Medicare services provided by cardiologists in the United States: 1999-2008

Prognostic Value of Stress Myocardial Perfusion Imaging Across the Spectrum of Cardiovascular Risk

BACKGROUND

Single-photon emission computed tomography (SPECT) myocardial perfusion imaging (MPI) is widely used to identify ischemia. There is limited research to evaluate if there is a risk threshold below which SPECT-MPI may not add significant prognostic value.

METHODS

Between January 1, 2012, and December 31, 2018, individuals who underwent SPECT-MPI were stratified into 4 risk groups. The primary outcome was acute myocardial infarction (MI) or death. Multivariable Cox proportional hazards regression analysis was used to calculated hazard ratios (HRs) with 95% confidence intervals (CIs).

RESULTS

Among 48,845 patients (52.3% male, median age 67 years), 8.5% were low risk, 4.8% borderline risk, 18.1% intermediate risk, and 68.6% high risk based on the American College of Cardiology pooled cohort equation. Ischemia was more commonly detected in the high-risk cohort (19.4% in high-risk vs 6.5% in low-risk). SPECT-MPI testing was associated with a significantly increased use of preventive medications such as statin therapy, regardless of stress test results. At a median follow-up of 4.2 years, there was no significant association between ischemia and death or MI in the low-risk cohort (adjusted HR, 1.91; 95% CI, 0.94-3.92) or the borderline-risk cohort (adjusted HR, 1.58; 95% CI, 0.79-3.15). Ischemia was associated with a higher risk of death or MI in the intermediate-risk (adjusted HR, 1.57; 95% CI, 1.24-1.99) and high-risk groups (adjusted HR, 1.54; 95% CI, 1.44-1.64).

CONCLUSIONS

SPECT-MPI was less useful for risk stratification among low-risk patients because of their low event rates regardless of test results.

Echocardiography Does not Reduce Mortality in Sepsis: A Re-Evaluation Using the Medical Information Mart for Intensive Care IV Dataset

OBJECTIVES

Echocardiography is commonly used for hemodynamic assessment in sepsis, but data regarding its association with outcome are conflicting. The aim of this study was to evaluate the association between echocardiography and outcomes in patients with septic shock using the Medical Information Mart for Intensive Care IV database.

DESIGN

Retrospective cohort study comparing patients who did or did not undergo transthoracic echocardiography within the first 5 days of admission for the primary outcome of 28-day mortality.

SETTING

Admissions to the Beth Israel Deaconess Medical Center intensive care from 2008 to 2019.

PATIENTS

Adults 16 years old or older with septic shock requiring vasopressor support within 48 hours of admission. Readmissions and patients admitted to the coronary care unit or cardiovascular intensive care were excluded, as well as patients with ST-elevation myocardial infarction or cardiac arrest.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Echocardiography was performed in 1,515 (27%) of 5,697 eligible admissions. The primary outcome was analyzed using a marginal structural model and rolling entry matching to adjust for baseline and time-varying confounders. Patients who underwent echocardiography showed no significant difference in 28-day mortality (adjusted hazard ratio 1.09; 95% CI, 0.95-1.25; p = 0.24). This was consistent across multiple sensitivity analyses. Secondary outcomes were changes in management instituted within 4 hours of imaging. Treatment changes occurred in 493 patients (33%) compared with 431 matched controls (29%), with the most common intervention being the administration of a fluid bolus.

CONCLUSIONS

Echocardiography in sepsis was not associated with a reduction in 28-day mortality based on observational data. These findings do not negate the utility of echo in cases of diagnostic uncertainty or inadequate response to initial treatment.

The Quality and Safety of Sedation and Monitoring in Adults Undergoing Nonoperative Transesophageal Echocardiography

Sedation is an essential component of the transesophageal echocardiography (TEE) procedure for patient comfort. The use and the clinical implications of cardiologist-supervised (CARD-Sed) versus anesthesiologist-supervised sedation (ANES-Sed) are unknown. We reviewed nonoperative TEE records from a single academic center over a 5-year period and identified CARD-Sed and ANES-Sed cases. We evaluated the impact of patient co-morbidities, cardiac abnormalities on transthoracic echocardiogram, and the indication for TEE on sedation practice. We analyzed the use of CARD-Sed versus ANES-Sed in light of institutional guidelines; the consistency in the documentation of preprocedural risk stratification; and the incidence of cardiopulmonary events, including hypotension, hypoxia, and hypercarbia. A total of 914 patients underwent TEE, with 475 patients (52%) receiving CARD-Sed and 439 patients (48%) receiving ANES-Sed. The presence of obstructive sleep apnea (p = 0.008), a body mass index of >45 kg/m2 (p <0.001), an ejection fraction of <30% (p <0.001), and pulmonary artery systolic pressure of more than 40 mm Hg (p = 0.015) were all associated with the use of ANES-Sed. Of the 178 patients (19.5%) with at least 1 caution to nonanesthesiologist-supervised sedation by the institutional screening guideline, 65 patients (36.5%) underwent CARD-Sed. In the ANES-Sed group, where intraprocedural vital signs and medications were documented in all cases, hypotension (n = 91, 20.7%), vasoactive medication use (n = 121, 27.6%), hypoxia (n = 35, 8.0%), and hypercarbia (n = 50, 11.4%) were noted. This single-center study revealed that 48% of the nonoperative TEE used ANES-Sed over 5 years. Sedation-related hemodynamic changes and respiratory events were not infrequently encountered during ANES-Sed.

Role of echocardiography in acute pulmonary embolism

Although pulmonary embolism (PE)-related mortality rate has decreased because of prompt diagnosis and effective therapy use, acute PE remains a potentially lethal disease. Due to its increasing prevalence, clinicians should pay attention to diagnosing and managing patients with acute PE. Echocardiography is the most commonly used method for diagnosing and managing acute PE; it also provides clues about hemodynamic instability in an emergency situation. It has been validated in the early risk stratification and impacts management strategies for treating acute PE. In hemodynamically unstable patients with acute PE, echocardiographic detection of right ventricular dysfunction is an indication for administering thrombolytics. In this review article, we discuss the role of echocardiography in the diagnosis and management of patients with acute PE.

Do I really need this transthoracic ECHO? An over-utilized test in trauma and surgical intensive care units

INTRODUCTION

Clinical use of transthoracic echocardiogram (TTE) in intensive care units (ICU) has dramatically increased without clear guidance on validated assessment indications, appropriateness, and patient value.

METHODS

A retrospective analysis of consecutive TTEs performed among patients admitted to a tertiary trauma/surgical ICU over 2.5 years was performed. A bivariate analysis and Poisson regression was used to compare patients who received a TTE. Sensitivity analysis was performed to assess patient factors that predict change in management based on TTE. An abnormal exam was defined as having at least one of the following: ejection fraction < 55%, wall motion, pericardial effusion, pericardial effusion, or other significant abnormality including filling defect. The effect on management was derived from clinical course. We hypothesize that these studies are usually normal and rarely lead to changes in clinical management.

RESULTS

912 TTEs were performed in 806 patients. The median age was 68 years (IQR 57, 77) and 63.5% were male. Syncope (21.7%) or hypotension/hypovolemia (20.5%) were the most common indications for a TTE. In total, 39.4% TTEs were abnormal and only 7.6% resulted in a change in management. Predictive factors associated with an abnormal exam included: age >50, serum troponin ≥0.1 ng/ml, abnormal ECG, and clinical suspicion of heart failure or acute myocardial infarction. A troponin cutoff level <0.25 ng/mL was the most reliable factor to predict no change in management after TTE with a negative predictive value of 94.3% (95% CI 93.1, 95.3).

CONCLUSION

TTE is commonly used for patient assessment in critically ill surgical patients but the majority of exams are normal without change in clinical management. Certain patient factors, such as troponin level, may help distinguish which patients would benefit from this diagnostic test. Given the considerable cost associated with TTE and the minimal effect on management, guidelines on appropriate use would provide improved patient value.

