medical Center

Data Registry: Get With The Guidelines^® - Heart Failure

Published Study: Quality of Care and Clinical Outcomes for Patients With Heart Failure at Hospitals Caring for a High Proportion of Black Adults: Get With The Guidelines-Heart Failure Registry. JAMA Cardiology.

When you started this study, what were you aiming to learn?
“Racial disparities in heart failure outcomes have been well established and eliminating them remains a national priority. Although hospitals caring for high proportions of Black patients have been shown to have worse heart failure outcomes (including readmissions and mortality) than other hospitals, it was unclear whether this is because they deliver worse quality of care, or alternatively, because they care for more medically and socially complex populations. Understanding this is critically important, given ongoing concern that federal value-based and pay-for-performance programs are disproportionately penalizing certain hospitals for the patients they serve rather than for poor quality care. Determining this also represented an opportunity to evaluate potential targets of improvement at the hospital level. Therefore, we wanted to understand whether there were within- and between-hospital differences in heart failure quality of care at hospitals caring for high proportions of Black patients vs. other hospitals.”

Who was your target patient population and how did you study them?
“We used the Get With The Guidelines Heart Failure (GWTG-HF) registry to evaluate quality of care for Black adults in the US as compared to other adults with heart failure. The GWTG-HF registry was also linked with the US Centers for Medicare & Medicaid Services administrative claims data to evaluate outcomes in the subset of patients 65 years or older enrolled in Medicare fee-for-service.”

Did the study results turn out as you had expected or were you surprised with what you found?
“We were not surprised to learn that there were some differences between hospitals caring for high proportions of Black patients compared to other hospitals because of the known differences in community infrastructure, timely follow-up appointments and resource allocation between the two hospital groups. Interestingly, overall quality of care was largely similar between the two hospital types on 11 of 14 pre-defined quality measures and for the overall metric of heart failure defect-free care. This finding indicates that the observed differences in clinical outcomes are more likely explained by broader system inequities and that the impact between socioeconomic factors and associated outcomes needs to be further elucidated.”

How do you see this study impacting the future of medicine and patient care?
“The differences we did find in our study including aldosterone antagonism at discharge, cardiac resynchronization-defibrillator or cardiac resynchronization-defibrillator therapy/pacing placement or prescription at discharge and follow-up visit made within 7 days or less do represent more granular opportunities for improvement at hospitals caring for high proportions of Black patients. Given the broader implications of our findings, the hope is that this study will generate more research into identifying and alleviating the known higher burden of social risk factors clustered within communities that utilize hospitals with high proportion of Black patients. Both hospital and population-level factors need to be addressed to see a trend toward more equitable heart failure outcomes.”

April 2023:

Yikuan Li, MSc
PhD Candidate
Division of Health and Biomedical Informatics, Department of Preventive Medicine Feinberg School of Medicine, Northwestern University

Data Registry: Get With The Guidelines^® - Heart Failure

Published Study: Improving Fairness in the Prediction of Heart Failure Length of Stay and Mortality by Integrating Social Determinants of Health; Circulation: Heart Failure.

When you started this study, what were you aiming to learn?
“The algorithmic biases are receiving more attention recently, as more studies have found that machine learning (ML) algorithms can perpetuate human biases and perform unfairly across different subpopulations. We would like to quantify and investigate if there was any algorithmic bias when building an ML model to predict the outcomes for HF patients. We were also interested in finding some methods to improve the fairness of models without sacrificing the performance.”

Who was your target patient population and how did you study them?
“We obtained 210 368 HF patients from the American Heart Association’s Get With The Guidelines® - Heart Failure (GWTG-HF) registry. The GWTG-HF registry was linked to two datasets, social deprivation index and area deprivation index, by using the zip code of each individual. Thus, 15 social determinants of health variables were added to the feature space, along with the clinical features extracted from (GWTG-HF) registry, to build machine learning models. All the experiments were performed at the Precision Medicine Platform, which was generously supported by the American Heart Association (AHA).”

Did the study results turn out as you had expected or were you surprised with what you found?
“We were surprised by how 'prejudiced' the machine learning (ML) models were. The ML models selectively under-diagnosed the social-demographic disadvantaged groups, e.g. Black, female and uninsured people. The discrimination of ML models would bring double-jeopardies to those if no intervention of de-biasing were to be applied.”

How do you see this study impacting the future of medicine and patient care?
“Our study provided an example framework to evaluate, quantify and improve the algorithmic biases of clinical predictive models. We urge peer researchers to improve the fairness problem when building clinical predictive models, which will facilitate the real-world applications of those models.”

November 2022:

James G. Jollis MD
Cardiologist, The Christ Hospital, Cincinnati, OH
Member of the AHA Coronary Artery Disease Systems of Care Advisory Work Group

Data Registry: Get With The Guidelines^® - Coronary Artery Disease

Published Study: Treatment time and in-hospital mortality among patients with ST elevation myocardial infarction, 2018-2021; JAMA

When you started this study, what were you aiming to learn?
“As a cardiologist, most of the important decisions that affect whether my heart attack patients live or die occur long before they see me. I rely on family members, tele-communicators, first responders, paramedics and front-line emergency department colleagues to diagnose and implement care that has the most impact on patient care and outcomes. In addition to these front-line colleagues, the only national organization that has organized, trained and can lead such great care is the American Heart Association. These important data, informed by quality-focused participating hospitals and the AHA GWTG-CAD registry are designed to assure that patients get great care before I meet them.

We were hoping to understand the current state of heart attack care using data collected by some of the best hospitals in the country.””

Who was your target patient population and how did you study them?
“Patients having heart symptoms who were diagnosed to have an ST-elevation myocardial infarction.

601 hospitals participating in a focused quality improvement project submitted data as part of their process to improve care.”

Did the study results turn out as you had expected or were you surprised with what you found?
“As two years of the study were conducted during the pandemic, we expected to find some declines in care. However, less than 1% of patients were diagnosed with SARS-CoV-2 infection, such that COVID only had a small direct role. Rather, I believe the pandemic potentially had a significant indirect role as hospitals were overwhelmed with patients, EMS and hospitals were challenged in maintaining paramedic and nurse staffing and intensive care bed availability, and patients experienced delayed care due to barriers to access or perceived fear of become entangled in an overwhelmed medical system.

The finding that patients were twice as likely to survive if they were treated within Mission: Lifeline time process goals was not surprising.”

How do you see this study impacting the future of medicine and patient care?
“I expect we will see a redoubling of regional efforts to improve heart attack care on the part of collaborating health care professionals in EMS, emergency departments, and catheterization laboratories.

Many emergency medical systems are currently challenged by diminished resources in the face of greater demands. By giving paramedics our strongest support, more lives will be saved.”

