Cardiac risk scores are a handy tool for hospitalists handling perioperative assessment, offering a convenient way to stratify patients into low- and high-risk groups. However, experts caution that it's important to understand how the tools were derived as well as their individual strengths and limitations.
“Hospitalists should know which populations were studied to create the tools, the complications and endpoints that were considered, and the time frame used,” said Steven L. Cohn, MD, MACP, professor emeritus at the University of Miami Miller School of Medicine and author of the Society of Hospital Medicine's Perioperative Cardiac Risk Assessment module. “It's easy to plug in numbers, but unless you really know what you're using and what it's really predicting, you may be making an error.”
In addition to needing background data to choose from the handful of available scores, hospitalists also have to remember to consider the broader context, including patient preferences regarding potential surgical outcomes. Add to that COVID-19, which, as usual, increases both the risks and the uncertainty.
Experts recently offered their advice on navigating these challenges and effectively using cardiac risk scores for perioperative evaluation.
The most recent guidance from the American College of Cardiology and the American Heart Association (ACC/AHA) recommends using one of three scoring systems to estimate patients' risk of major cardiovascular events following noncardiac surgery: the Revised Cardiac Risk Index (RCRI), the American College of Surgeons' surgical risk calculator (ACS-SRC), or the Gupta Perioperative Risk for Myocardial Infarction or Cardiac Arrest (MICA) calculator.
The RCRI is the simplest to use, experts said, and the only one based on prospective data. The latter two are based on data collected in the ACS National Surgical Quality Improvement Program (NSQIP) database, which is continuously updated.
The RCRI is also the most frequently used of the three recommended tools, according to a review article in the Winter 2021 International Anesthesiology Clinics coauthored by Dr. Cohn.
It includes six risk factors for complications: high-risk surgery, history of ischemic heart disease, congestive heart failure, history of cerebrovascular disease, use of insulin for diabetes, and preoperative serum creatinine level greater than 2 mg/dL. Patients are assigned a score between zero (lowest risk) and six based on the presence of these risk factors.
“If the RCRI is zero and the patient is asymptomatic, you usually don't have to worry about underestimating risk,” said Kurt Pfeifer, MD, chief of perioperative and consultative medicine at the Medical College of Wisconsin in Milwaukee. “We often use RCRI as a quick first look at a patient. If their score is higher than zero, we may turn to one of the other tools.”
RCRI is the go-to tool for many hospitalists who specialize in perioperative medicine, said Paul J. Grant, MD, FACP, director of perioperative and consultative medicine in the division of hospital medicine at the University of Michigan in Ann Arbor. Although it is now more than 20 years old, RCRI is the most validated of the three recommended calculators and is still widely used in perioperative medicine research.
“Current studies that examine common issues—such as whether or not we should use statins or beta-blockers to decrease risk perioperatively—often use the same risk factors as used in the RCRI,” said Dr. Grant, coauthor of an Annals for Hospitalists article on cardiovascular risk assessment published Feb. 20, 2018, by Annals of Internal Medicine. “It's helpful to know that the patients in subsequent risk reduction studies meet the same risk criteria used in the RCRI study.”
That said, it's also useful to understand that the RCRI was derived from a single-center study that included patients with expected hospital stays of two days or more, thus excluding most ambulatory surgery and low-risk procedures, said Dr. Cohn, author of a new book, “Decision Making in Perioperative Medicine: Clinical Pearls,” co-published by ACP. In addition, it only considers complications that occurred in the hospital, whereas the other major tools look at outcomes up to 30 days after discharge.
The ACS-SRC, which was developed using standardized clinical data from hospitals participating in NSQIP, is the most comprehensive of the three scoring systems. It includes 21 risk factors to predict 30-day outcomes and allows physicians to input CPT codes for specific surgical procedures.
There are several notable benefits to using ACS-SRC compared with RCRI, according to an article in the Oct. 1, 2018, Journal of Cardiothoracic and Vascular Anesthesia (JCVA). For example, it was developed using a diverse surgical cohort of over 1 million patients from hospitals across the country, making it more robust and generalizable than RCRI. Additionally, the ACS-SRC includes many more variables that influence postoperative cardiac complications, the authors noted. For example, unlike the RCRI, the ACS-SRC accounts for functional status, which is a known predictor of adverse postoperative outcomes.
The ACS-SRC also has the benefit of being periodically recalibrated using based on new data, Dr. Pfeifer noted. “In fact, it was just recently recalibrated again. This is a key point in my opinion,” he said. “As validations of the RCRI over time have shown, the rates of complications are evolving and an RCRI score that was once considered low risk no longer is.”
While RCRI may seem simpler and less cumbersome compared to more comprehensive tools, the JCVA article noted, the latter have become more user-friendly over time thanks to web-based interfaces and integration with electronic health record systems (EHRs). All of the scores have online calculators, but availability in one's EHR varies by hospital.
The MICA calculator is based on a study that used a validated cohort of patients from the ACS NSQIP database to identify five independent risk factors for major cardiovascular events within 30 days of surgery: type of surgery, dependent functional status, serum creatinine level more than 1.5 mg/dL, American Society of Anesthesiologists (ASA) physical status class, and increasing age.