The effect of body mass index on high versus low administered activity protocol myocardial perfusion imaging scan time and effective dose using a cadmium zinc telluride camera in clinical practice

Cadmium Zinc Telluride (CZT) crystal-based myocardial perfusion imaging (MPI) cameras have increased count sensitivity compared to Anger cameras and can be used to lower either the injected activity or the image acquisition time. Institutions adopting CZT cameras need to decide whether to lower the injected activity or imaging time or attempt to lower both with a compromise. The aim of our study was to compare the scan time required to obtain similar count images using high activity protocol (HAP) versus low activity protocol (LAP) stratified by body mass index (BMI) and assess the impact on effective dose and our clinic workflow. Using a CZT camera, a cohort of 100 consecutive clinical patients imaged with LAP rest-stress MPI with approximately 185 MBq and 555 MBq activity was retrospectively compared to a similar cohort of 100 consecutive clinical patients imaged with HAP rest-stress MPI using approximately 370 MBq and 1110 MBq. Administered activity and BMI both had a statistically significant effect on scan time and radiation effective dose. LAP scans took an average of 9 min longer than HAP scans overall, P < 0.0001 and larger BMIs took longer than smaller BMIs, P < 0.0001. In addition, scan times were longer in men than women, P = 0.007. Effective dose was inversely proportional to BMI with an overall decrease of approximately 50% comparing LAP to HAP. For the same CZT camera, the LAP increased scan time while lowering the radiation effective dose when compared to HAP. The increase in scan time increased proportionally to BMI. The effective dose was inversely proportional to BMI. This increase in time did not have a significant impact on our local workflow, but its implications should be considered in the setting of LAP implementation, especially in obese or high patient volume practices.

Spending on Cardiovascular Disease and Cardiovascular Risk Factors in the United States: 1996 to 2016

BACKGROUND

Spending on cardiovascular disease and cardiovascular risk factors (cardiovascular spending) accounts for a significant portion of overall US health care spending. Our objective was to describe US adult cardiovascular spending patterns in 2016, changes from 1996 to 2016, and factors associated with changes over time.

METHODS

We extracted information on adult cardiovascular spending from the Institute for Health Metrics and Evaluation's disease expenditure project, which combines data on insurance claims, emergency department and ambulatory care visits, inpatient and nursing care facility stays, and drug prescriptions to estimate >85% of all US health care spending. Cardiovascular spending (2016 US dollars) was stratified by age, sex, type of care, payer, and cardiovascular cause. Time trend and decomposition analyses quantified contributions of epidemiology, service price and intensity (spending per unit of utilization, eg, spending per inpatient bed-day), and population growth and aging to the increase in cardiovascular spending from 1996 to 2016.

RESULTS

Adult cardiovascular spending increased from $212 billion in 1996 to $320 billion in 2016, a period when the US population increased by >52 million people, and median age increased from 33.2 to 36.9 years. Over this period, public insurance was responsible for the majority of cardiovascular spending (54%), followed by private insurance (37%) and out-of-pocket spending (9%). Health services for ischemic heart disease ($80 billion) and hypertension ($71 billion) led to the most spending in 2016. Increased spending between 1996 and 2016 was primarily driven by treatment of hypertension, hyperlipidemia, and atrial fibrillation/flutter, for which spending rose by $42 billion, $18 billion, and $16 billion, respectively. Increasing service price and intensity alone were associated with a 51%, or $88 billion, cardiovascular spending increase from 1996 to 2016, whereas changes in disease prevalence were associated with a 37%, or $36 billion, spending reduction over the same period, after taking into account population growth and population aging.

CONCLUSIONS

US adult cardiovascular spending increased by >$100 billion from 1996 to 2016. Policies tailored to control service price and intensity and preferentially reimburse higher quality care could help counteract future spending increases caused by population aging and growth.

Differences in High- and Low-Value Cardiovascular Testing by Health Insurance Provider

Background

Quality of care incentives and reimbursements for cardiovascular testing differ between insurance providers. We hypothesized that there are differences in the use of guideline‐concordant testing between Medicaid versus commercial insurance patients <65 years, and between Medicare Advantage versus Medicare fee‐for‐service patients ≥65 years.

Methods and Results

Using data from the Colorado All‐Payer Claims Database from 2015 to 2018, we identified patients eligible to receive a high‐value test recommended by guidelines: assessment of left ventricular function among patients hospitalized with acute myocardial infarction or incident heart failure, or a low‐value test that provides minimal patient benefit: stress testing prior to low‐risk surgery or routine stress testing within 2 years of percutaneous coronary intervention or coronary artery bypass graft surgery. Among 145 616 eligible patients, 37% had fee‐for‐service Medicare, 18% Medicare Advantage, 22% Medicaid, and 23% commercial insurance. Using multilevel logistic regression models adjusted for patient characteristics, Medicaid patients were less likely to receive high‐value testing for acute myocardial infarction (odds ratio [OR], 0.84 [0.73–0.98]; P=0.03) and heart failure (OR, 0.59 [0.51–0.70]; P<0.01) compared with commercially insured patients. Medicare Advantage patients were more likely to receive high‐value testing for acute myocardial infarction (OR, 1.35 [1.15–1.59]; P<0.01) and less likely to receive low‐value testing after percutaneous coronary intervention/ coronary artery bypass graft (OR, 0.63 [0.55–0.72]; P<0.01) compared with Medicare fee‐for‐service patients.

Conclusions

Guideline‐concordant testing was less likely to occur among patients with Medicaid compared with commercial insurance, and more likely to occur among patients with Medicare Advantage compared with fee‐for‐service Medicare. Insurance plan features may provide valuable targets to improve guideline‐concordant testing.

Facility Variation in Troponin Ordering Within the Veterans Health Administration

BACKGROUND

Current United States guidelines recommend troponin as the preferred biomarker in assessing for acute coronary syndrome, but recommendations are limited about which patients to test. Variations in troponin ordering may influence downstream health care utilization.

METHODS

We performed a cross-sectional analysis of 3,308,131 emergency department (ED) visits in all 121 acute care facilities within the Veterans Health Administration from 2015 to 2017. We quantified the degree to which case mix and facility characteristics accounted for variations in facility rates in troponin ordering. We then assessed the association between facility quartiles of risk-adjusted troponin ordering and downstream resource utilization [inpatient admissions, noninvasive testing (stress tests, echocardiograms), and invasive procedures (coronary angiograms, percutaneous coronary interventions, and coronary artery bypass grafting surgeries)].

RESULTS

The proportion of ED visits with troponin orders ranged from 2.2% to 64.5%, with a median of 37.1%. Case mix accounted for 9.5% of the variation in troponin orders; case mix and differences in facility characteristics accounted for 34.6%. Facilities in the highest quartile of troponin ordering, as compared with those in the lowest quartile, had significantly higher rates of inpatient admissions, stress tests, echocardiograms, coronary angiograms, and percutaneous coronary intervention.

CONCLUSIONS

Significant variation in troponin utilization exists across Veterans Health Administration facilities and that variation is not well explained by case mix alone. Facilities with higher rates of troponin ordering were associated with more downstream resource utilization.

Factors Influencing Temporal Trends in Pediatric Inpatient Imaging Utilization

BACKGROUND

Concern exists over exponential growth in cardiac imaging in adults, but there is paucity of such data for cardiac imaging trends in pediatric patients. The aims of this study were to determine temporal trends in the use of noninvasive cardiac imaging and compare these with trends in the use of noncardiac imaging and to identify factors influencing those trends using the Pediatric Health Information Service database.

METHODS

Pediatric inpatient encounter data from January 2004 to December 2017 at 35 pediatric hospitals were extracted from the Pediatric Health Information Service database. Temporal imaging utilization trends in cardiac and noncardiac ultrasound or echocardiography, magnetic resonance imaging (MRI), and computed tomography (CT) were assessed using linear mixed-effects models. Models were adjusted for case-mix index, complex chronic conditions, patient age, length of stay, payer source, and cardiac surgical volume.

RESULTS

A total of 5,869,335 encounters over 14 years were analyzed (median encounters per center per year, 11,411; median patient age, 4 years; median length of stay, 3 days). From 2004 to 2017, the rates of pediatric inpatient cardiac and noncardiac ultrasound and MRI increased, whereas the rate of noncardiac CT decreased. Cardiac CT use increased beginning in 2014 (+0.264 cardiac CT encounters per 1,000 encounters per year), surpassing the rate of rise of cardiac MRI. Case-mix index, cardiac surgical volume, and payer source affected the largest number of imaging trends.