September 2022:

Rishi Wadhera, MD, MPP, MPhil
Section Head of Health Policy and Equity, Richard A. and Susan F. Smith Center for Outcomes Research
Assistant Professor of Medicine, Harvard Medical School

Data Registry: COVID-19 CVD Registry

Published Study: County-Level Social Vulnerability is Associated With In-Hospital Death and Major Adverse Cardiovascular Events in Patients Hospitalized With COVID-19; Circ Cardiovasc Qual Outcomes.

When you started this study, what were you aiming to learn?
“The COVID-19 pandemic has disproportionately affected low-income and minoritized populations in the US. We wanted to understand whether hospitalized patients with COVID-19 who live in socially vulnerable communities experience higher rates of death and/or major adverse cardiovascular events.”

Who was your target patient population and how did you study them?
“Our study included patients with COVID-19 in the American Heart Association COVID-19 Cardiovascular Disease Registry across 107 US hospitals between January 14, 2020 to November 30, 2020.”

Did the study results turn out as you had expected or were you surprised with what you found?
“We found that patients hospitalized with COVID-19 from the most socially vulnerable counties were ≈25% more likely to experience an in-hospital death compared with individuals living in the least vulnerable counties. These associations were most pronounced during the first wave of the COVID-19 pandemic. In addition, socially vulnerable patients were also significantly more likely to experience major adverse cardiovascular events than their less vulnerable counterparts.”

August 2022:

Matthew Segar, MD, MS
Cardiology Fellow
Texas Heart Institute

Data Registry: Get With The Guidelines^® - Heart Failure

Published Study: Machine Learning-Based Models Incorporating Social Determinants of Health vs Traditional Models for Predicting In-Hospital Mortality in Patients With Heart Failure; JAMA Cardiol.

When you started this study, what were you aiming to learn?
“Prior to this study, race has been a common covariate included in many risk prediction models. Our team had previously shown that race-specific models outperformed models that included race as a covariate when assessing long-term risk of heart failure. What we didn’t know, however, was if race-specific or race-agnostic models would be superior when assessing short-term, in-hospital mortality risk. Additionally, we hypothesized that including ZIP-code level social determinants of health could improve model performance”

Who was your target patient population and how did you study them?
“Our study utilized data from the American Heart Association’s Get With The Guidelines® - Heart Failure (GWTG-HF) registry. The AHA also has a fantastic resource with the Precision Medicine Platform (PMP) that provides researchers with an environment to conduct their research. Now having used the PMP for multiple studies, it provides an unparalleled opportunity to perform very high-level and sophisticated analyses that would traditionally fail on personal computers.“

Did the study results turn out as you had expected or were you surprised with what you found?
"Based on our previous studies, we hypothesized that race-specific and race-agnostic models would outperform traditional methods that included race as a covariate. However, I was surprised by the social determinants of health (SDoH) data. As physicians, we learn the basic of how important SDoH is to adverse health outcomes. What was surprising, however, was exactly how much SDoH contributed to the risk of in-hospital mortality and how different the contribution was between race groups. The improvement in risk performance with the inclusion of SDoH measures was also an exciting finding."

How do you see this study impacting the future of medicine and patient care?
“I think the current method of including race as a covariate is outdated and does not take into consideration the underlying social and biologic risk factors that contribute to in-hospital mortality risk. Excluding race and including SDoH measures is a promising alternative that not only provides a more accurate risk assessment, but also provides actionable opportunities to modify a patient’s risk. Our stud provides additional support that improving a patient’s environment, whether that’s improving their access to transportation, housing, or food, can improve their cardiovascular health”

July 2022:

Amelia K. Adcock, M.D.
Associate Professor
Director, Center for Telestroke and Teleneurology
Associate Director, Comprehensive Stroke Center
West Virginia University School of Medicine
Department of Neurology Cerebrovascular Division

Data Registry: Get With The Guidelines^® - Stroke

Published Study: Trends in Use, Outcomes, and Disparities in Endovascular Thrombectomy in US Patients With Stroke Aged 80 Years and Older Compared With Younger Patients; JAMA Netw Open.

When you started this study, what were you aiming to learn?
"Is endovascular therapy (EVT) offered at similar rates regardless of age? Are there any differences between the patients and/or their clinical outcomes stratified by age group?"

Who was your target patient population and how did you study them?
"Our target population was all acute ischemic stroke patients who were potentially eligible for EVT and those who received EVT nationwide. As vascular imaging was not recorded during the study time period, we substituted a National Institute of Health Stroke Scale (NIHSS) score of at least 6 as a surrogate for higher probability of large vessel occlusion stroke. This was an observational study utilizing clinical and demographic data reported by hospitals participating in the Get With The Guidelines-Stroke Registry. We compared the rates of patients undergoing EVT who were 80 years or older with those patients younger than 80 years of age as well as the corresponding patient/hospital characteristics and clinical outcomes.”

Did the study results turn out as you had expected or were you surprised with what you found?
"Overall, our study found that the older group underwent EVT less often than their younger counterparts and experienced worse clinical outcomes. However, no increase in symptomatic intracerebral hemorrhage (sICH) was observed. These findings were in line with previous work and clinical experience. What was surprising was the dramatic narrowing of the age-associated treatment effect over the study period (2012-2019). At the beginning of the study, older patients were only half as likely to receive EVT but by the end of the study, the relative rate had increased to three-quarters as likely to receive it. Furthermore, no plateau was observed, implying this particular disparity may continue to shrink between older and younger patients eligible for EVT."

How do you see this study impacting the future of medicine and patient care?
"It is vitally important to recognize the treatment disparities in healthcare that continue to persist. The reasons behind them are complex but awareness remains central to create a more balanced healthcare system. I hope this work helps to serve as a reminder to providers and patients that while age can be a barrier to the most effective acute stroke therapies, it need not impede their implementation. Older stroke patients tend to experience worse clinical outcomes overall, however sICH rates are not higher in this population. In addition, this work highlights the value of carrying out future within-group outcome comparisons between older eligible patients who underwent EVT and those who did not."

June 2022:

Kershaw V. Patel, MD, MSCS
Assistant Professor of Cardiology
Houston Methodist Academic Institute
Houston Methodist DeBakey Heart & Vascular Center

Data Registry: Get With The Guidelines^® - Heart Failure

Published Study: Association of readmission penalty amount with subsequent 30-day risk standardized readmission and mortality rates among patients hospitalized with heart failure: An analysis of Get With The Guidelines - Heart Failure participating centers (Am Heart Journal)

When you started this study, what were you aiming to learn?
"Hospitals with readmission rates for heart failure that were higher than expected incurred financial penalties under the Hospital Readmissions Reductions Program. We were interested in understanding the implications of these financial penalties on readmissions and death."

Who was your target patient population and how did you study them?
"Our study included patients who were at least 65 years of age and hospitalized for heart failure. We studied these patients using data from the American Heart Association’s Get With The Guidelines – Heart Failure registry."

