Questioning the KCCQ: Experts Debate Its Validity as QoL Metric

BOSTON, MA—The Kansas City Cardiomyopathy Questionnaire (KCCQ) has been integrated into numerous heart failure (HF) and transcatheter device trials to provide insights into the effects of various interventions on quality of life, but whether those results can be trusted is another question entirely.

Gregg Stone, MD (Icahn School of Medicine at Mount Sinai, New York, NY), in a debate last week at THT 2025, showed data from blinded and unblinded trials in HF and structural heart disease challenging the validity of KCCQ as a measure of quality of life and suggested much of the change in this metric can be attributed to a placebo effect.

“This is a real shame,” he said in a recorded presentation. “I’m sorry to actually give this talk because I really like the KCCQ and I think it should measure heart failure symptomatology. But it does seem like we have a problem.”

But David Cohen, MD (St. Francis Hospital & Heart Center, Roslyn, NY, and Cardiovascular Research Foundation, New York, NY)—who worked with the creator of the KCCQ, John Spertus, MD (Saint Luke’s Mid America Heart Institute, Kansas City, MO), for many years—defended the endpoint, arguing that “the KCCQ is a valid tool for measuring what your patients care about.”

Questioning KCCQ

Changes in patient-reported outcome (PRO) measures like the KCCQ have been associated with changes in endpoints such as death, hospitalization for heart failure, NYHA class, and health status, said Stone during his talk. However, he said, most of these validation studies were performed in unblinded studies susceptible to placebo effects and other biases. “So the real question is: are quality-of-life measures such as KCCQ useful in blinded trials and can they be trusted in unblinded trials?”

In the open-label COAPT trial, for which Stone was one of the principal investigators, adding the MitraClip (Abbott) to medical therapy provided large reductions in mortality and HF hospitalization compared with medical therapy alone in patients with HF and severe mitral regurgitation, in addition to a big gain in KCCQ score; there was minimal change in quality of life in the control arm. A higher KCCQ score at 1 month was associated with lower risks of death and HF hospitalization from 30 days to 2 years.

But questions about the link between KCCQ and clinical outcomes emerged during another trial that Stone helped lead—the sham-controlled RELIEVE-HF study. The trial showed that an interatrial shunt did not improve prognosis or reduce symptoms in patients with HF and did not impact quality of life as measured by KCCQ. The KCCQ scores improved to a similar extent—roughly 10 points—in both the intervention and control arms, with that bounce sustained out to 2 years.

Treatment effects differed based on whether patients had HF with reduced ejection fraction (HFrEF) or preserved ejection fraction (HFpEF). The shunt procedure appeared to improve clinical outcomes in patients with HFrEF and worsen outcomes in those with HFpEF. In both subgroups, however, KCCQ scores increased overall, without a difference between trial arms.

That spurred Stone and his colleagues to look at what happened in other blinded trials, and across several studies of HF drugs. In EMPEROR-Reduced, EMPEROR-Preserved, DAPA-HF, PARADIGM-HF, and FINEARTS-HF, there were significant reductions in hard outcomes, but only small increases in KCCQ scores ranging from 1.3 to 2.8 points. That demonstrates a disconnect between clinical outcomes like mortality and HF hospitalization and quality of life in blinded trials.

To examine unblinded trials, Stone turned to the TRILUMINATE Pivotal trial of TriClip (Abbott) and the TRISCEND II trial of transcatheter tricuspid valve replacement with Evoque (Edwards Lifesciences). Both trials demonstrated large improvements in KCCQ scores (13.5 points in TRILUMINATE and 17.8 points in TRISCEND II). “And that’s what we always see in unblinded trials,” Stone said. But there were no between-group differences in death or HF hospitalization in either study.

Both of those tricuspid regurgitation (TR) devices were approved by the US Food and Drug Administration on the basis of their effects on KCCQ, Stone noted. “But is this all a placebo effect in an unblinded trial? Or perhaps devices that reduce TR are the rare devices that improve symptoms without reducing death or heart failure hospitalization. I honestly don’t know.”

Wrapping up, however, he reiterated that the results of unblinded trials may be affected by placebo and Hawthorne effects and various biases. “Soft endpoints, such as PROs and exercise performance, are especially at risk from these confounding influences, and KCCQ falls into this category,” Stone said, arguing that future randomized trials or devices should be sham-controlled whenever possible to eliminate uncertainty about the findings.