CONCLUSIONS

Among pediatric inpatients, utilization of cardiac and noncardiac ultrasound and MRI has steadily increased. Noncardiac CT use declined and cardiac CT use increased after 2014. Factors influencing imaging trends include case-mix index, cardiac surgical volume, and payer source. This study lays a foundation for investigations of imaging-related resource utilization and outcomes among pediatric inpatients.

Diagnostic and treatment utility of echocardiography in the management of the cardiac patient

BACKGROUND

Echocardiograms are an incredibly useful diagnostic tool due to their lack of harmful radiation, the relative ease and speed with which they can be performed, and their almost ubiquitous availability. Unfortunately, the advantages that support the use of echocardiography can also lead to the overuse of this technology. We sought to evaluate the physician perceived impact echocardiography has on patient management.

AIM

To evaluate the physician perceived impact echocardiography has on patient management.

METHODS

Surveys were distributed to the ordering physician for echocardiograms performed at our institution over a 10-wk period. Only transthoracic echocardiograms performed on the inpatient service were included. Surveys were distributed to either the attending physician or the resident physician listed on the echocardiogram order. The information requested in the survey focused on the indication for the study and the perceived importance and effect of the study. Observational statistical analysis was performed on all of the answers from the collected surveys.

RESULTS

A total of 103 surveys were obtained and analyzed. The internal medicine (57%) and cardiology (37%) specialties ordered the most echocardiograms. The most common reason for ordering an echocardiogram was to rule out a diagnosis (38.2%). Only 27.5% of physicians reported that the echocardiogram significantly affected patient care, with 18.6% reporting a moderate effect, and 30.4% reporting a mild effect. A total of 19.6% of physicians stated that there was no effect on patient management. Additionally, 43.1% of physicians reported that they made changes in patient management due to no change having occurred in the disease, 11.8% reported that changes in management were based on the recommendation of a specialist, and only 9.8% reported that further imaging was ordered due to the results of the echocardiogram. The majority of physicians (67.6%) considered an echocardiogram to be “somewhat essential” in the management of adult inpatients, with only 15.7% considering it “essential”.

CONCLUSION

The majority of physicians surveyed report the echocardiogram had only a mild effect on management with only 27.5% reporting a significant effect. However, the majority of physicians (83.3%) perceived an echocardiogram to be somewhat or entirely essential for management. Only 9.8% reported the echo led to further imaging. These insights into ordering physician reasoning should help guide better definition of the optimal and ideal use of echocardiography.

Validating Publicly Available Crosswalks for Translating ICD-9 to ICD-10 Diagnosis Codes for Cardiovascular Outcomes Research

Background On October 1, 2015, the Center for Medicare and Medicaid Services transitioned from the International Classification of Diseases, Ninth Revision ( ICD-9) to the ICD, Tenth Revision ( ICD-10) compendium of codes for diagnosis and billing in health care, but translation between the two is often inexact. Here we describe a validated crosswalk to translate ICD-9 codes into ICD-10 codes, with a focus on complications after carotid revascularization and endovascular aortic aneurysm repair. Methods and Results We devised an 8-step process to derive and validate ICD-10 codes from existing ICD-9 codes. We used publicly available sources, including the General Equivalence Mapping database, to translate ICD-9 codes used in prior work to ICD-10 codes. We defined ICD-10 codes as validated if they were concordant with the initial ICD-9 codes after manual comparison by two physicians. Our primary validation measure was the percent of valid ICD-10 codes out of the total ICD-10 codes obtained during translation. We began with 126 ICD-9 diagnosis codes used for complication identification after carotid revascularization procedures, and 97 ICD-9 codes for complications after endovascular aortic aneurysm procedures. Translation generated 143 ICD-10 codes for carotid revascularization, a 14% increase from the initial 126 codes. Manual comparison demonstrated 98% concordance, with 99% agreement between the reviewers. Similarly, we identified 108 ICD-10 codes for endovascular aortic aneurysm repair, an 11% increase from the initial 97 ICD-9 codes. We again noted excellent concordance and agreement (98% and 100%, respectively). Manual review identified 4 ICD-10 codes incorrectly translated from ICD-9 codes for carotid revascularization, and 3 codes incorrectly translated for endovascular aortic aneurysm repair. Conclusions Algorithms to crosswalk lists of ICD-9 codes to ICD-10 codes can leverage electronic resources to minimize the burden of code translation. However, manual review for code validation may be necessary, with collaboration across institutions for researchers to share their efforts.

Trends in High- and Low-Value Cardiovascular Diagnostic Testing in Fee-for-Service Medicare, 2000-2016

IMPORTANCE

Owing to a rapid increase in rates of diagnostic cardiovascular testing in the 1990s and early 2000s, the Centers for Medicare & Medicaid Services implemented a series of payment changes intended to reduce overall spending on fee-for-service testing. Whether guideline-concordant testing has been subsequently affected is unknown to date.

OBJECTIVE

To determine whether changes in overall rates of use of diagnostic cardiovascular tests were associated with changes in high-value testing recommended by guidelines and low-value testing that is expected to provide minimal benefits.

DESIGN, SETTING, AND PARTICIPANTS

This retrospective cohort study assessed a national 5% random sample of Medicare fee-for-service beneficiaries aged 65 to 95 years from January 1, 1999, through December 31, 2016. Data were analyzed from February 15, 2018, through August 15, 2019.

EXPOSURES

Eligibility to receive high-value testing (assessment of left ventricular systolic function among patients hospitalized with acute myocardial infarction or heart failure) and low-value testing (stress testing before low-risk noncardiac surgery and routine stress testing within 2 years of coronary revascularization not associated with acute care visits).

MAIN OUTCOMES AND MEASURES

Age- and sex-adjusted annual rates of overall, high-value, and low-value diagnostic cardiovascular testing.

RESULTS

Mean (SD) age was similar over time (75.57 [7.32] years in 2000-2003; 74.82 [7.79] years in 2012-2016); the proportion of women slightly declined over time (63.23% in 2000 to 2003; 57.27% in 2012 to 2016). The rate of overall diagnostic cardiovascular testing per 1000 patient-years among the 5% sample of Medicare beneficiaries increased from 275 in 2000 to 359 in 2008 (P < .001) and then declined to 316 in 2016 (P < .001). High-value testing increased steadily over the entire study period for patients with acute myocardial infarction (85.7% to 89.5%; P < .001) and heart failure (72.6% to 80.1%; P < .001). Low-value testing among patients undergoing low-risk surgery increased from 2.4% in 2000 to 3.8% in 2008 (P < .001) but then declined to 2.5% in 2016 (P < .001). Low-value testing within 2 years of coronary revascularization slightly increased from 47.4% in 2000 to 49.2% in 2003 (P = .03) but then declined to 30.8% in 2014 (P < .001).

CONCLUSIONS AND RELEVANCE

Rates of overall and low-value diagnostic cardiovascular testing appear to have declined considerably and rates of high-value testing have increased slightly. Payment changes intended to reduce spending on overall testing may not have adversely affected testing recommended by guidelines.

Association Between Medicare Expenditure Growth and Mortality Rates in Patients With Acute Myocardial Infarction: A Comparison From 1999 Through 2014

Importance

Many studies have considered the association between Medicare spending and health outcomes at a point in time; few have considered the association between the long-term growth in spending and outcomes.

Objective

To assess whether components of growth in Medicare expenditures are associated with mortality rates between January 1, 1999, and June 30, 2014, for beneficiaries hospitalized for acute myocardial infarction.

Design, Setting, and Participants

Cross-sectional analysis of a random 20% sample of fee-for-service Medicare beneficiaries from January 1, 1999, through December 31, 2000 (n=72 473) and January 1, 2004, through December 31, 2004 (n=38 248), and 100% sample from January 1, 2008, through December 31, 2008 (n=159 558) and January 1, 2013, through June 30, 2014 (n=209 614) admitted with acute myocardial infarction to 1220 hospitals.