Did the study results turn out as you had expected or were you surprised with what you found?
"We hypothesized that hospitals that incurred higher financial penalties would have lower readmissions and higher mortality over follow-up. In our study, we found that readmissions and mortality remained mostly constant over time across hospitals that incurred varying amounts of financial penalties."

How do you see this study impacting the future of medicine and patient care?
"Our study findings highlight that a large proportion of hospitals received financial penalties due to excess readmissions but this policy was not associated with meaningful improvements in patient outcomes. Future studies are needed to evaluate the impact of healthcare policies on patients with heart failure prior to widespread adoption."

March 2022:

Saket Girotra, MD, SM
Associate Professor of Medicine
University of Iowa

Data Registry: Get With The Guidelines - Resuscitation

Published Study: Association of COVID-19 Infection With Survival After In-Hospital Cardiac Arrest Among US Adults; JAMA Network Open.

When you started this study, what were you aiming to learn?
"Early during the COVID-19 pandemic, several single-center studies reported poor survival in COVID-19 patients who experienced a cardiac arrest. As a result, many health systems began considering withholding CPR in patients with COVID-19 infection, out of concerns of PPE shortage and transmission of infection. We wanted to learn whether these findings were consistent among a diverse group of U.S. hospitals.”

Who was your target patient population and how did you study them?
"We included adult patients aged 18 years and older with in-hospital cardiac arrest during 2020 in GWTG-Resuscitation. We compared survival outcomes in patients who were suspected or confirmed to have COVID-19 infection at the time of arrest with non-COVID patients after adjusting for baseline differences between these groups.”

Did the study results turn out as you had expected or were you surprised with what you found?
"We were surprised that among ~25,000 patients included in our study, approximately 1 in 4 were COVID+ at the time of arrest. This highlights the commitment of healthcare professionals on the front lines in caring for these really sick patients, even when they arrest. We found that survival in COVID+ patients with cardiac arrest was 11.9% compared to 23.5% survival in non-COVID patients. However, we were surprised to learn that cardiac arrest survival in COVID+ patients was not as grim as the <3% survival reported in earlier studies.”

How do you see this study impacting the future of medicine and patient care?
"Sharing these data with the research community are important as they highlight that all CPR in COVID-19 patients is not futile."

February 2022:

Joseph E. Tonna, MD, MS
Associate Professor with Tenure
Section Head, Cardiothoracic Critical Care
Medical Director, Cardiovascular ICU/Cardiothoracic Critical Care
Associate Director, ECMO Services
University of Utah Health

Data Registry: Get With The Guidelines - Resuscitation

Published Study: Resuscitation Using ECPR During In-Hospital Cardiac Arrest (RESCUE-IHCA) Mortality Prediction Score and External Validation; JACC Cardiovasc Interv.

When you started this study, what were you aiming to learn?
"We were hoping to understand the patient population who would most benefit from extracorporeal cardiopulmonary resuscitation (ECPR), and likewise define the types of patients that ECPR may not benefit. At the time we started the research, ECPR had not been shown superior to conventional cardiopulmonary resuscitation for any population in a randomized controlled trial (RCT), and so we hoped that by defining the population who would most benefit, we could optimize clinical trial enrollment."

Who was your target patient population and how did you study them?
"Our target population included all adults with in-hospital cardiac arrest from the AHA GWTG-R database. We retrospectively analyzed these patients to first understand the way in which ECPR was utilized (Tonna, JAHA, 2020) and then to identify the characteristics of the ECPR patients and of their resuscitation that predicted their survival"

Did the study results turn out as you had expected or were you surprised with what you found?
"We were surprised as to how much more likely ECPR was to be utilized among surgical and cardiac patients than among non-surgical and non-cardiac patients. We suspected that ECPR would be used more in this population, but the magnitude of the difference was surprising. We expected many of the characteristics that predicted survival, but were again a little surprised by their magnitude, especially the influence of patient type and time of day on survival, and on how important rapid ECPR cannulation (short resuscitation) was.”

How do you see this study impacting the future of medicine and patient care?
"The findings demonstrate that survival after ECPR is much greater among select types of adult patients, and likewise lower among other types. As ECPR is a complex therapy with high potential for benefit, but also potential for harm and cost without benefit, our research should inform future clinical trials by helping define the optimal population to enroll. These trials are needed to prospectively confirm the thresholds of benefit based on arrest characteristics (rhythm, duration, time of day) and patient characteristics (age, medical conditions). There are likely patient populations in whom ECPR is largely superior, but also populations in whom it isn’t. Our findings suggest what each of these groups look like, but now we need RCTs to confirm and further define the benefit. Future research is also needed to understand the optimal way to implement ECPR, including team composition, case volume, location, cannulation training, etc, as ECPR survival is highly dependent on the team/providers performing it and subsequently caring for these patients."

December 2021:

Saate S. Shakil, MD
Research Fellow, Division of Cardiology
University of Washington

Data Registry: COVID-19 CVD Registry

Published Study: Stroke Among Patients Hospitalized With COVID-19: Results From the American Heart Association COVID-19 Cardiovascular Disease Registry; Stroke.

When you started this study, what were you aiming to learn?
"A high risk of cardiovascular complications, including stroke, has been reported among coronavirus disease 2019 (COVID-19) patients. We sought to explore prevalence and predictors of acute ischemic stroke among hospitalized COVID-19 patients.”

Who was your target patient population and how did you study them?
"To explore this phenomenon, we designed and implemented a retrospective cohort analysis of the American Heart Association’s COVID-19 Cardiovascular Disease Registry of 21,073 patients hospitalized with SARS-CoV-2 infection. This registry contains patient-level data from 101 hospitals across the United States, and followed patients for the duration of their COVID-19 hospitalization. We examined data on demographics, medical history, COVID-19 illness severity, and in-hospital outcomes, and compared these features between hospitalized COVID-19 patients who experienced an ischemic stroke or transient ischemic attack (TIA) immediately prior to or during their admission and COVID-19 patients who did not have a stroke. We also examined other stroke subtypes in this population.”

Did the study results turn out as you had expected or were you surprised with what you found?
"The prevalence of acute ischemic stroke/TIA in our study was 0.75%, while the prevalence of other stroke subtypes was 0.61%. As expected, we found that some traditional ischemic stroke risk factors, including history of hypertension and atrial fibrillation, were predictive of ischemic stroke/TIA in this population. In addition, requiring ICU admission and mechanical ventilation, both markers of COVID-19 illness severity, were also associated with ischemic stroke/TIA. However, after adjusting for demographics, comorbidities, and acute illness severity, ischemic stroke risk was highest in middle-aged individuals, rather than the oldest adults in the registry, which was unexpected given that atherosclerosis, which is generally more prevalent with age, is a key underlying mechanism for ischemic stroke. This is remarkable and suggests a potential mechanism for vascular events in COVID-19 that is independent of age-related atherosclerotic pathways.”

