Deferred Testing in Low-Risk Patients Safe, May Curb Symptoms: PRECISE

These data support the idea that deferral, when informed by CT, is a “reasonable strategy,” says Pamela Douglas.

Deferred Testing in Low-Risk Patients Safe, May Curb Symptoms: PRECISE

For stable, symptomatic patients with suspected CAD who are deemed to be at “minimal risk” using a CT-based tool, deferral of diagnostic testing is just as safe as the usual approach and is even associated with similar symptom relief, according to a prespecified subanalysis of the PRECISE trial.

The original PRECISE results show that following a pathway informed by the PROMISE Minimal Risk Score (PMRS)—derived from the PROMISE trial—reduced the composite clinical endpoint of death, nonfatal MI, and catheterization without obstructive CAD at 1 year by about 70% compared with usual care among the entire population of patients presenting with stable angina.

Here, the emphasis is on those considered minimal risk, who made up 20% of that larger group.

Senior author Pamela Douglas, MD (Duke Clinical Research Institute, Durham, NC), told TCTMD the prespecified subanalysis, published online today in JAMA Cardiology, was designed to help “figure out how to evaluate chest pain better.

Our guidelines have long said you don't need to test low-risk people,” she continued. “But there has never been a trial to say that it's feasible and it's safe and acceptable to patients, and to physicians for that matter.”

Without any evidence to say deferral of diagnostic testing is safe and feasible, most clinicians “have probably erred on the side of testing,” Douglas said. “This will provide docs and patients with a great deal of comfort that a strategy of not testing immediately is a reasonable strategy and will give them the tool in terms of the risk score and the cut point of how to do that.”

Similarly, Leslie Cho, MD (Cleveland Clinic, OH), told TCTMD the results are reassuring for both physicians and patients alike. “We live in a very litigious society where . . . sometimes patients demand testing, but I think one really great thing about this study is that in low-risk patients—and this is really low risk—that you can defer testing and they do great, they do just fine,” she said. “What is really interesting is the chest pain or the angina pain that they had goes away by a year regardless of whether you do testing or not.”

Deferred Testing Safe, Effective

For the analysis, James Udelson, MD (Tufts Medical Center, Boston, MA), Douglas, and colleagues looked at the 422 PRECISE participants who, as virtue of being deemed at minimal risk by the PMRS, had been randomized to either deferred (n = 214) or usual testing (n = 208). Notably, these patients tended to be younger (mean age 46 years) and more often were women (72%).

Over the trial’s median follow-up period of 11.7 months, two-thirds (64%) of patients randomized to deferred testing never underwent any testing. Of the 36% who did undergo testing, 24% had coronary CT angiography (CCTA) with or without CT-derived fractional flow reserve (FFR-CT), 11% had a functional test, and 0.5% were referred directly for invasive coronary angiography. Reasons associated with testing in the deferred-testing arm included worsening symptoms (30%), uncontrolled symptoms (10%), or new symptoms (6%) as well as changing clinician (19%) or patient preference (10%). These tests were performed at a median 48 days after randomization, and in 96% neither obstructive CAD nor inducible ischemia were found.

For those randomized to usual testing, 83% had a functional stress test, 3% were referred direction for invasive coronary angiography (with obstructive CAD found in none), 1.4% underwent CCTA with or without FFR-CT, and 13% had no testing done over a median follow-up period of 12 months. Initial tests in this group took place a median of 15 days post-randomization, with 93% of them showing neither obstructive CAD nor inducible ischemia.

The primary composite endpoint—all-cause death, nonfatal MI, or catheterization without obstructive CAD—was reported in 0.9% of the deferred-testing arm and 6.3% of the usual care group (adjusted HR 0.16; 95% CI 0.04-0.70).

There has never been a trial to say that it's feasible and it's safe and acceptable to patients, and to physicians for that matter. Pamela Douglas

None of the deferred testing participants reported death or MI, but one of each of these events occurred in the usual testing group. Fewer patients in the deferred testing group had catheterizations without obstructive CAD compared with usual care (0.9% vs 5.8%; P = 0.02).

Frequent angina was noted in 70% of all patients at baseline, with a similar improvement in this endpoint to less than 20% at 12 months observed in both groups.

This reduction in symptoms especially among those who did not undergo any testing, was notable to Udelson. “It was somewhat unexpected and somewhat, as many things are in trials like this, unexplained,” he told TCTMD. “For all these hundreds and hundreds of people, you can't track everything that is done to them. But I would find that very reassuring if I was a patient or as a clinician.”