Main Outcomes and Measures

Primary exposure measures include the growth of 180-day expenditure components (eg, inpatient, physician, and postacute care) and early percutaneous coronary intervention by hospitals adjusted for price differences and inflation. The primary outcome is the risk-adjusted 180-day case fatality rate.

Results

Patients in each of the years 2004, 2008, and 2013-2014 (relative to those in 1999-2000) were qualitatively of equivalent age, less likely to be white or female, and more likely to be diabetic (all P < .001). Adjusted expenditures per patient increased 13.9% from January 1, 1999, through December 31, 2000, and January 1, 2013, through June 30, 2014, but declined 0.5% between 2008 and 2013-2014. Mean (SD) expenditures in the 5.0% of hospitals (n = 61) with the most rapid expenditure growth between 1999-2000 and 2013-2014 increased by 44.1% ($12 828 [$2315]); for the 5.0% of hospitals with the slowest expenditure growth (n = 61), mean expenditures decreased by 18.7% (-$7384 [$4141]; 95% CI, $8177-$6496). The growth in early percutaneous coronary intervention exhibited a negative association with 180-day case fatality. Spending on cardiac procedures was positively associated with 180-day mortality, while postacute care spending exhibited moderate cost-effectiveness ($455 000 per life saved after 180 days; 95% CI, $323 000-$833 000). Beyond spending on noncardiac procedures, growth in other components of spending was not associated with health improvements.

Conclusions and Relevance

Health improvements for patients with acute myocardial infarction varied across hospitals and were associated with the diffusion of cost-effective care, such as early percutaneous coronary intervention and, to a lesser extent, postacute care, rather than overall expenditure growth. Interventions designed to promote hospital adoption of cost-effective care could improve patient outcomes and, if accompanied by cuts in cost-ineffective care (inside and outside of the hospital setting), also reduce expenditures.

Understanding by General Providers of the Echocardiogram Report

Echocardiograms are the second most frequently utilized cardiac test after electrocardiograms and are most commonly ordered by noncardiology providers. Echocardiogram reports are designed to communicate a comprehensive interpretation of cardiac function; however, it is not known how well these reports are understood by ordering providers. In order to identify gaps in understanding and target potential areas for improvement, we developed a questionnaire testing various topics reported on a standard transthoracic echocardiogram report. This questionnaire was administered to general medicine and cardiology trainees and attending physicians at 2 large academic institutions. Questionnaire response rate was 81%. There were several topics that were not well understood by general providers; these included viability of an akinetic region, pulmonary artery systolic pressure, left ventricular filling pressure, recognition of abnormal structures, and method of identifying of intracardiac thrombus. In conclusion, strategies such as improved communication techniques and adjustment of reporting format should be implemented to increase the clinical value of the echocardiogram.

Predicting Survival From Large Echocardiography and Electronic Health Record Datasets: Optimization With Machine Learning

OBJECTIVES

The goal of this study was to use machine learning to more accurately predict survival after echocardiography.

BACKGROUND

Predicting patient outcomes (e.g., survival) following echocardiography is primarily based on ejection fraction (EF) and comorbidities. However, there may be significant predictive information within additional echocardiography-derived measurements combined with clinical electronic health record data.

METHODS

Mortality was studied in 171,510 unselected patients who underwent 331,317 echocardiograms in a large regional health system. The authors investigated the predictive performance of nonlinear machine learning models compared with that of linear logistic regression models using 3 different inputs: 1) clinical variables, including 90 cardiovascular-relevant International Classification of Diseases, Tenth Revision, codes, and age, sex, height, weight, heart rate, blood pressures, low-density lipoprotein, high-density lipoprotein, and smoking; 2) clinical variables plus physician-reported EF; and 3) clinical variables and EF, plus 57 additional echocardiographic measurements. Missing data were imputed with a multivariate imputation by using a chained equations algorithm (MICE). The authors compared models versus each other and baseline clinical scoring systems by using a mean area under the curve (AUC) over 10 cross-validation folds and across 10 survival durations (6 to 60 months).

RESULTS

Machine learning models achieved significantly higher prediction accuracy (all AUC >0.82) over common clinical risk scores (AUC = 0.61 to 0.79), with the nonlinear random forest models outperforming logistic regression (p < 0.01). The random forest model including all echocardiographic measurements yielded the highest prediction accuracy (p < 0.01 across all models and survival durations). Only 10 variables were needed to achieve 96% of the maximum prediction accuracy, with 6 of these variables being derived from echocardiography. Tricuspid regurgitation velocity was more predictive of survival than LVEF. In a subset of studies with complete data for the top 10 variables, multivariate imputation by chained equations yielded slightly reduced predictive accuracies (difference in AUC of 0.003) compared with the original data.

CONCLUSIONS

Machine learning can fully utilize large combinations of disparate input variables to predict survival after echocardiography with superior accuracy.

Damage Caps and Defensive Medicine: Reexamination with Patient-Level Data

Physicians often claim that they practice "defensive medicine," including ordering extra imaging and laboratory tests, due to fear of malpractice liability. Caps on noneconomic damages are the principal proposed remedy. Do these caps in fact reduce testing, overall health-care spending, or both? We study the effects of "third-wave" damage caps, adopted in the 2000s, on specific areas that are expected to be sensitive to med mal risk: imaging rates, cardiac interventions, and lab and radiology spending, using patient-level data, with extensive fixed effects and patient-level covariates. We find heterogeneous effects. Rates for the principal imaging tests rise, as does Medicare Part B spending on laboratory and radiology tests. In contrast, cardiac intervention rates (left-heart catheterization, stenting, and bypass surgery) do not rise (and likely fall). We find some evidence that overall Medicare Part B rises, but variable results for Part A spending. We find no evidence that caps affect mortality.

Evidence Basis for a Point-of-Care Ultrasound Examination to Refine Referral for Outpatient Echocardiography

BACKGROUND

Few data exist on the potential utility of a cardiac point-of-care ultrasound (POCUS) examination in the outpatient setting to assist diagnosis of significant cardiac disease. Using a retrospective sequential cohort design, we sought to derive and then validate a POCUS examination for cardiac application and model its potential use for prognostication and cost-effective echo referral.

METHODS

For POCUS examination derivation, we reviewed 233 consecutive outpatient echo studies for 4 specific POCUS "signs" contained therein representing left ventricular systolic dysfunction, left atrial enlargement, inferior vena cava plethora, and lung apical B-lines. The corresponding formal echo reports were then queried for any significant abnormality. The optimal POCUS examination for identifying an abnormal echo was determined. We then reviewed 244 consecutive outpatient echo studies from another institution for associations between the optimal POCUS examination, clinical variables, and referral source with major adverse cardiac events and all-cause mortality in univariate and multivariate models. Assuming a referral model where the absence of POCUS signs or variables would negate initial echo referral, theoretical cost savings were expressed as a percentage in reduction of echo studies.

RESULTS

In the derivation cohort, the combination of two signs, denoting left atrial enlargement and inferior vena cava plethora resulted in the highest accuracy of 72% [95% CI: 65%, 78%] in detecting an abnormal echocardiogram. In the validation cohort, mortality at 5.5 years was 14.6% overall, 23% in patients with the left atrial enlargement sign (OR 3.5 [1.6, 7.6]), 25% with inferior vena cava plethora sign (OR 2.2 [0.8, 6.0]), and 8.0% (OR 0.3 [0.2, 0.7]) in those lacking both signs. After adjusting for age, both diabetes (OR 4.8 [2.0, 11.6]), and the left atrial enlargement sign (OR 2.4 [1.1, 5.4]) remained independently associated with mortality (p<0.05). In the referral model, patients younger than 65 years of age without diabetes and without the left atrial enlargement sign would not have received echo referral, resulting in a 33% reduction in total echo cost and would have constituted a low-risk group with a 1.2% 5.5-year mortality.

CONCLUSIONS

A quick-look sign for left atrial enlargement is associated with 5-year mortality and could function as an easily obtained outpatient POCUS examination to help in identifying patients in need of echo referral.

Fully Automated Echocardiogram Interpretation in Clinical Practice

BACKGROUND

Automated cardiac image interpretation has the potential to transform clinical practice in multiple ways, including enabling serial assessment of cardiac function by nonexperts in primary care and rural settings. We hypothesized that advances in computer vision could enable building a fully automated, scalable analysis pipeline for echocardiogram interpretation, including (1) view identification, (2) image segmentation, (3) quantification of structure and function, and (4) disease detection.