Dr. Cohn's review noted that one of the biggest limitations of the MICA calculator and ACS-SRC is their narrow definition of cardiac events (myocardial infarction and cardiac arrest). In contrast, the RCRI uses a broader definition of major cardiovascular events, including pulmonary edema, ventricular fibrillation or primary cardiac arrest, and complete heart block during the index hospitalization.
Tools in practice
Most hospitalists adopt whichever tool they used during residency training, said Dr. Grant. In general, he recommends that physicians become familiar with all three but choose one tool that they're most comfortable with to use regularly.
MICA is the go-to tool for Avital O’Glasser, MD, FACP, associate professor of medicine and medical director of the pre-op clinic at Oregon Health & Science University in Portland. It's simple to use, she said, while also taking into account factors not included in the RCRI, such as functional status and ASA score. She uses ACS-SRC as a backup because it is much more granular in defining surgical type according to CPT code.
“I don't use MICA for all patients, because the bucket categories are very broad compared to those used in ACS,” she said. For example, in a patient scheduled for spinal surgery, MICA does not distinguish between the risk associated with repair of a herniated disc (typically performed in under an hour) and a much longer operation for scoliosis.
“Because I use MICA every day, I can approach it with this level of insight,” said Dr. O’Glasser. “If I feel like MICA may be over- or underestimating risk based on those big bucket categories or procedures, then I may turn to the ACS.”
Another potential advantage of ACS-SRC is its use of color-coded graphics, said Dr. Cohn. “With ACS, I can print out a colorful graphic that I can show to patients to help them understand what their percentage predicted risk is and whether it is lower or higher than average compared with other similar patients,” he said. The graphics can potentially help hospitalists interpret risk scores for patients and more clearly explain their health care options.
This can be critical, considering the difficulty of patient-physician communication about risk, highlighted by a study published in the December 2018 Perioperative Care and Operating Room Management. More than half of patients surveyed believed high risk corresponded to a greater than 50% chance of an adverse event, whereas most clinicians generally defined it as greater than 5%.
“That paper's findings highlight that percentages can be very abstract for patients, and we shouldn't necessarily use them in discussions with patients,” Dr. O’Glasser said. “Instead of telling someone they have a 10% risk of heart failure, for example, I try to create a narrative of best- and worst-case scenarios using patient-centered language.”
Other ways to measure
While each of the three tools recommended by ACC/AHA have advantages, it is difficult to compare them directly because each is based on different types of patient populations, surgeries, and predictive outcomes, said Dr. Pfeifer. As a result, they should be used only as aids to help guide clinical decision making, along with a thorough history and physical exam.
“We'd all like to have one tool that allows us to plug in information and get a full picture of a complex patient care scenario, but that's just not realistic,” he said. “Also, we have to keep in mind that these tools are only designed to capture ischemic cardiac risk—you have to think about the other issues separately.”
For example, a patient's risk score does not account for risks that stem from other types of cardiac disease, such as pulmonary hypertension and severe aortic stenosis, he said. In such cases, the hospitalist should consult with the anesthesiologist, surgeon, and cardiologist to determine whether the surgery can proceed.
These commonly used risk calculators may also become less relevant over time, partly because cardiac risk has declined due to safer surgical procedures and practices, said Dr. Grant. The most serious postoperative events—massive myocardial infarction and death—are now rare. At the same time, he added, there is growing evidence that biomarkers may be effective in predicting cardiac risk.
In the Annals for Hospitalists article, Dr. Grant noted that the most recent Canadian Cardiovascular Society guidelines recommend measuring brain natriuretic peptide (BNP) and N-terminal proBNP, which are simple and inexpensive risk predictors, before surgery. (Troponin levels can also be used postoperatively to identify mortality risk after surgery in asymptomatic high-risk patients.)
Emerging data also suggest that having surgery within 30 days of a diagnosis of COVID-19 is associated with worse clinical outcomes, postsurgical complications, and mortality, according to a commentary coauthored by Dr. O’Glasser, published Jan. 7 by Perioperative Medicine. The article introduces a protocol for preoperative assessment of COVID-19 survivors, stratified by surgery and degree of illness.
However, COVID-19's impact on cardiac risk is still unclear. “We know that operating shortly after a COVID diagnosis is fraught with peril,” said Dr. Pfeifer. “It's just another condition that, even though it's not yet part of any tool, should make anyone pause and wonder if it's really the right time for a procedure or if other options should be considered.”
Clinicians often put a lot of emphasis on a patient's score, but it should be just one part of the overall assessment, noted Dr. O’Glasser. She offered the example of an elderly patient who had been scheduled for oral cancer surgery that would include major reconstruction and insertion of a feeding tube. Based on her cardiac risk score alone, the woman did not require a stress test and could be medically cleared for surgery. However, Dr. O’Glasser saw reason not to give a green light.
“Based on what we knew about the patient and this type of surgery, we estimated that she could be in the hospital for well over a week and have an 80% chance of going to rehab,” she said. “That wasn't consistent with her personal preference to avoid a long hospital stay and remain in her home; ultimately, she opted to try a new oral chemotherapy instead of pursuing surgery.”