How do you see this study impacting the future of medicine and patient care?
"These results illustrate the importance of rapid implementation and analysis of disease registries to guide clinical decisions and identify high-risk populations for vascular complications during an evolving pandemic. Recognizing the limitations associated with all observational research, hospital-based disease registries can yield important hypothesis-generating findings that can lay the foundation for further rigorous longitudinal research.”

November 2021:

Fahd Yunus, MD
Cardiologist, Palo Alto Medical Foundation
Mountain View, CA

Data Registry: Get With The Guidelines - AFib

Published Study: Sex Differences in Ablation Strategy, Lesion Sets, and Complications of Catheter Ablation for Atrial Fibrillation: An Analysis From the GWTG-AFIB Registry; Circ Arrhythm Electrophysiol.

When you started this study, what were you aiming to learn?
"We have previously shown that women present to ablation with more advanced afib and have a greater risk of complications. At the same time, others have shown that women can often have electrical triggers of afib that are different than men and that pulmonary vein isolation might not be enough. We wanted to understand if the ablation that women were receiving might tie into some of these observations.”

Who was your target patient population and how did you study them?
"Our study population included patients with atrial fibrillation who underwent catheter ablation of afib. We studied patients from the Get With the Guidelines-AFib registry, which is a multi-center registry with the aim of improving quality of care of patients with atrial fibrillation. We worked with a stellar team at the AHA, Duke Clinical Research Institute, and collaborators who are all involved with GWTG-AFib.”

Did the study results turn out as you had expected or were you surprised with what you found?
"A key finding is that among patients undergoing ablation for non-paroxysmal atrial fibrillation, women were more likely to receive additional lesion sets beyond pulmonary vein isolation (which is the core ablation lesion set performed in almost all cases). For non-paroxysmal afib, we found that women also had a numerically high risk of complications, although this risk was not higher when accounting for baseline patient factors. So, on one hand, women were getting more ablation lesion sets because they are known to have triggers than were different than men. On the other hand, the complication rate was numerically higher.”

How do you see this study impacting the future of medicine and patient care?
"Our hope is that this study can raise awareness to the current practices within the atrial fibrillation ablation field. We believe that there isn’t a one-size-fits-all approach to atrial fibrillation ablation. We know that women have different electrical triggers for afib. What we don’t know is whether going after these triggers is the right thing to do. Randomized trials — specifically in women — to help determine whether certain lesion sets should be used — is something we should be thinking about.”

October 2021:

Blake Tyler McGee, PhD, MPH, RN
Assistant Professor
Byrdine F. Lewis College of Nursing and Health Professions
Georgia State University
Atlanta, GA

Data Registry: Get With The Guidelines - Stroke

Published Study: Associations of Medicaid Expansion with Access to Care, Severity, and Outcomes for Acute Ischemic Stroke; Circ Cardiovas Qual Outcomes

When you started this study, what were you aiming to learn?
"A dozen states have declined to expand Medicaid under the terms of the Affordable Care Act (ACA), and these states are located primarily in the Stroke Belt, which is concentrated in the Southeast. Therefore, thousands of low-income patients at risk of stroke remain uninsured. We wanted to understand what relationships the Medicaid expansion decision has had with outcomes for ischemic stroke patients, because health insurance not only protects against catastrophic financial loss, but also is linked to greater use of preventive care and fewer delays in seeking emergency care. Medicaid specifically is associated with lower reported inability to afford medications, better blood pressure control, and—for stroke survivors—higher odds of receiving institution-based rehabilitation."

Who was your target patient population and how did you study them?
"We began with over 340,000 patients hospitalized for a first-ever ischemic stroke from 2012 through 2018 at one of the 2,000+ hospitals participating in the Get With The Guidelines–Stroke program. The patients had to be under age 65, because older people are usually eligible for Medicare regardless of state Medicaid policy, and we excluded a handful of states that already offered widespread health insurance to low-income adults before the ACA’s Medicaid expansions began. In this sample, we looked at how insurance coverage and type of payer changed over time. Then, we looked closer at the 28% of patients with Medicaid or no insurance, because they were the likeliest to have low incomes and be affected by Medicaid expansion policy. Among these 95,000-odd patients, we studied differences between Medicaid expansion and non-expansion states with respect to changes in stroke severity, time from symptom onset to hospital arrival, odds of EMS arrival, in-hospital outcomes (e.g., length of stay and mortality), level of function at discharge, and odds of discharge to a rehab facility."

Did the study results turn out as you had expected or were you surprised with what you found?
"As expected, states that expanded Medicaid saw a net decline of 3.5% in uninsured ischemic stroke hospitalizations and a net increase of 5.3% in stroke hospitalizations covered by Medicaid. After risk adjustment, Medicaid expansion was associated with 33% increased odds of discharge to a skilled nursing facility (SNF) in the low-income group, and 24% higher odds of transfer to any rehab facility (if eligible). However, I was surprised that we did not observe an association between Medicaid expansion and stroke severity or the timeliness of emergency care, because prior studies had suggested that Medicaid improves access to clinical management of stroke risk factors and shortens prehospital delays during heart attacks. I was less surprised not to observe a relationship with in-hospital outcomes—to the extent that quality of inpatient care is unaffected by insurance status—or with receipt of inpatient rehab, because Medicaid coverage of inpatient rehab (vs. SNF care) is variable, and Medicaid members may lack the resources for the up-front discharge plan that inpatient rehab facilities require."

How do you see this study impacting the future of medicine and patient care?
"Clinicians should be aware of the larger policy environment in which they deliver care, especially the impact of policy choices on vulnerable populations. Our study suggests that Stroke Belt states which have not expanded Medicaid have more patients and/or hospitals at risk of financial hardship, and that expanding Medicaid may lead to better functional outcomes for low-income stroke survivors by raising the chance that a patient will be transferred to a SNF rather than discharged straight home. However, our study may also manage expectations that Medicaid expansion will solve all access-to-care issues. In the future, it would be useful to directly examine the role of expanded Medicaid in stroke prevention."

September 2021:

Anoop Mayampurath, PhD
Assistant Professor, Biostatistics & Medical Informatics
University of Wisconsin-Madison
Madison, WI

Data Registry: Get With The Guidelines - Resuscitation

Published Study: Comparison of machine learning methods for predicting outcomes after in-hospital cardiac arrest; Critical Care Medicine

When you started this study, what were you aiming to learn?
"Prognostication of neurological outcomes for in-hospital cardiac arrest survivors is challenging. Researchers have developed risk scores to help determine the likelihood that a resuscitated patient will survive to discharge with a favorable neurological status. However, these scores were developed using standard statistical techniques. We wanted to explore if we could better identify resuscitated patients with favorable neurological outcomes using machine learning methods."