METHODS

Using 14 035 echocardiograms spanning a 10-year period, we trained and evaluated convolutional neural network models for multiple tasks, including automated identification of 23 viewpoints and segmentation of cardiac chambers across 5 common views. The segmentation output was used to quantify chamber volumes and left ventricular mass, determine ejection fraction, and facilitate automated determination of longitudinal strain through speckle tracking. Results were evaluated through comparison to manual segmentation and measurements from 8666 echocardiograms obtained during the routine clinical workflow. Finally, we developed models to detect 3 diseases: hypertrophic cardiomyopathy, cardiac amyloid, and pulmonary arterial hypertension.

RESULTS

Convolutional neural networks accurately identified views (eg, 96% for parasternal long axis), including flagging partially obscured cardiac chambers, and enabled the segmentation of individual cardiac chambers. The resulting cardiac structure measurements agreed with study report values (eg, median absolute deviations of 15% to 17% of observed values for left ventricular mass, left ventricular diastolic volume, and left atrial volume). In terms of function, we computed automated ejection fraction and longitudinal strain measurements (within 2 cohorts), which agreed with commercial software-derived values (for ejection fraction, median absolute deviation=9.7% of observed, N=6407 studies; for strain, median absolute deviation=7.5%, n=419, and 9.0%, n=110) and demonstrated applicability to serial monitoring of patients with breast cancer for trastuzumab cardiotoxicity. Overall, we found automated measurements to be comparable or superior to manual measurements across 11 internal consistency metrics (eg, the correlation of left atrial and ventricular volumes). Finally, we trained convolutional neural networks to detect hypertrophic cardiomyopathy, cardiac amyloidosis, and pulmonary arterial hypertension with C statistics of 0.93, 0.87, and 0.85, respectively.

CONCLUSIONS

Our pipeline lays the groundwork for using automated interpretation to support serial patient tracking and scalable analysis of millions of echocardiograms archived within healthcare systems.

Trends in the use of echocardiography in pulmonary embolism

Pulmonary embolism (PE) is a devastating diagnosis which carries a high mortality risk. Echocardiography is often performed to risk stratify patients diagnosed with PE, and guide management strategies. Trends in the performance of echocardiography among patients with PE and its role in influencing outcomes is unknown.We analyzed the 2005 to 2014 National Inpatient Sample Database to identify patients with primary diagnosis of PE or secondary diagnosis of PE and ≥1 of the following diagnoses: syncope, thrombolysis, acute deep vein thrombosis, acute cardiorespiratory failure, and secondary pulmonary hypertension. Trends in the performance of echocardiography and in-hospital mortality were analyzed. The admissions were divided into 2 groups with echocardiography, and without echocardiography, and 1:2 propensity score matching (PSM) was performed for comparison. The primary end-point was in-hospital mortality. The secondary endpoints were length of stay and total hospitalization costs. Odd ratios (OR) with confidence intervals (CI) were reported.A total of 299,536 unweighted PE cases were studied. Performance of echocardiography among patients with PE patients increased from 3.5% to 5.6%, whereas in-hospital mortality decreased from 4.2% to 3.7% between years 2005 and 2014. Before matching, patients who received an echocardiogram were more likely to be younger, African American, admitted to a large, urban teaching institute, and had higher rates of concurrent acute deep vein thrombosis, and acute respiratory failure. Post-PSM, patients who received echocardiography during hospitalization had lower in-hospital mortality (odds ratio 0.75, 95% confidence intervals (CI) 0.68-0.83; P < 0.001), longer length of stay (median 6 days vs 5 days; P < .001) and higher mean hospitalization costs ($34,379 vs $27,803; P < .001) compared to those without echocardiography.Performance of echocardiography among patients with a PE is increasing and is associated with lower in-hospital mortality.

A systematic review of Appropriate Use Criteria for transthoracic echocardiography: are they relevant outside the United States?

INTRODUCTION

The 2011 Appropriate Use Criteria (AUC) were developed by the American Society of Echocardiography (ASE) to provide guidance for referring physicians in response to growing concerns about unnecessary transthoracic echocardiogram (TTE) requests. Very few studies have assessed how medical centers overseas perform against AUC. Evidence is now emerging that inappropriate referral rates in Europe are similar to those reported in the US.

OBJECTIVE

This study systematically reviewed published evidence to identify (1) whether the 2011 AUC are applicable to medical centers outside the US (2) the level of adherence to the AUC across multiple centers, (3) the main factors which cause deviation from AUC, (4) any changes in referral rates since the publication of AUC, and (5) any factors and/or intervention strategies which promote adherence to AUC.

METHODS AND RESULTS

Electronic databases were systematically searched for papers related to AUC and cardiac imaging. Following screening and application of eligibility criteria, data was extracted from ten reports involving 8561 TTE studies. Classification rates were 99.5 and 98% for US studies and studies outside the US respectively. Overall, 7119 TTE studies were classified as appropriate (83.1%) of which 3724 were US referrals (84.7%) and 3395 originated outside the US (81.5%). Six of the included studies independently observed significantly more appropriate referrals among inpatients compared to outpatients (p < 0.001). US centers observed no significant difference in appropriate referral rates between physician specialties while one UK study showed cardiac surgeons ordered inappropriate TTEs more frequently than other specialties (p < 0.05). This review found no obvious trend in appropriate referral rates between 2012 and 2015 indicating no temporal change in physician ordering patterns. Only one educational interventional study met the author's criteria which showed that while intervention was effective during its implantation (26% reduction in TTEs ordered per day), TTE referral rates regressed to pre-intervention levels overtime.

CONCLUSIONS

In conclusion, the American guideline AUC are applicable to centers outside the US and their implementation across five international centers suggest almost 1 in 5 scans could be avoided.

Comparative Utilization and Temporal Trends in Cardiac Stress Testing in U.S. Medicare Beneficiaries With and Without Chronic Kidney Disease

OBJECTIVES

The authors aimed to analyze temporal trends in cardiac stress testing in U.S. Medicare beneficiaries from 2008 to 2012, types of stress testing, and comparative utilization related to the presence and severity of chronic kidney disease (CKD).

BACKGROUND

A long-held perception depicts patients with CKD as being treated less intensively for cardiovascular disease than nonrenal patients. We wondered whether use of diagnostic testing for ischemic heart disease is affected by the presence of CKD.

METHODS

Using the 20% Medicare sample, we assembled yearly cohorts of Medicare beneficiaries (∼4,500,000 per year) from 2008 to 2012. Beneficiaries 66 years or older undergoing a first cardiac stress test, with no previous history of coronary revascularization and no acute coronary syndrome within 60 days, were identified, as was the type of stress test. We analyzed temporal trends and compared testing rates related to CKD stage versus no CKD. A Poisson regression model estimated the likelihood of stress testing in 2012 by CKD stage, adjusted for demographic characteristics and comorbid conditions.

RESULTS

Approximately 480,000 older patients (∼29,000 with CKD) underwent stress tests in 2008, progressively declining to ∼400,000 in 2012 (∼38,000 with CKD). In 2008 to 2012, 78% to 80% of all stress testing in non-CKD patients used nuclear imaging, as did 87% to 88% in CKD patients. Rates of stress testing declined progressively for non-CKD and CKD patients in 2008 to 2012: 11.5 to 9.4 per 100 patient-years and 16.8 to 13.4 per 100 patient-years, respectively. The adjusted Poisson model, with non-CKD as the reference, showed an increasing likelihood of stress testing with worsening CKD: incidence rate ratio 1.01 for stages 1 to 2 (p = NS), 1.05 for stage 3 (p < 0.0001), 1.01 for stage 4 (p = NS), 1.04 for stage 5 nondialysis (p = NS), and 1.15 for stage 5 dialysis (p < 0.0001).

CONCLUSIONS

Overall rates of cardiac stress testing (over three-fourths using nuclear imaging) declined in 2008 to 2012 among Medicare beneficiaries 66 years or older but were consistently higher for CKD than for non-CKD patients. The effect of screening algorithms for transplant candidates was unknown. Our data refute underutilization of cardiac stress testing in CKD patients.

