Stress Testing Before Low-Risk Surgery: So Many Recommendations, So Little Overuse

The Choosing Wisely (CW) campaign, which commenced in 2011, focuses on reducing medical services that are of questionable value or may be harmful.¹ In 1996 and 2002, guidelines from the American College of Cardiology and the American Heart Association implied that routine stress testing before low-risk surgeries should be avoided; this was codified in the 2007 guidelines² because the risk of cardiac complications from these surgeries is very low. Consequently, 7 specialty societies for the CW campaign now recommend not performing cardiac stress testing prior to low-risk surgery.³ Recently, Thilen et al⁴ documented that the rate of preoperative consultation for cataract surgery in 2006 approached 20% among Medicare patients, but they did not comment on the use of stress testing. Therefore, we sought to determine the prevalence of cardiac stress testing before low-risk surgeries, prior to commencement of the CW campaign, to estimate the potential effect of the recommendations on future use of resources. We examined the use of preoperative stress testing in the 2 largest US federally sponsored health care programs: the Department of Veterans Affairs (VA) and fee-for-service Medicare.

We performed a retrospective cohort study using data from the VA’s Corporate Data Warehouse and from a nationally representative 5% sample of Medicare fee-for-service claims. The Ann Arbor VA Human Studies Committee and the Kaiser Permanente of the Mid-Atlantic States institutional review board approved this study with waiver of informed consent. Using Current Procedural Terminology codes, we identified all asymptomatic patients 65 years or older who underwent 1 or more cataract surgical procedures, knee arthroscopies, or shoulder arthroscopies from February 1 to December 31, 2009. Using an approach similar to that used by Schwartz et al,⁵ we then examined the proportion of these patients who had an exercise or pharmacologic electrocardiographic treadmill, echocardiographic, or nuclear stress test in the 28-day period before their first low-risk surgery. To isolate routine preoperative stress tests, we excluded stress tests during that period that also occurred 0 to 30 days after a hospitalization or 0 to 3 days after an emergency department visit. We also examined a more sensitive measure without such exclusions. To assess regional variation, we estimated stress test rates by hospital referral region using 2-level empty logit models.

A total of 22 670 VA patients and 109 270 Medicare patients had cataract surgery, knee arthroscopy, or shoulder arthroscopy from February to December 2009. The mean age of patients was approximately 75 years in both cohorts (Table 1). A routine preoperative stress test preceded 1 of the 3 low-risk surgeries in only 0.67% of VA patients and 2.14% Medicare patients (Table 2). Estimated stress test rates by hospital referral region ranged in the VA from 0.3% to 2.0% (interquartile range, 0.4%-0.7%) and in Medicare from 1.5% to 3.1% (interquartile range, 1.8%-2.1%). Applying the more sensitive measure, 0.8% of VA patients and 2.4% of Medicare patients had stress testing before surgery.

We found that the use of routine preoperative stress testing before low-risk surgery in both VA and Medicare patients was very low and varied little across geographic regions, even before the CW campaign started. Although the rates in the Medicare group were 3 times as high as those in the VA group, these low absolute numbers suggest that interventions to further decrease use of the testing would minimally improve quality while diverting attention away from higher-yield interventions that would more strongly affect care. It appears that most physicians had already incorporated guidelines about appropriate preoperative stress testing into their practices before the CW recommendations became available. Although this is good news for patients, it is not helpful for a campaign that aims to improve appropriateness.⁶ Specialty societies should focus future CW recommendations on services that have high baseline rates of inappropriate care to call attention to areas where interventions can best improve quality.

Corresponding Author: Eve A. Kerr, MD, MPH, Veterans Affairs Center for Clinical Management Research, PO Box 130170, Ann Arbor, MI 48113 (ekerr@umich.edu).

Published Online: February 9, 2015. doi:10.1001/jamainternmed.2014.7877.

Author Contributions: Drs Kerr and Chen 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.

Study concept and design: Kerr, Chen, Sussman, Nallamothu.

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

Drafting of the manuscript: Kerr, Klamerus.

Critical revision of the manuscript for important intellectual content: Kerr, Chen, Sussman, Nallamothu.

Obtained funding: Kerr.

Administrative, technical, or material support: Chen, Klamerus.

Study supervision: Kerr, Chen.

Conflict of Interest Disclosures: None reported.

Funding/Support: This study was supported in part by the Veterans Health Administration’s Office of Informatics and Analytics, grant DIB 98-001 from the VA Diabetes Quality Enhancement Research Initiative, and Career Development award 1K08HS018781-01 (Dr Chen) from the Agency for Healthcare Research and Quality.

Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation or approval of the manuscript; and decision to submit the manuscript for publication. The VA Office of Informatics and Analytics reviewed these results prior to submission.

Disclaimer: The views expressed herein are those of the authors and do not necessarily represent the US Department of Veterans Affairs, the University of Michigan, or Kaiser Permanente.

Additional Contributions: Rob Holleman, MPH, VA Center for Clinical Management Research, and Haihong Hu, MPH, Kaiser Permanente Mid-Atlantic States, provided assistance with data management and analysis. The contributors did not receive financial compensation.