Who was your target patient population and how did you study them?
"From the GWTG-Resuscitation data registry, we obtained a cohort of 117,674 patients who were successfully resuscitated within a hospital and had recorded neurological outcomes at discharge. We used patient characteristics, pre-existing conditions, pre-arrest interventions, and peri-arrest variables to predict survival with a favorable neurological outcome. We developed several machine learning models, such as logistic regression, random forests, support vector machines, neural networks and extreme gradient boosted (XGB) models, and compared their performance to each other and to the previously published CASPRI score in terms of discrimination, sensitivity, specificity, positive and negative predictive values, and calibration."

Did the study results turn out as you had expected or were you surprised with what you found?
"The discriminative performance of our logistic regression model was equivalent to that of more advanced machine learning models and higher than the CASPRI score. This is similar to other studies wherein the data that are fed into the model are highly structured. However, in the end, the XGB model was the winner when considering other clinically relevant prediction metrics."

How do you see this study impacting the future of medicine and patient care?
"Our study highlights the ability of machine learning to predict neurological outcomes in resuscitated patients. The scores returned from our model could be used for quality initiatives and patient risk adjustment. As we continue to improve and validate these types of models, in the future they could also be useful for clinicians and families at the bedside."

August 2021:

David M. Tehrani, MD, MS
Fellow in Cardiovascular Diseases
Division of Cardiology, Department of Medicine
Ronald Reagan UCLA Medical Center
Los Angeles, CA

Data Registry: COVID-19 CVD Registry

Published Study: Impact of cancer and cardiovascular disease on in-hospital outcomes of COVID-19 patients: results from the american heart association COVID-19 cardiovascular disease registry; Cardio-Oncology

When you started this study, what were you aiming to learn?
"Underlying cardiovascular disease (CVD) has been previously associated with poor outcomes in patients hospitalized with Coronavirus Disease 2019 (COVID-19). However, data on patients with CVD and concomitant cancer--both leading causes of morbidity and mortality in the United States--is limited. As oncologists and cardio-oncologists consider medical and surgical treatments of cancer (including potential cardiotoxic complications) in a patient population at high risk for COVID-19 infections, there is a paucity of outcome data to help guide this complex clinical decision-making. Therefore, we sought to address this evidence gap and evaluate the association of underlying CVD, history of cancer, and recent cancer-related therapy with clinical outcomes adult patients hospitalized with COVID-19."

Who was your target patient population and how did you study them?
"We obtained patient-level data for hospitalized adults with COVID-19 among 86 US hospitals registered in the American Heart Association Get with the Guidelines COVID-19 CVD registry. We examined the association of CVD risk factors, underlying CVD, and history of cancer with in-hospital mortality as well as other key clinical endpoints. Additionally, we evaluated if patients with recent cancer therapy, within two weeks prior to hospitalization, had worse clinical outcomes."

Did the study results turn out as you had expected or were you surprised with what you found?
"First, we found that a history of cancer was in itself independently associated with increased in-hospital mortality. However, we were surprised to find that among cancer patients, having a history of CVD did not necessarily translate to worse outcomes. However, those cancer patients with concomitant CVD risk factors did have higher in-hospital mortality compared to their counterparts. Importantly, a history of recent cancer therapy was associated with increased in-hospital mortality among cancer patients."

How do you see this study impacting the future of medicine and patient care?
"It is important to note the limitations of the registry: 1) the type of malignancy and whether it was active or not was not defined, and 2) cancer treatments, which vary in their cardiotoxic and immunosuppressive properties, were not captured. Nonetheless, as the COVID-19 pandemic continues both nationally and worldwide with potential compromise of hospital resources devoted to COVID-19 related care, our findings will hopefully be useful for oncologists and cardio-oncologists as they weigh the risks and benefits therapeutic strategies for balancing cancer therapy, cardiovascular care, and/or cardiotoxicity surveillance for their patients."

July 2021:

Sara C. Handley, MD, MSCE
Instructor of Pediatrics, Perelman School of Medicine, University of Pennsylvania
Attending Neonatologist, The Children's Hospital of Philadelphia
Philadelphia, PA

Data Registry: Get With The Guidelines - Resuscitation (Pediatric) Registry

Published Study: Epidemiology and outcomes of infants after cardiopulmonary resuscitation in the neonatal or pediatric intensive care unit from a national registry; Resuscitation

When you started this study, what were you aiming to learn?
"The research surrounding risk factors and outcomes of infants after a CPR event is limited. We hypothesized that there are a variety of factors that may influence infant outcomes. Our goal was to understand how different types of factors, including patient factors (e.g., markers of illness prior to the CPR event), CPR event factors (e.g., event interventions), unit factors (e.g., the location of the event), and hospital factors (e.g., hospital type) are associated with infant outcomes after a CPR event."

Who was your target patient population and how did you study them?
"We were interested in studying infants (≤365 days old) without congenital heart disease who experienced a CPR event in the intensive care unit (ICU). The GWTG-Resuscitation registry allowed us to identify these high-risk infants who experienced a CPR event in the neonatal or pediatric ICU."

Did the study results turn out as you had expected or were you surprised with what you found?
"We were most surprised by the difference in outcomes between the NICU and PICU, as the unit location was the factor most strongly associated with mortality prior to discharge. This finding needs to be considered in the context of our models, as we were able to adjust for the infant's age and weight at the time of the event, but not their gestational age at birth or birth weight."

How do you see this study impacting the future of medicine and patient care?
"The differences in outcomes between care units highlights the opportunity to further study and understand differences in the approaches to CPR management in the neonatal and pediatric ICU as well as differences in other unit-based practices, processes, and policies."

June 2021:

Ravi B. Patel MD, MSc
Division of Cardiology, Department of Medicine
Northwestern University Feinberg School of Medicine
Chicago, IL

Data Registry:Get With The Guidelines - Heart Failure Registry

Published Study: Kidney Function and Outcomes in Patients Hospitalized with Heart Failure; Journal of the American College of Cardiology

When you started this study, what were you aiming to learn?
"We have come to understand that chronic kidney disease (CKD) frequently coexists with heart failure (HF), and patients with both CKD and HF have worse clinical outcomes than those without CKD. However, contemporary data surrounding risk associated with CKD in patients hospitalized for HF were lacking. In a population of patients hospitalized for HF, we aimed to understand differences in in-hospital mortality based upon degree of kidney dysfunction at admission to the hospital. Additionally, among patients with HF with reduced ejection fraction (HFrEF), we wanted to understand if there were differences in the prescription rates of certain evidence-based medications at discharge based on degree of kidney dysfunction."