Contrast-Enhanced Stress Echocardiography and Myocardial Perfusion Imaging in Patients Hospitalized With Chest Pain: A Randomized Study

Ultrasound contrast-enhanced stress echocardiography improves endocardial visualization, but diagnostic test rates versus stress myocardial perfusion imaging (MPI) have not been studied. A prospective randomized trial was performed between April 2012 and October 2014 at a single-center, safety net hospital. Hospitalized patients referred for noninvasive stress imaging were randomized 1:1 to stress echocardiography or stress MPI. The primary outcome was diagnostic test rate defined as interpretable images and achievement of >85% of age-predicted maximal heart rate (for dobutamine and exercise). Rates were assessed among those completing testing and then based solely on image interpretability. Charges and length of stay were secondary outcomes. A total of 240 patients were randomized, and 229 completed testing. Diagnostic test rates were similar for stress echocardiography versus MPI {89.4% [95% confidence interval (CI), 82.2-94.4] vs. 94.8% [95% CI, 89.1-98.1], P = 0.13} and did not differ with multivariable adjustment. Modalities requiring a diagnostic heart rate criteria were more frequently ordered with stress echocardiography (100% vs. 26%; P < 0.001). Therefore, an imaging-based analysis without the 12 individuals who failed to achieve target heart rate (n = 217) was evaluated with diagnostic test rates of 100% versus 94.8% (95% CI, 89.1%-98.1%; P = 0.03) for stress echocardiography and MPI, respectively. Median length of stay did not differ. Median (interquartile range) test-related charges were lower with stress echocardiography: $2,424 ($2400-$2508) versus $3619 ($3584-$3728), P < 0.0001. Overall, tests were positive for ischemia in 8% of patients. In conclusion, contrast-enhanced stress echocardiography provides comparable diagnostic test rates to MPI with lower associated charges.

Left ventricular end-diastolic volume predicts exercise capacity in patients with a normal ejection fraction

BACKGROUND

Exercise capacity is a powerful predictor of all-cause mortality. The duration of exercise with treadmill stress testing is an important prognostic marker in both healthy subjects and patients with cardiovascular disease. Left ventricular (LV) structure is known to adapt to sustained changes in level of physical activity.

HYPOTHESIS

Poor exercise capacity in patients with a preserved LV ejection fraction (LVEF) should be reflected in smaller LV dimensions, and a normal exercise capacity should be associated with larger LV dimensions, irrespective of comorbidities.

METHODS

This hypothesis was first tested in a cross-sectional analysis of 201 patients with normal chamber dimensions and preserved LVEF who underwent a clinically indicated treadmill stress echocardiogram using the Bruce protocol (derivation cohort). The best LV dimensional predictor of exercise capacity was then tested in 1285 patients who had a Bruce-protocol treadmill exercise stress test and a separate transthoracic echocardiogram (validation cohort).

RESULTS

In the derivation cohort, there was a strong positive relationship between exercise duration and LV end-diastolic volume deciles (r ²  = 0.85; P < 0.001). Regression analyses of several LV dimensional parameters revealed that the body surface area-based LV end-diastolic volume index was best suited to predict exercise capacity (P < 0.0001). In a large validation cohort, LV end-diastolic volume was confirmed to predict exercise capacity (P < 0.0001).

CONCLUSIONS

Among patients referred for outpatient stress echocardiography who have a preserved LVEF and no evidence of myocardial ischemia, we found a strong positive association between LV volume and exercise capacity.

Clinical Outcomes After Cardiac Stress Testing Among US Patients Younger Than 65 Years

Background

Scientific statements have championed the measurement of clinical outcomes after cardiac stress testing to better define their value. Using contemporary national data, we sought to describe the characteristics of patients who experience outcomes after stress testing.

Methods and Results

Using administrative claims from a large national private insurer, we conducted an observational cohort study of patients without cardiovascular disease aged 25 to 64 years who underwent stress testing from 2006 to 2011 and had at least 1 year of membership in the insurance company before and after testing. We used Kaplan–Meier time‐to‐event analyses to determine rates of acute myocardial infarction (AMI), elective coronary revascularization, and coronary angiography without revascularization in the year following testing. We used logistic regression to determine factors associated with outcomes, and stratified the cohort into quintiles based on likelihood of experiencing AMI and/or revascularization to describe the characteristics of patients at highest and lowest risk. Among 553 027 patients who underwent stress testing (mean age 50 years, 49% women, 73% white), 0.8% were hospitalized for AMI, 1.8% underwent elective coronary revascularization, and 2.5% underwent coronary angiography without revascularization within 1 year. Patients who were older, male, and white were more likely to undergo subsequent revascularization. Patients in the lowest likelihood quintile were young (mean age 40 years), frequently women (84.7%), had a low incidence of coexisting conditions (5.2% with diabetes mellitus), and had a 0.5% rate of AMI and/or revascularization.

Conclusions

The proportion of US patients younger than 65 who had AMI and/or coronary revascularization after stress testing was low. Assessing risk of subsequent outcomes may be useful in improving patient referrals for stress testing.

Derivation of a Clinical Model to Predict Unchanged Inpatient Echocardiograms

BACKGROUND

Transthoracic echocardiography (TTE) is one of the most commonly ordered tests in healthcare. Repeat TTE, defined as a TTE done within 1 year of a prior TTE, represents 24% to 42% of all studies. The purpose of this study was to derive a clinical prediction model to predict unchanged repeat TTE, with the goal of defining a subset of studies that are unnecessary.

METHODS

Single-center retrospective cohort study of all hospitalized patients who had a repeat TTE between October 1, 2013, and September 30, 2014.

RESULTS

Two hundred eleven of 601 TTEs were repeat studies, of which 78 (37%) had major changes. Five variables were independent predictors of major new TTE changes, including history of intervening acute myocardial infarction, cardiothoracic surgery, major new electrocardiogram (ECG) changes, prior valve disease, and chronic kidney disease. Using the β-coefficient for each of these variables, we defined a clinical prediction model that we named the CAVES score. The acronym CAVES stands for chronic kidney disease, acute myocardial infarction, valvular disease, ECG changes, and surgery (cardiac). The prevalence of major TTE change for the full cohort was 35%. For the group with a CAVES score of -1, that probability was only 5.6%; for the group with a score of 0, the probability was 17.7%; and for the group with a score ≥1, the probability was 55.3%. The bootstrap corrected C statistic for the model was 0.78 (95% confidence interval, 0.72-0.85), indicating good discrimination.

CONCLUSIONS

Overall, the CAVES score had good discrimination and calibration. If further validated, it may be useful to predict repeat TTEs that are unlikely to have major changes.

Comparison of Rate of Utilization of Medicare Services in Private Versus Academic Cardiology Practice

Cardiovascular services are the third largest source of Medicare spending. We examined the rate of cardiovascular service utilization in the community of Glendale, CA, compared with the nearest academic medical center, the University of Southern California. Publicly available utilization data released by Medicare for the years 2012 and 2013 were used to identify all inpatient and outpatient cardiology services provided in each practice setting. The analysis included 19 private and 17 academic cardiologists. In unadjusted analysis, academic physicians performed half as many services per Medicare beneficiary per year as those in private practice: 2.3 versus 4.8, p <0.001. Other factors associated with higher utilization included male physician, international (vs US) medical school graduate, interventional (vs general) cardiologist, and more years in practice. Factors independently associated with higher utilization rates by multivariable analysis included private practice setting (odds ratio [OR] 1.84, 95% confidence interval [CI] 1.30 to 2.61, p <0.001), male physician (OR 1.64, 95% CI 1.00 to 2.67, p = 0.049), and international medical school graduate (OR 1.37, 95% CI 1.07 to 1.78, p = 0.014). In conclusion, in this analysis of 2 cardiology practice settings in southern California, medical service utilization per Medicare beneficiary was nearly 2-fold higher in private practice than in the academic setting, suggesting that there may be opportunity for substantially reducing costs of cardiology care in the community setting.