Douglas agreed. “That's real confirmation that it is safe and feasible to not test our lowest-risk patients as identified by the risk score that we use,” she said.

But overall, the way the results will be incorporated into practice will depend upon patients for the most part, according to Udelson. “There are some people who have symptoms and if you tell them, ‘Okay, we have run you through this tool, and it is extremely unlikely that you have disease or anything bad is going to happen and it is okay not to do a test,’ they are like ‘Great thank you!’ They're thrilled,” he said. “Other people, that would make them very anxious and they are made less anxious by a normal test result. So, you do have to feel that out with patients.”

Patient Acceptability, Physician Preference

It’s not rare to see patients who find a deferred testing strategy acceptable, Cho said, but many still go on to see a cardiologist after being told they don’t need testing by a primary care physician. This is where shared decision-making comes into play. “We have to be cognizant,” she said. “Some people are very afraid because they have a family history. Even though they are at super low risk, they are just very afraid. And sometimes talking to them, reassuring them does not necessarily help.”

“The issue of patient acceptability is an important one,” Douglas said. “Sometimes our patients say, ‘My neighbor or my cousin or my whatever had pains like this and got a stress test, why don't I?’ . . . And so I think it's very encouraging that in this patient-centered world that we now live in, that not testing is an acceptable strategy to patients.”

In situations like these, Cho recommends showing patients the data from this study to help convince them to wait.

What is really interesting is the chest pain or the angina pain that they had goes away by a year regardless of whether you do testing or not. Leslie Cho

But physician preference also matters. “Sometimes talking to patients and not getting a test takes longer,” Cho explained. “You are in the office and it is much easier to get a test and get them out of the room than to have a long discussion. . . . Sometimes not getting a test is much harder than getting a test, and I think we have to be cognizant of the pressures that our physicians face too.”

That’s where a system-wide management plan to reduce inappropriate catheterization would be helpful, according to Udelson. “As more and more people might be covered with risk-based contracts where physicians or clinicians might be at risk for medical expenses, this is something that would be really attractive as a way for a system to manage down appropriately the number of people who get cathed unnecessarily,” he said. For that to happen, “you need buy-in from the clinicians, and this is always about a discussion with patients.”

‘NONOBSTCAD’ an Appropriate Endpoint?

In an accompanying editorial, Raymond J. Gibbons, MD (Mayo Clinic, Rochester, MN), writes that while he agrees with the rationale of the study, the lower rate of nonobstructive CAD on catheterization (NONOBSTCAD) in the deferred-testing arm does not strengthen the evidence in favor of that strategy. “It is not a measure of patient safety, but rather a measure of physician preference,” he argues.Although Udelson et al do not indicate how many patients in the substudy were treated by physicians who favored an invasive strategy, in the usual testing group, 3.3% (seven of 208) were sent directly to catheterization, suggesting that they might have been treated by such physicians. All had NONOBSTCAD.

“In my opinion, NONOBSTCAD should have been a secondary endpoint in this substudy,” Gibbons continues. “Its use as a component of the primary endpoint is not justified and potentially misleading.”

Even so, the rest of the evidence in the substudy indicate that “deferred testing is safe,” he concludes. “Hopefully, this new evidence will enable physicians to better educate such patients regarding their options. Such shared decision-making could potentially lead to fewer unnecessary tests.”

In response, Udelson said he could see why Gibbons found the use of NONOBSTCAD as part of the primary endpoint “somewhat inappropriate.” However, he said, “our thinking was we are trying to blend having the structure of a clinical trial with the pragmatic effectiveness of as real life as we can be. And we know from many studies that a lot of people who go to the cath lab for an invasive catheterization don't have coronary disease. So, it's just something that is done and when you show a way to reduce that, that seems to be okay.”

Similarly, Douglas said, “we could argue that you should never [go direct to cath], and maybe Dr. Gibbons never does that, but some people do in the world. And our data show, although the numbers are very small, that those are really unlikely to have obstructive disease and that cath is perhaps a bit of an overkill. But we haven't had data before.”

Going forward, she would like to see these findings confirmed in larger data sets, whether observational or randomized. “I think people need to use our method and apply it in practice and see how it works for them, as with any new tool,” Douglas said. “And I'd like to see us reduce testing of the lower risk patients.

  • The PRECISE study was sponsored by HeartFlow.
  • Udelson and Douglas report receiving grants from HeartFlow during the conduct of the study.
  • Gibbons reports receiving personal fees from AlphaSight.