Who was your target patient population and how did you study them?
"We aimed to study a population of patients hospitalized for HF. For this, we turned to the GWTG-HF registry, which has granular data on hospitalization for HF across several hundred US sites, and encompasses >500,000 patients. The GWTG-HF registry collects data regarding kidney function as measured by estimated glomerular filtration rate (eGFR) at admission and discharge. We categorized patients within GWTG-HF based on degree of eGFR both at admission and discharge by current societal recommendations (KDIGO guidelines). We then evaluated the association of eGFR category with in-hospital mortality and rates of prescription medical therapies at discharge."

Did the study results turn out as you had expected or were you surprised with what you found?
"While there were some expected results, there were several aspects that were surprising. Approximately 60% of patients hospitalized for HF have eGFR <60 mL/min/1.72m2 (at least Stage 3 CKD) and 5% are on dialysis. These patterns of CKD prevalence have remained stable over time. Worse kidney function at admission for HF was linked to higher in-hospital mortality across all subtypes of heart failure. Interestingly, this association between kidney dysfunction and mortality in the hospital was strongest among patients with HF with reduced ejection fraction. Additionally, we noted that all classes of evidence-based medical therapies for HFrEF were prescribed less frequently at discharge among individuals with worse kidney function. Importantly, prescription rates declined across kidney function categories even among groups that would not have contraindications to these medications based on eGFR level."

How do you see this study impacting the future of medicine and patient care?
"Our findings highlight that despite advances in HF care, CKD is highly prevalent and identifies a particularly high-risk cohort for in-hospital mortality. Despite this higher risk, patients with CKD were less likely to receive evidence based medical therapies for HFrEF. Given that several of these classes of medications have recently shown favorable long-term effects upon kidney function, further efforts to mitigate risk in this cohort and understand barriers to implementation of medications among patients with HFrEF and CKD are required."

May 2021:

Gregory Roth MD MPH, FACC FAHA
Associate Professor of Medicine-Cardiology
Associate Professor of Global Health and Health Metrics Science (Adjunct)
Attending Cardiologist, Harborview Medical Center
Division of Cardiology, Department of Medicine, University of Washington
Institute for Health Metrics and Evaluation
Seattle, WA

Data Registry: COVID-19 CVD Registry

Published Study: Trends in Patient Characteristics and COVID-19 In-Hospital Mortality in the United States During the COVID-19 Pandemic; JAMA Network Open

When you started this study, what were you aiming to learn?
"The AHA’s new COVID-19 registry offered a unique opportunity to understand hospital outcomes across more than a hundred hospitals in an up-to-the-moment way. Big changes in mortality were starting to show up in single-center studies but we didn’t know if this was just because of the different populations coming into the hospital. Were they younger, healthier, presenting earlier? The registry offered an opportunity to figure that out.".

Who was your target patient population and how did you study them?
"The AHA’s COVID-19 hospital registry includes inpatient adults with COVID-19. The registry has detailed data on past history, clinical characteristics, treatment, lab values and hospital disposition."

Did the study results turn out as you had expected or were you surprised with what you found?
"We were struck by how large the decline in mortality really was, and how early in the pandemic it occurred. Also, it is very clear that the decline in mortality was not due to any changes in the kind of patients being hospitalized. Unfortunately, after early summer there was very little further decline in that death rate."

How do you see this study impacting the future of medicine and patient care?
"The AHA registry clearly offers the chance for ongoing surveillance. With 10% of hospitalized patients dying of their COVID-19, this remains a supremely dangerous and mortal infection. We need to keep a laser focus on improving care and making sure the best practices and treatments are adopted and implemented in every hospital."

April 2021:

Pratyaksh K. Srivastava, MD
Cardiology Fellow
UCLA Medical Center
Los Angeles, CA

Data Registry: Get With The Guidelines^®-Stroke

Published Study: Acute Ischemic Stroke in Patients With COVID-19: An Analysis From Get With The Guidelines-Stroke; Stroke

When you started this study, what were you aiming to learn?
"The COVID-19 pandemic has created significant challenges in the delivery of acute stroke care. We sought to evaluate the clinical impact of the pandemic on those presenting with acute ischemic stroke (AIS)".

Who was your target patient population and how did you study them?
"We evaluated the clinical characteristics, treatment patterns, and outcomes of 41,971 patients from 458 hospitals in the Get With The Guidelines-Stroke registry with a diagnosis of acute ischemic stroke between February 4th, 2020 and June 29th, 2020. Of this population, 1,143 patients were also positive for COVID-19 (AIS/COVID-19)."

Did the study results turn out as you had expected or were you surprised with what you found?
"We found that patients with AIS/COVID-19 were more likely to be non-Hispanic Black, Hispanic, or Asian, had a greater proportion of large vessel occlusions, and presented with higher National Institutes of Health Stroke Scale scores compared to patients with AIS/no COVID-19. While rates of thrombolysis and thrombectomy were similar between the two groups, door to computed tomography, door to needle, and door to endovascular therapy times were all longer in the AIS/COVID-19 cohort. In adjusted models, patients with AIS/COVID-19 had decreased odds of discharge with a modified Rankin Scale score ≤2, and increased odds of in-hospital mortality when compared to those with AIS/no COVID-19."

How do you see this study impacting the future of medicine and patient care?
"These findings demonstrate the challenges associated with delivering acute stroke care to those with COVID-19, and highlight a potential need for enhanced protocols and interventions for those who present with AIS and a simultaneous diagnosis of COVID-19. "

March 2021:

Ying Xian, MD, PhD, FAHA
Associate Professor of Neurology and Medicine
Duke University
Durham, NC

Data Registry: Get With The Guidelines^®-Stroke

Published Study: Clinical Characteristics and Outcomes Associated With Oral Anticoagulant Use Among Patients Hospitalized With Intracerebral Hemorrhage;JAMA Network Open

When you started this study, what were you aiming to learn?
"Factor Xa (FXa) inhibitors such as apixaban, rivaroxaban, and edoxaban have increasingly gained popularity as first-line agents for stroke prevention in high-risk individuals with atrial fibrillation. Despite their improved safety profiles over warfarin, up to 0.7% of patients taking FXa inhibitors still experience an intracranial bleeding event with each year of treatment. Because of the low rate of bleeding events, there remains limited experience with FXa inhibitors-related intracerebral hemorrhage (ICH), the most feared complication of oral anticoagulation therapy (OAC)".

Who was your target patient population and how did you study them?
"Using data from the Get With The Guidelines® Stroke program, the largest stroke registry in the nation, we studied patients who experienced a non-traumatic ICH. We compared patients with preceding use of FXa inhibitors against those on warfarin or those without any OAC."