Variation in the Echocardiographic Surveillance of Primary Mitral Regurgitation

BACKGROUND

Clinical outcomes after surgical treatment of mitral regurgitation are worse if intervention occurs after deterioration of left ventricular size and function. Transthoracic echocardiographic (TTE) surveillance of patients with mitral regurgitation is indicated to avoid adverse ventricular remodeling. Overly frequent TTEs can impair patient access and reduce value in care delivery. This balance between timely surveillance and overutilization of TTE in valvular disease provides a model to study variation in the delivery of healthcare services. We investigated patient and provider factors contributing to variation in TTE utilization and hypothesized that variation was attributable to provider practice even after adjustment for patient characteristics.

METHODS AND RESULTS

We obtained records of all TTEs from 2001 to 2016 completed at a large echocardiography laboratory. The outcome variable was time interval between TTEs. We constructed a mixed-effects linear regression model with the individual physician as the random effect in the model and used intraclass correlation coefficient to assess the proportion of outcome variation because of provider practice. Our study cohort was 55 773 TTEs corresponding to 37 843 intervals ordered by 635 providers. The mean interval between TTEs was 12.4 months, 17.0 months, 18.3 months, and 17.4 months for severe, moderate, mild, and trace mitral regurgitation, respectively, with 20% of providers deemed overutilizers of TTEs and 25% underutilizers.

CONCLUSIONS

We conclude that there is substantial variation in follow-up intervals for TTE assessment of mitral regurgitation, despite risk-adjustment for patient variables, likely because of provider factors.

Cardiovascular Event Prediction by Machine Learning: The Multi-Ethnic Study of Atherosclerosis

RATIONALE

Machine learning may be useful to characterize cardiovascular risk, predict outcomes, and identify biomarkers in population studies.

OBJECTIVE

To test the ability of random survival forests, a machine learning technique, to predict 6 cardiovascular outcomes in comparison to standard cardiovascular risk scores.

METHODS AND RESULTS

We included participants from the MESA (Multi-Ethnic Study of Atherosclerosis). Baseline measurements were used to predict cardiovascular outcomes over 12 years of follow-up. MESA was designed to study progression of subclinical disease to cardiovascular events where participants were initially free of cardiovascular disease. All 6814 participants from MESA, aged 45 to 84 years, from 4 ethnicities, and 6 centers across the United States were included. Seven-hundred thirty-five variables from imaging and noninvasive tests, questionnaires, and biomarker panels were obtained. We used the random survival forests technique to identify the top-20 predictors of each outcome. Imaging, electrocardiography, and serum biomarkers featured heavily on the top-20 lists as opposed to traditional cardiovascular risk factors. Age was the most important predictor for all-cause mortality. Fasting glucose levels and carotid ultrasonography measures were important predictors of stroke. Coronary Artery Calcium score was the most important predictor of coronary heart disease and all atherosclerotic cardiovascular disease combined outcomes. Left ventricular structure and function and cardiac troponin-T were among the top predictors for incident heart failure. Creatinine, age, and ankle-brachial index were among the top predictors of atrial fibrillation. TNF-α (tissue necrosis factor-α) and IL (interleukin)-2 soluble receptors and NT-proBNP (N-Terminal Pro-B-Type Natriuretic Peptide) levels were important across all outcomes. The random survival forests technique performed better than established risk scores with increased prediction accuracy (decreased Brier score by 10%-25%).

CONCLUSIONS

Machine learning in conjunction with deep phenotyping improves prediction accuracy in cardiovascular event prediction in an initially asymptomatic population. These methods may lead to greater insights on subclinical disease markers without apriori assumptions of causality.

CLINICAL TRIAL REGISTRATION

URL: http://www.clinicaltrials.gov. Unique identifier: NCT00005487.

Temporal Trends of Single-Photon Emission Computed Tomography Myocardial Perfusion Imaging in Patients With Coronary Artery Disease

Background—

There has been a gradual decline in the prevalence of abnormal stress single-photon emission computed tomography (SPECT) myocardial perfusion imaging studies among patients without history of coronary artery disease (CAD). The trends of SPECT studies among patients with known CAD have not been evaluated previously.

Methods and Results—

We assessed the Mayo Clinic nuclear cardiology database for all stress SPECT tests performed between January 1991 and December 2012 in patients with history of CAD defined as having previous myocardial infarction, percutaneous coronary intervention, and coronary artery bypass grafting. The study cohort was divided into 5 time periods: 1991 to 1995, 1996 to 2000, 2001 to 2005, 2006 to 2010, and 2011 to 2012. There were 19 373 patients with a history of CAD who underwent SPECT between 1991 and 2012 (mean age, 66.2±10.9 years; 75.4% men). Annual utilization of SPECT in these patients increased from an average of 495 tests per year in 1991 to 1995 to 1425 in 2003 and then decreased to 552 tests in 2012 without evidence for substitution with other stress modalities. Asymptomatic patients initially increased until 2006 and then decreased. Patients with typical angina decreased, whereas patients with dyspnea and atypical angina increased. High-risk SPECT tests significantly decreased, and the percentage of low-risk SPECT tests increased despite decreased SPECT utilization between 2003 and 2012. Almost 80% of all tests performed in 2012 had a low-risk summed stress score compared with 29% in 1991 ( P <0.001).

Conclusions—

In Mayo Clinic, Rochester, annual SPECT utilization in patients with previous CAD increased between 1992 and 2003, but then decreased after 2003. High-risk SPECT tests declined, whereas low-risk tests increased markedly. Our results suggest that among patients with a history of CAD, SPECT was being increasingly utilized in patients with milder CAD. This trend parallels reduced utilization of other stress modalities, coronary angiography, reduced smoking, and greater utilization of optimal medical therapy for prevention and treatment of CAD.

Cardiac Stress Test Trends Among US Patients Younger Than 65 Years, 2005-2012

Importance

After a period of rapid growth, use of cardiac stress testing has recently decreased among Medicare beneficiaries and in a large integrated health system. However, it is not known whether declines in cardiac stress testing are universal or are confined to certain populations.

Objective

To determine trends in rates of cardiac stress testing among a large and diverse cohort of commercially insured patients.

Design, Setting, and Participants

A serial cross-sectional study with time trends was conducted using administrative claims from all members aged 25 to 64 years belonging to a large, national managed care company from January 1, 2005, to December 31, 2012. Linear trends in rates were determined using negative binomial regression models with procedure count as the dependent variable, calendar quarter as the key independent variable, and the size of the population as a logged offset term. Data analysis was performed from January 1, 2005, to December 31, 2012.

Main Outcomes and Measures

Age- and sex-adjusted rates of cardiac stress tests per calendar quarter (reported as number of tests per 100 000 person-years).

Results

A total of 2 085 591 cardiac stress tests were performed among 32 921 838 persons (mean [SD] age, 43.2 [10.9] years; 16 625 528 women [50.5%] and 16 296 310 [49.5%] men; 7 604 945 nonwhite [23.1%]). There was a 3.0% increase in rates of cardiac stress testing from 2005 (3486 tests; 95% CI, 3458-3514) to 2012 (3589 tests; 95% CI, 3559-3619; P = .01 for linear trend). Use of nuclear single-photon emission computed tomography decreased by 14.9% from 2005 (1907 tests; 95% CI, 1888-1926) to 2012 (1623 tests; 95% CI, 1603-1643; P = .03). Use of stress echocardiography increased by 27.8% from 2005 (709 tests; 95% CI, 697-721) to 2012 (906 tests; 95% CI, 894 to 920; P < .001). Use of exercise electrocardiography increased by 12.5% from 2005 (861 tests; 95% CI, 847-873) to 2012 (969 tests; 95% CI, 953-985; P < .001). Use of other stress testing modalities increased 65.5% from 2006 (55 tests; 95% CI, 51-59) to 2012 (91 tests; 95% CI, 87-95; P < .001). For individuals aged 25 to 34 years, rates of cardiac stress testing increased 59.1% from 2005 (543 tests; 95% CI, 532-554) to 2012 (864 tests; 95% CI, 852-876; P < .001). For individuals aged 55 to 64 years, rates of cardiac stress testing decreased by 12.3% from 2005 (7894 tests; 95% CI, 7820-7968) to 2012 (6923 tests; 95% CI, 6853-6993; P < .001).

