Does Following Perioperative Cardiovascular Evaluation Guidelines Increase Perioperative Costs?

Lee A. Fleisher, MD; Joshua A. Beckman, MD, MS; Duminda N. Wijeysundera, MD, PhD


Anesth Analg. 2019;128(2):202-203. 

In 1996, the American College of Cardiology (ACC) and the American Heart Association (AHA) first published multispecialty Guidelines on Preoperative Evaluation for Noncardiac Surgery. The guidelines took the novel approach, at the time, that testing should be reserved for those patients in whom the results would impact care. This philosophy has continued through all of the subsequent iterations of the guidelines and is the foundation on which committee members (which include the 3 authors of this editorial) evaluated the evidence and developed the algorithm. In the last 2 decades, improvements in preoperative risk assessment, anesthetic care, surgical technique, and postoperative surveillance have resulted in a significant decrease in postoperative mortality.[1] The question remains as to whether the approach advocated by the ACC/AHA guidelines is cost-effective. In this issue of the Anesthesia & Analgesia, Pappas et al[2] used the well-validated American College of Surgeons (ACS)–National Surgical Quality Improvement Project (NSQIP) database and 3 clinical risk indices to determine both concordance among the indices and financial burden resulting from using these testing algorithms. The authors concluded that the rates of cardiovascular testing are a function of the index or calculator used and that the potential financial burden of testing could be >$1.7 billion annually.

The work by Pappas et al[2] points out the limitations of guidelines, which require accurate interpretation. Misinterpretation of the guidelines may even raise potential hazards, even with the incorporation of the class of recommendation placed directly in the care algorithm. The 2014 ACC/AHA algorithm does not advocate any class I indications for preoperative testing.[3] The decision to refer for preoperative noninvasive stress testing is a class IIa recommendation reserved for patients who are at elevated risk, who have poor or unknown functional capacity, and in whom test results will impact care. Multiple societies have included class III recommendations against routine preoperative noninvasive cardiac testing in the Choosing Wisely campaign. With evidence demonstrating minimal value for preoperative coronary revascularization, medical therapy, and diagnostic monitoring, the indications for further testing are limited to a small subset of the high-risk patients.

The authors correctly point out the target population where noninvasive stress testing might be considered varies considerably based on the choice of preoperative risk index. The 2014 ACC/AHA guidelines recommend that clinicians use either the revised cardiac risk index or ACS–NSQIP risk calculator to screen for low-risk patients (ie, <1% predicted risk of postoperative death or myocardial infarction) in whom further cardiac testing is inappropriate. The current article by Pappas et al,[2] as well as another recent study that also used the ACS–NSQIP registry,[4] have shown that these 2 tools are not interchangeable approaches to screen out low-risk patients. Specifically, in about one-fifth of cases, these 2 risk indices disagree as to whether an individual patient is low risk or otherwise. This consistent finding across 2 studies points to the need for a high-quality comparison between these 2 preoperative clinical risk assessment tools. Both instruments have important strengths and limitations. For example, while the Revised Cardiac Risk Index is a simple and well-validated tool,[5] was derived in a single-center cohort study performed >2 decades ago, includes components with uncertain prognostic value (eg, diabetes requiring insulin therapy),[6] and employs a simple categorization scheme for operative risk. By comparison, the ACS–NSQIP risk calculator was derived in a large multicenter sample and includes more gradation of operative risk. Conversely, it has not undergone extensive external validation outside the ACS–NSQIP registry and was developed in a registry that did not mandate standardized postoperative surveillance for myocardial infarction. Importantly, the absence of standardized postoperative surveillance may have led to a significant underestimation of postoperative myocardial infarction rates. What is needed to address this area of clinical uncertainty is concurrent external validation of the 2 risk indices in a cohort study that implements standardized contemporary postoperative surveillance for cardiac events (ie, clinical evaluation, electrocardiograms, troponin concentrations).

Even if the patient is shown to be at elevated risk with poor exercise tolerance, it is rare that the test will change management. Current therapeutic approaches to mitigate the risk of noncardiac surgery are limited. In the recent past, large randomized controlled studies have demonstrated no benefit for β-adrenergic blockade, clonidine, aspirin, preoperative coronary revascularization, and routine pulmonary artery catheterization.[7–11] The evaluation of subgroups has shown that aspirin provides benefit for patients with previous percutaneous coronary intervention,[12] and coronary artery bypass may be beneficial for patients with left main coronary artery disease.[13] The former should be well known after clinician interview, while the latter is rarely clinically silent in the absence of a severe, secondary functional limitation. In that case, testing in the guidelines is advocated: the patient who may have severe coronary artery disease but whose greatest functional limitation arises as a result of a different problem. As far as we know, the results of testing will rarely instigate the use of or change in medical therapies to reduce risk, so the purpose of testing is almost exclusively limited to the discovery of previously unknown, severe coronary artery disease in the patient who is willing to undergo surgical revascularization or identifying those in whom testing may lead to changes in the surgical procedure or use of palliative measures; a small population indeed. This reality is demonstrated in the literature. Of the 194,444 patients who underwent preoperative coronary angiography from 2009 to 2014 in the National Cardiac Data Registry CathPCI Registry, preoperative stress testing was performed in 65% of the subjects.[14] Left main coronary artery occlusive disease was found in 5.6% of the patients, of whom half were revascularized. This finding does not seem unreasonable, particularly in the face of preoperative cardiac stress testing rates being ≈2% in the United States from 2002 to 2010.[15] It seems likely that the estimates of testing by Pappas et al[2] overestimate actual practice. Thus, the fear of a significant excess of superfluous testing raised by Pappas et al[2] may not be reflected in actual practice. We agree that the further reduction of testing and cardiac catheterization should be pursued because type 1 myocardial infarction remains a modest cause of postoperative mortality and stricter adherence to guidelines would result in even lower preoperative cardiac testing utilization.[16] Indeed, the benefit of coronary revascularization performed in 40% of the patients with single- and 53% of the patients with 2-vessel asymptomatic coronary artery disease in the National Cardiac Data Registry remains unknown in the face of the Coronary Artery Revascularization Prophylaxis trial, which showed that revascularization did not reduce 30-day or 3-year mortality.[10]

In conclusion, the article by Pappas et al[2] pointed out the potential financial burden and excess low value testing that would occur if a large majority of patients who were at elevated risk (>1%) underwent cardiovascular testing. We urge clinicians to utilize the guidelines to identify those who should proceed directly to surgery and reserve further cardiovascular testing for those patients with poor or unknown exercise tolerance in whom care will truly change and the probability of a major cardiovascular event may plausibly be reduced. The use of such a strategy will result in many fewer tests ordered and more cost-effective tests.