Did the study results turn out as you had expected or were you surprised with what you found?
"We found 27% ICH patients with preceding use of FXa inhibitors died during hospitalization. Although the mortality rates were higher than ICH with no OAC, patients on FXa inhibitors had 24% lower odds of death than ICH due to warfarin use. In addition, individuals on FXa inhibitors were more likely to be discharged home and have better functional outcomes at discharge than those treated with warfarin. Concomitant OAC and antiplatelet therapy were quite common among ICH patients. Among ICH on warfarin, both single and dual antiplatelet therapy were associated with worse outcomes as compared with those on warfarin alone. To our surprise, such incremental risk was not found in patients taking FXa inhibitors. This is clinically relevant because some patients may be required to take both anticoagulant and antiplatelet agents for a short period of time in certain clinical scenarios such as atrial fibrillation with recent PCI."

How do you see this study impacting the future of medicine and patient care?
"Although quite rare, FXa inhibitor-related ICH remains a devastating complication of oral anticoagulation. Having said that, many patients are required to take OAC to prevent thromboembolic events. Our findings suggest that FXa inhibitors may be a better choice than warfarin when OAC is warranted. Additionally, further study is warranted to identify the best treatment strategies, such as the anticoagulation reversal treatment, for FXa inhibitor-related ICH."

February 2021:

Daniel J. Friedman, MD
Assistant Professor of Medicine (Cardiology)
Yale New Haven Hospital
New Haven, CT

Data Registry: Get With The Guidelines^®- AFIB

Published Study: Procedure characteristics and outcomes of atrial fibrillation ablation procedures using cryoballoon versus radiofrequency ablation: A report from the GWTG-AFIB registry; Journal of Cardiovascular Electrophysiology

When you started this study, what were you aiming to learn?
"Radiofrequency ablation had long been the standard for catheter ablation of atrial fibrillation. However, over the last decade, cryoballoon ablation has increasingly gained popularity, particularly for paroxysmal atrial fibrillation. Despite this, there was relatively little real-world data on use of cryoballoon ablation, as compared to radiofrequency ablation. We sought to describe patterns of use and periprocedural outcomes associated with use of radiofrequency and cryoballoon ablation in the US. ".

Who was your target patient population and how did you study them?
"We studied all patients in the GWTG-AFIB Registry who underwent ablation of atrial fibrillation with either a radiofrequency catheter or a cryoballoon. Our study included patients with paroxysmal and persistent atrial fibrillation, including those undergoing de novo and repeat procedures."

Did the study results turn out as you had expected or were you surprised with what you found?
"The study demonstrated that reassuringly, rates of phrenic nerve injury, which was a concern with use of the cryoballoon, were significantly less common in the real world compared to the pivotal FIRE and ICE trial. However, phrenic nerve injury was still more common with cryoablation compared to radiofrequency ablation. We did observe that radiofrequency ablation cases, which use catheters that emit saline to prevent overheating and clot formation, more commonly had issues associated with this, including more issues with periprocedural volume overload. There were no differences in major complications like pericardial tamponade, stroke, or death. Overall, cryoballoon ablation was associated with fewer complications, although this finding was largely driven by fluid overload related complications in patients with persistent atrial fibrillation.

We were overall surprised to find that although the cryoballoon was initially studied for achieving pulmonary vein isolation in paroxysmal atrial fibrillation, it was being used very frequently in the setting of more advanced forms of atrial fibrillation and in the context of extra-pulmonary vein ablation."

How do you see this study impacting the future of medicine and patient care?
"This study demonstrates that the cryoballoon has an excellent safety profile and that rates of phrenic nerve injury are even lower in the real world than what were observed in the pivotal FIRE and ICE trial. These data are particularly important in light of the recent STOP-AF and EARLY-AF trials, which compared cryoballoon ablation with medical therapy for recently diagnosed paroxysmal atrial fibrillation. The safety data from our study suggests that, in the current era, both radiofrequency and cryoballoon ablation demonstrate excellent safety among patients with paroxysmal atrial fibrillation."

December 2020:

Frederik Dalgaard, MD, PhD
Herlev & Gentofte Hospital, Copenhagen, Denmark
Duke Clinical Research Institute, Durham, NC, USA

Data Registry: Get With The Guidelines^®- AFIB

Published Study: Management of Atrial Fibrillation in Older Patients by Morbidity Burden: Insights From Get With The Guidelines‐Atrial Fibrillation; JAHA

When you started this study, what were you aiming to learn?
"As clinicians, treating multimorbid patients is a complex and growing challenge. Older patients with atrial fibrillation(AF) often present with a high number of comorbidities and receive a higher number of prescriptions. Furthermore, multimorbidity increases the risk of stroke in patients with AF. These multiple comorbidities, which are indicative of an increasingly complex patient population, may influence therapeutic choices and quality of care. With sparse guidelines, it is unknown how multimorbid older AF patients are managed. We had two main objectives. Firstly, we wanted to investigate if the prevalence of patients with atrial fibrillation and multimorbidity was growing. Secondly, we wanted to investigate if the comorbidity burden affected the treatment decisions for patients with atrial fibrillation".

Who was your target patient population and how did you study them?
"We investigated hospitalized patients with AF who were more than 65 years old from the Get With The Guidelines®-Atrial Fibrillation registry from 2013 to 2019. The patients had to be eligible for oral anticoagulation therapy. We used a temporal trend analysis to conclude if the multimorbidity was growing in patients with AF and we used logistic regression models to determine whether multimorbidity was associated with oral anticoagulant treatment at discharge.."

Did the study results turn out as you had expected or were you surprised with what you found?
"We found that multimorbidity was extremely common and prevalent in almost two thirds of the patients with AF over 65 years of age. During the study period, multimorbidity increased with 4.9% which was more than we had expected. The most common comorbidities were hypertension, heart failure, coronary artery disease, and diabetes. The overall prescription rate of oral anticoagulants was quite high, 8 out of 10 received oral anticoagulation at discharge. However, having a high multimorbidity burden was associated with fewer prescription of oral anticoagulation at discharge. To our surprise, the most common reason for non-prescription was frequent falls and frailty. Reasons that research have shown for non-use of oral anticoagulants, in particular in the era of novel oral anticoagulants, are rarely good. The very high rates of oral anticoagulation used in this study may not reflect other healthcare settings, since the patients in this study was recruited in a quality of care database that aims to improve stroke prevention in patients with AF. This could potentially limit the generalizability of the results.."

How do you see this study impacting the future of medicine and patient care?
"Healthcare workers and physicians should be aware that multimorbidity in AF is common and rising in prevalence. Furthermore, our study highlights a gap between current knowledge of atrial fibrillation management and clinical misperceptions regarding the benefits and risks of oral anticoagulation therapy. The study implies the need for initiatives that can close this gap between knowledge and clinical practice to improve the use of oral anticoagulation therapy in patients with high multimorbidity, mitigating the high stroke risk in this population."