Conclusions and Relevance

In contrast to declines in the use of cardiac stress testing in some health care systems, we observed a small increase in its use among a nationally representative cohort of commercially insured patients. Our findings suggest that observed trends in the use of cardiac stress testing may have been driven more by unique characteristics of populations and health systems than national efforts to reduce the overuse of testing.

Value of Exercise ECG for Risk Stratification in Suspected or Known CAD in the Era of Advanced Imaging Technologies

Exercise stress electrocardiography (ExECG) is underutilized as the initial test modality in patients with interpretable electrocardiograms who are able to exercise. Although stress myocardial imaging techniques provide valuable diagnostic and prognostic information, variables derived from ExECG can yield substantial data for risk stratification, either supplementary to imaging variables or without concurrent imaging. In addition to exercise-induced ischemic ST-segment depression, such markers as ST-segment elevation in lead aVR, abnormal heart rate recovery post-exercise, failure to achieve target heart rate, and poor exercise capacity improve risk stratification of ExECG. For example, patients achieving ≥10 metabolic equivalents on ExECG have a very low prevalence of inducible ischemia and an excellent prognosis. In contrast, cardiac imaging techniques add diagnostic and prognostic value in higher-risk populations (e.g., poor functional capacity, diabetes, or chronic kidney disease). Optimal test selection for symptomatic patients with suspected coronary artery disease requires a patient-centered approach factoring in the risk/benefit ratio and cost-effectiveness.

Association of Liability Concerns with Decisions to Order Echocardiography and Cardiac Stress Tests with Imaging

BACKGROUND

Professional societies have made efforts to curb overuse of cardiac imaging and decrease practice variation by publishing appropriate use criteria. However, little is known about the impact of physician-level determinants such as liability concerns and risk aversion on decisions to order testing.

METHODS

A web-based survey was administered to cardiologists and general practice physicians affiliated with two academic institutions. The survey consisted of four clinical scenarios in which appropriate use criteria rated echocardiography or stress testing as "may be appropriate." Respondents' degree of liability concerns and risk aversion were measured using validated tools. The primary outcome variable was tendency to order imaging, calculated as the average likelihood to order an imaging test across the clinical scenarios (1 = very unlikely, 6 = very likely). Linear regression models were used to evaluate the association between tendency to order imaging and physician characteristics.

RESULTS

From 420 physicians invited to participate, 108 complete responses were obtained (26% response rate, 54% cardiologists). There was no difference in tendency to order imaging between cardiologists and general practice physicians (3.46 [95% CI, 3.12-3.81] vs 3.15 [95% CI, 2.79-3.51], P = .22). On multivariate analysis, a higher degree of liability concerns was the only significant predictor of decisions to order imaging (mean difference in tendency to order imaging, 0.36; 95% CI, 0.09-0.62; P = .01).

CONCLUSION

In clinical situations in which performance of cardiac imaging is rated as "may be appropriate" by appropriate use criteria, physicians with higher liability concerns ordered significantly more testing than physicians with lower concerns.

Non-invasive assessment of low- and intermediate-risk patients with chest pain

Coronary artery disease (CAD) remains a significant global public health burden despite advancements in prevention and therapeutic strategies. Common non-invasive imaging modalities, anatomic and functional, are available for the assessment of patients with stable chest pain. Exercise electrocardiography is a long-standing method for evaluation for CAD and remains the initial test for the majority of patients who can exercise adequately with a baseline interpretable electrocardiogram. The addition of cardiac imaging to exercise testing provides incremental benefit for accurate diagnosis for CAD and is particularly useful in patients who are unable to exercise adequately and/or have uninterpretable electrocardiograms. Radionuclide myocardial perfusion imaging and echocardiography with exercise or pharmacological stress provide high sensitivity and specificity in the detection and further risk stratification of patients with CAD. Recently, coronary computed tomography angiography has demonstrated its growing role to rule out significant CAD given its high negative predictive value. Although less available, stress cardiac magnetic resonance provides a comprehensive assessment of cardiac structure and function and provides a high diagnostic accuracy in the detection of CAD. The utilization of non-invasive testing is complex due to various advantages and limitations, particularly in the assessment of low- and intermediate-risk patients with chest pain, where no single study is suitable for all patients. This review will describe currently available non-invasive modalities, along with current evidence-based guidelines and appropriate use criteria in the assessment of low- and intermediate-risk patients with suspected, stable CAD.

Utility of Echocardiogram in the Evaluation of Heart Murmurs

Valvular heart disease is a common condition in today's patient population. Accurate characterization of vital cardiac structures has become crucial to early diagnosis and varied treatment options. The advent of ultrasound technology has had a large impact in cardiovascular medicine, particularly in the assessment of valvular heart disease. Today its versatility and availability have allowed it to become one of the most frequently ordered imaging tests for cardiovascular indications. Despite the tremendous evidence that suggests that clinical examinations are still standard of care, a large volume of referrals for echocardiograms suggests differently.

Assessment of inpatient multimodal cardiac imaging appropriateness at large academic medical centers

Background

Responding to concerns regarding the growth of cardiac testing, the American College of Cardiology Foundation (ACCF) published Appropriate Use Criteria (AUC) for various cardiac imaging modalities. Single modality cardiac imaging appropriateness has been reported but there have been no studies assessing the appropriateness of multiple imaging modalities in an inpatient environment.

Methods

A retrospective study of the appropriateness of cardiac tests ordered by the inpatient General Internal Medicine (GIM) and Cardiology services at three Canadian academic hospitals was conducted over two one-month periods. Cardiac tests characterized were transthoracic echocardiography (TTE), transesophageal echocardiography (TEE), single-photon emission tomography myocardial perfusion imaging (SPECT), and diagnostic cardiac catheterization.

Results

Overall, 553 tests were assessed, of which 99.8 % were classifiable by AUC. 91 % of all studies were categorized as appropriate, 4 % may be appropriate and 5 % were rarely appropriate. There were high rates of appropriate use of all modalities by GIM and Cardiology throughout. Significantly more appropriate diagnostic catheterizations were ordered by Cardiology than GIM (93 % vs. 82 %, p = <0.01). Cardiology ordered more appropriate studies overall (94 % vs. 88 %, p = 0.03) but there was no difference in the rate of rarely appropriate studies (3 % vs. 6 %, p = 0.23).

Conclusion

The ACCF AUC captured the vast majority of clinical scenarios for multiple cardiac imaging modalities in this multi-centered study on Cardiology and GIM inpatients in the acute care setting. The rate of appropriate ordering was high across all imaging modalities. We recommend further work towards improving appropriate utilization of cardiac imaging resources focus on the out-patient setting.

Variation in use of echocardiography among veterans who use the Veterans Health Administration vs Medicare

BACKGROUND

Rapid growth in the provision of cardiac imaging tests has led to concerns about overuse. Little is known about the degree to which health care delivery system characteristics influence use and variation in echocardiography.

METHODS

We analyzed administrative claims of veterans with heart failure older than 65 years from 2007 to 2010 across 34 metropolitan service areas (MSAs). We compared overall rates and geographic variation in use of transthoracic echocardiography (TTE) between veterans who used the Veterans Health Administration (VA) and propensity-matched veterans who used Medicare. "Dual users" were excluded.

RESULTS

There were no significant differences in clinical characteristics or mortality between the propensity-matched cohorts (overall n = 30,404 veterans, mean age 76 years, mortality rate 52%). The Medicare cohort had a significantly higher overall rate of TTE use compared with the VA cohort (1.25 vs 0.38 TTEs per person-year, incidence rate ratio 2.89 [95% CI 2.80-3.00], both P < .001), but a similar coefficient of variation across MSAs (0.36 [95% CI 0.27-0.45] vs 0.48 [95% CI 0.37-0.59]). There was a moderate to strong correlation in variation at the MSA level between cohorts (Spearman r = 0.58, P < .001).

CONCLUSION

Overall rates of TTE use were significantly higher in a Medicare cohort compared with a propensity score-matched VA cohort of veterans with heart failure living in urban areas, with similar relative degrees of geographic variation and moderate to strong regional correlation. Rates of TTE use may be strongly influenced by health care system characteristics, but local practice styles influence echocardiography rates irrespective of health system.