“For now, given the discordance between the change in KCCQ and cardiovascular outcomes in blinded trials, the utility and interpretation of KCCQ in these trials that are blinded—and, by implication, in unblinded trials—is uncertain, and it’s probably best to avoid using this PRO,” Stone argued.

KCCQ Has Value

Kicking off his defense of the KCCQ, Cohen agreed with Stone that “a sham-controlled trial is definitely the ideal way to evaluate health status after a procedure.”

But that “doesn’t mean that unblinded studies are completely worthless if you’re willing to think,” he added. “I’m going to defend the KCCQ as best I can as a measure that provides value sometimes, even in unblinded trials.”

He stressed that the KCCQ is not a surrogate for mortality or any other outcome. “I don’t care if it’s not a surrogate outcome for mortality,” he said. “Mortality is mortality. Quality of life is quality of life. Patients care about both. We should measure both.”

Cohen pointed to the Tri-QOL study to show that among patients with severe TR, the most common symptoms—dyspnea, fatigue, and leg edema—are captured well with the KCCQ. Though there are other symptoms like loss of appetite and upper-extremity edema that are not assessed with the KCCQ, they typically occur in the presence of symptoms that are. “What that means from a development standpoint is the KCCQ still should be reasonably valid because it’s measuring the things that are most important to our patients and the other things that do occasionally crop up don’t occur in isolation,” he said.

Another analysis of the Tri-QOL study, published last year in JAMA Cardiology with Cohen as senior author, validated the KCCQ in patients included in 11 studies of transcatheter tricuspid valve interventions. The tool “met all of the standards for a valid health status instrument,” with consistent results in patients with HFrEF, HFpEF, and hypertrophic obstructive cardiomyopathy, Cohen said. “It’s measuring what our patients care about and it’s measuring it in a valid and reliable way.”

Cohen disputed that placebo-controlled trials can provide an estimate of the size of the placebo effect, using the SUMMIT trial to illustrate his point. Treatment with tirzepatide resulted in a larger improvement in KCCQ score than did placebo (19.5 vs 13 points). The difference represents the treatment effect, but the gain in the placebo group cannot all be attributed to a placebo effect, Cohen indicated, saying that it’s driven by a combination of placebo effect, Hawthorne effect, and regression to the mean.

A typical placebo effect is about one-quarter of a standard deviation, which is about a 5-point change in the KCCQ, said Cohen. In the tricuspid intervention trials, the gain in KCCQ has been closer to 10 or 20 points, “so it’s certainly larger than what we would expect” if it were all attributed to a placebo effect, he said.

Moreover, with a typical placebo effect, the benefits would be expected early on and waning over time. In TRISCEND II, however, there is an early increase in KCCQ but it grows even larger by 6 months. “That’s not typical for a placebo effect,” Cohen said. “That’s what medications do. That’s what effective therapies do. So again, to me, this is evidence, at least circumstantial evidence, in favor of it not all being a placebo effect.”

Cohen also presented data showing that quality of life in patients with tricuspid valve disease is associated with hard clinical outcomes as well as objective measures of disease severity. In TRILUMINATE, the gain in KCCQ was 4 points for every one-grade reduction in TR severity, and in TRISCEND II, patients with the most-severe TR at baseline had the largest increases in KCCQ scores a year after the intervention. That is “strongly suggesting a dose-response effect, which is very unlikely to be mediated by placebo,” Cohen said.

Finally, Cohen said there is a correlation between the change in KCCQ and the difference in death or HF hospitalization in various device and drug trials.

“For many older patients, improved health status and quality of life are at least as important as improved survival, and understanding whether a device impacts these outcomes is essential,” Cohen concluded. Any study that employs KCCQ as part of the primary endpoint “should be blinded to remove all doubt and to convince the skeptics, which includes payers,” he said, reiterating the need for sham-controlled trials.

One of the panelists at the debate session, Noam Josephy, MD, a vice president and general manager for Abiomed, said knowing whether the KCCQ is a valid measure is important because it “carries a huge burden” for industry sponsors when introduced into studies in terms of translation, maintenance of databases, and monitoring. He called for guidance for industry on “where to use it, where to not use it, and how to use it.”