October 2020:

Feras Akbik, MD, PhD
Assistant Professor, Neurology and Neurosurgery
Emory University School of Medicine

Data Registry: Get With The Guidelines^®-Stroke

Published Study: Trends in Reperfusion Therapy for In-Hospital Ischemic Stroke in the Endovascular Therapy Era; JAMA Neurology(link opens in new window)(link opens in new window)(link opens in new window)

When you started this study, what were you aiming to learn?
"As a resident physician, I often found that inpatient stroke alerts were a frustrating mixture of confusion and delay, while strokes in the emergency department were routine. I wanted to better understand whether this was a more widespread experience, and if so, what the impact of these delays were on patient outcomes. The hope was to identify tractable targets to guide efforts to improve stroke therapy for patients already under our care in the hospital".

Who was your target patient population and how did you study them?
"Leveraging the national Get With The Guidelines-Stroke registry, we found that the in-hospital onset strokes are increasingly reported from 2008-2018. Our target population was patients who received reperfusion therapy, either with mechanical thrombectomy and or intravenous thrombolysis. We dichotomized these patients by site of stroke onset, either in or out of the hospital, and then compared quality metrics as well as functional outcomes."

Did the study results turn out as you had expected or were you surprised with what you found?
"We predicted that the use of intravenous thrombolytics would be fairly constant throughout the study period, while endovascular therapy would significantly increase in 2015 (after publication of the pivotal clinical trials). We did see a significant increase in endovascular therapy starting in 2015, but we were surprised to see a steady, year-over-year increase in intravenous thrombolysis, more than doubling over the 10-year period."

How do you see this study impacting the future of medicine and patient care?
"Hospitals that have acknowledged the importance and morbidity of inpatient strokes have developed and practiced inpatient stroke protocols, minimizing the disparities in treatment and outcomes after in-hospital onset strokes. We hope that by highlighting the persistence of these disparities nationwide helps our colleagues around the country revisit their inpatient stroke protocols. If they do not have one, this is a call to develop and implement one. If they do have one, our data highlight the importance of reevaluating and practicing the protocol to maximize patient outcomes. Triage will always be faster in the emergency department; it's what they do all day, every day. But we can do better upstairs."

September 2020:

Madeline Sterling, MD, MPH, MS
Associate Professor of Medicine
Weill Cornell Medicine, New York, NY

Data Registry: Get With The Guidelines^®-Heart Failure

Published Study: Home Health Care Use and Post-Discharge Outcomes After Heart Failure Hospitalizations JACC: Heart Failure

When you started this study, what were you aiming to learn?
"There have been a few studies to suggest that home health care (HHC) use has been increasing among older adults discharged home after a heart failure (HF) hospitalization, but few contemporary studies have investigated this at the national level. HHC provides patients with home-based personal and skilled medical care after discharge, so our team wanted to understand how these services were being used and how they influenced post-discharge outcomes among HF patients. In that context, we had two main objectives. First, we aimed to describe utilization patterns of HHC among Medicare beneficiaries after a HF hospitalization. Second, we aimed to examine associations among HHC and post-discharge outcomes, including short-term readmission and mortality.

Who was your target patient population and how did you study them?
"We studied Medicare beneficiaries hospitalized at American Heart Association’s Get With The Guidelines-HF (GWTG-HF) centers between 2005 and 2015 with a primary discharge diagnosis of HF, who had Medicare-linked data available, and were discharged home alive. Propensity score matching and Cox proportional hazards models were used to evaluate the associations between HHC and post-discharge outcomes."

Did the study results turn out as you had expected or were you surprised with what you found?
"We found that more than one third of Medicare beneficiaries hospitalized for HF in the US were discharged home with HHC; this is equivalent to an absolute increase of 5% from 2005 to 2015. HHC recipients were older, more likely to be female, and had more comorbidities than those who did not receive HHC after discharge. We found that HHC recipients had a higher risk of 30 and 90-day all-cause and HF-specific readmission and a higher risk of all-cause mortality than those who did not receive HHC. This surprised us, as we hypothesized that HHC would be associated with a decrease in the risk of these outcomes. However, we urge readers to interpret the results cautiously, as the allocation of HHC was not random, and unmeasured confounding may have persisted. That is, while we adjusted for many covariates in our analyses, we were unable to account for some important ones (such as patients’ functional needs and cognition, HF severity, and social support) which may affect HHC and post-discharge outcomes.

How do you see this study impacting the future of medicine and patient care?
"Health care providers and health systems should be aware of the increasing utilization of HHC among older adults discharged home after a HF hospitalization in the US. Further research is needed to determine the intensity and type of services patients received with their HHC, something we were unable to do in this study, but which may impact outcomes. Additionally, research is needed to understand the appropriateness and preventability of readmissions among patients with HF sick enough to receive HHC after discharge."

August 2020:

Saket Girotra, MD, SM
Associate Professor of Medicine
University of Iowa Carver College of Medicine

Data Registry: Get With The Guidelines^®- Resuscitation

Published Study: Association of Hospital-Level Acute Resuscitation and Post resuscitation Survival With Overall Risk-Standardized Survival to Discharge for In-Hospital Cardiac Arrest JAMA Network Open

When you started this study, what were you aiming to learn?
"Prior work from GWTG-Resuscitation has shown that survival for survival for in-hospital cardiac arrest (IHCA) varies by more than 3-fold. We wanted to find out whether high survival at top-performing hospitals is because they excel in running codes, excel in post-resuscitation care, or both.

Who was your target patient population and how did you study them?
"We used data from 290 hospitals participating in GWTG-R during 2015-2018. For each hospital in our study, we calculated risk-adjusted rates of acute resuscitation survival, defined as return of spontaneous circulation (ROSC) for at least 20 minutes, post-resuscitation survival, defined as survival to discharge, in patients who achieved ROSC and overall survival using a validated methodology. We examined the extent to which hospital rates of overall survival were correlated with rates of acute resuscitation and post-resuscitation survival."

Did the study results turn out as you had expected or were you surprised with what you found?
"We found that a hospital’s overall risk-standardized survival rate (RSSR) was correlated with both acute resuscitation and post-resuscitation survival which was expected because in order for patients to survive, they must achieve ROSC and survive during the post-resuscitation phase. However, the strength of the correlation between overall RSSR and post-resuscitation survival was much stronger compared to its correlation with acute resuscitation survival (rho=0.90 vs. 0.50). Additionally, we found no correlation between acute resuscitation and post-resuscitation survival. This finding was rather unexpected and suggested that hospitals that excel in IHCA survival, either excel in acute resuscitation, or post-resuscitation survival but not consistently both phases.

How do you see this study impacting the future of medicine and patient care?
"Our findings have important implications for quality improvement for resuscitation care. The current AHA GWTG-R quality measures that incentivize hospitals are focused only on acute resuscitation. Our findings suggest that in order for us to achieve continued improved in overall IHCA survival, there is an urgent need to identify best practices for post-resuscitation survival."