Association Between Offering Limited Left Ventricular Ejection Fraction Echocardiograms and Overall Use of Echocardiography

Medicare spends more than $900 million annually on echocardiography.¹ Despite the development of appropriate use criteria for echocardiography to limit inappropriate testing, at least 5% to 20% of studies may remain inappropriate.^2-5 One overlooked mechanism of reducing costs is to improve test efficiency. For simple indications, a brief echocardiogram with minimal views and measurements may answer the diagnostic question.

In 2006, the Veterans Affairs Palo Alto Health Care System (PAVA) added a new limited transthoracic echocardiogram (TTE) option that evaluated only the left ventricle (LV-TTE). We evaluated the use of this new option and its association with overall use of TTE.

The new electronic medical order option included the estimated costs for a full TTE ($240) and for LV-TTE ($60). No additional barriers or incentives for either form of TTE were introduced. For LV-TTE, the sonographer obtained 5 views of the left ventricle without color or Doppler imaging. If a major abnormality was noted, further imaging was performed at the discretion of the sonographer. The study was approved by the Stanford Human Subjects Institutional Review Board, which granted a waiver of consent for this retrospective data analysis.

The number of annual TTEs and outpatient visits and the mean outpatient age were extracted from facility-specific administrative data between January 1, 2005, and December 31, 2015, for the PAVA and from 118 additional Veterans Affairs hospitals that performed more than 100 TTEs in 2005; these 118 Veterans Affairs hospitals comprised the control group. Transthoracic echocardiograms were classified as full or limited studies (including, but not restricted to, LV-TTEs).

We compared TTE volume and TTEs per 100 outpatient visits between the PAVA group and controls. We evaluated the association between the introduction of LV-TTEs at the PAVA and TTE volume with multivariable linear regression, controlling for hospital fixed effects, number of outpatient visits, and mean patient age. We tested a level change (time independent) in use of TTEs after 2005. Because the rate of adoption of LV-TTEs likely increased over time, we also evaluated the difference in the slope of TTE use at the PAVA.

Transthoracic echocardiogram volume, number of outpatient visits, and mean age were similar at the PAVA and control facilities in 2005 (Table). Limited TTEs as a percentage of total TTEs increased from 0.8% (19 of 2395) in 2005 to 19.5% (2840 of 14541) between 2011 and 2015 at the PAVA. Use of limited TTE increased from 0.04 per 100 outpatient visits in 2005 to 0.87 between 2011 and 2015 (Figure). There was no significant change in total use of TTEs: 4.55 TTEs per 100 visits in 2005 and 4.45 between 2011 and 2015.

With multivariable regression, the introduction of LV-TTEs was associated with 349.5 (95% CI, 331.2-367.7; P < .001) more limited TTEs but no significant change in total TTEs (–82.4; 95% CI, –185.3 to 20.6; P = .12). With the second model, the order was associated with 66.7 (95% CI, 63.4-69.9; P < .001) additional limited TTEs per year after implementation, without significantly affecting the change in total TTEs.

Evaluation of left ventricular ejection fraction is one of the most common indications for a TTE. The order for LV-TTE gave clinicians the option of a simplified test that addressed this indication in a fraction of the time and number of images of a full study. Less resource-intensive diagnostics can increase or decrease resource use, depending on whether they are used for patients who would not have been tested or as a substitute for more comprehensive testing. Our results suggest that the LV-TTE was substituted in place of a full TTE.

Focused handheld echocardiography is another means of less-intensive testing with lower equipment costs. However, there are also advantages to laboratory-based studies.⁶ For example, these studies are performed by trained sonographers, images are recorded for future review, and a trained echocardiographer generates a report for the electronic medical record. These features are often lacking with focused handheld echocardiography studies. The tradeoffs between studies using handheld echocardiography and those using laboratory-based echocardiography for a health system warrant further evaluation.

We did not have data before 2005 to evaluate existing trends, and there may be unmeasured factors affecting the use of limited and full TTEs. Because we facilitated a less-intensive option at the discretion of the ordering physician—without restricting full studies—we believe that adverse clinical effects were unlikely. However, further study is warranted given the potential implications of overlooking key findings.

A focus of high-value care has been eliminating unnecessary testing. Another approach would be to identify lower-cost alternatives. Our study suggests that echocardiography testing can be redesigned to answer focused questions important to clinicians while reducing health care spending.

Accepted for Publication: May 25, 2018.

Corresponding Author: Alexander T. Sandhu, MD, MS, Division of Cardiology, Stanford University, 300 Pasteur Dr, Stanford, CA 94305 (ats114@stanford.edu).

Published Online: July 23, 2018. doi:10.1001/jamainternmed.2018.3317

Author Contributions: Drs Sandhu and Heidenreich had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Sandhu, Heidenreich.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Sandhu.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Sandhu, Parizo, Heidenreich.

Obtained funding: Heidenreich.

Administrative, technical, or material support: Sandhu, Heidenreich.

Supervision: Heidenreich.

Conflict of Interest Disclosures: None reported.