Author + information
- Published online June 24, 2019.
- D. Marshall Brinkley, MD∗ ( and )
- Lynne W. Stevenson, MD
- ↵∗Address for correspondence:
Dr. D. Marshall Brinkley, Vanderbilt University Medical Center, Division of Cardiovascular Medicine, 1215 21st Avenue South, MCE 5th Floor, Nashville, Tennessee 37232.
“The woods are lovely, dark, and deep, but I have promises to keep.”
—Robert Frost, (1)
The Rise of Patient-Reported Outcomes
We strive to help our patients feel better and live longer. Guideline-directed medical therapies (GDMTs) for heart failure (HF) have substantially decreased hospitalizations and mortality. However, the effects of GDMT on function and quality of life have been difficult to ascertain, because most benefits demonstrated in this domain have been to reduce future decline from disease progression. Multiple health organizations now agree that “the definition of health and concepts of patient-centered care directly support the measurement of patient health status as a key metric of cardiovascular health” (2).
Patient-reported outcomes (PROs) are provided directly by patients about how they feel and function (2). PROs reproducibly track symptom burden, functional status, and health-related quality of life (HRQOL), providing insight into the full impact of disease. Their impact has been best validated for single conditions dominated by interventions to relieve symptoms, as in orthopedic clinics. For the complexity of HF, the Minnesota Living With Heart Failure Questionnaire and the Kansas City Cardiomyopathy Questionnaire (KCCQ) are HF-specific tools with psychometric validation (3). These PROs support regulatory approval, facilitate comparative effectiveness study, and correlate with mortality.
It is not yet known how these tools would be effective in routine practice. In their introduction, the investigators of the present study (4) in this issue of JACC: Heart Failure, describe 2 separate challenges: 1) to ascertain whether PROs are actionable in clinical practice; and 2) whether they can serve as a performance measure for which providers should be held accountable. Could KCCQ scores integrate both HRQOL and GDMT performance metrics?
Can We Link Patient-Reported Health Status to Appropriate Use of GDMT?
To provide new information on how PROs may relate to therapies, Thomas et al. (4) compared changes in HF medications within 7 days of baseline with changes in KCCQ score (12-item version) between baseline and 3 months in a multicenter HF registry of 3,313 outpatients with ejection fractions ≤40% on at least 1 goal-directed medication. This registry provides us with valuable information about the contemporary population and therapies for heart failure with reduced ejection fraction.
The median improvement in KCCQ-12 score by 3 months was higher in the 22% of patients with medication changes than in the 78% of patients with no medication changes. A very large improvement of ≥20 points occurred in 11% of patients, of whom 26% had medication changes and 14% did not. Changes of ≥5 points, the threshold considered clinically meaningful, occurred in 70% of patients, three-quarters of whom had no medication changes. Among the 1,098 patients (42%) with improvements of ≥10 points, fewer than 1 in 3 had medication changes. Thus, most of the changes in PROs and medications were uncoupled. Moreover, it is not clear how changes in KCCQ score aligned with changes in therapy. Although decreases in therapy were less common, they were associated with more HRQOL improvement than therapy increases. The magnitude of HRQOL improvement was highest with diuretic agents but almost as high with sacubitril/valsartan. These data do not provide strong support for the conclusion that “health status–based performance measures can quantify the benefits of titrating medicines.”
Over a longer time than 3 months (Figure 1), HRQOL and GDMT should converge to better health in patients who are adherent and responsive. The kinetics may differ substantially in patients with more advanced disease than in this study, dominated by class I and II patients, as changes in medications were more common after recent hospitalization, and improvements were more likely from worse baseline HRQOL (4).
How Can HF PROs Be Made Actionable?
This study was not designed to determine how routine use of PROs might improve clinical care, as the PROs were not presented to the providers. Practical challenges were discussed on November 6, 2018, at a forum on the integration of PROs into cardiovascular care sponsored by the American College of Cardiology. Administrators and payers wanted to know how PROs could be used to measure value. Patients spoke eloquently of the need to see providers respond to the information. Providers emphasized the importance of seeing PROs in a trackable format that would be actionable. Implementation requires serious commitment, as demonstrated by Stehlik et al. (5); despite extensive staff training, only 58% of his patients completed PROs and electronic assessment took an average of 6.7 min prior to the visit.
The KCCQ score is collapsed into a global summary score for which high absolute scores and positive changes correlate with better prognosis. However, the summary scores do not serve to direct therapy beyond vigilance and discussion of advanced therapies or goals of care.
The score includes domains for symptoms and severity, physical activity and roles, and overall HRQOL. Even the scores from these individual domains are rarely “actionable.” The domain most likely to trigger action is symptoms, but linking overall symptom severity with intervention requires that symptoms be linked with physiology that can be addressed (Figure 1). For example, positive answers to questions about edema and orthopnea might focus the examination toward signs of congestion. Decongestion is the intervention most likely to alleviate symptoms. New analysis of the PARADIGM trial showed inverse changes of 5 KCCQ points for appearance or resolution of each of 4 physical findings of congestion (6). Additionally, the number of these signs of congestion predicted hospitalization and mortality independently from natriuretic peptide levels or calculated survival scores. In contrast, fatigue often limits function and HRQOL but is influenced by many factors and is difficult to address.
How Do PROs Link to Outcomes?
Incorporation of PROs into standard HF management could improve outcomes if they are reviewed and the review triggers actions that improve outcomes. Of the outcomes, patient-reported HRQOL has face validity as an important outcome in itself. It may for some patients be more important than hospitalizations and predicted mortality. Changes in HRQOL also have face validity, but here the magnitude of change is important. Small transient decreases in a good baseline HRQOL are generally acceptable during up-titration of GDMT for mild HF, but further decline from a poor baseline during up-titration in advanced disease may be unacceptable to patients.
We should not expect close coupling between change in PROs at 3 months and future outcomes related to disease progression, which will be favorably influenced by ongoing attention to GDMT. Although early responses may be seen for cardiac resynchronization and sometimes hydralazine or nitrates, the overall benefit of GDMT to improve HRQOL will be greatest over the long term.
What Qualities of Care Can Be Tracked for PROs?
Should changes in PROs be used to measure provider performance? Most patients in the present study had changes of ≥5 points during the 3 months, but few of these followed medication changes. Quality of life with HF is influenced by many factors, including other medical diagnoses and psychosocial factors (7). Even in the longer term, risk adjustments will never be sufficient to capture all the components of decision making and outcomes for GDMT and will capture even fewer drivers of HRQOL.
Consensus is growing that symptoms and quality of life as reported by patients remain vital topics for the clinical encounter. PROs provide opportunity for both patients and providers to prioritize the limited time in face-to-face visits. When asked about their hopes for PROs, patients reported that they “want to be heard.” A reasonable expectation may be that physicians consider and discuss PROs before the end of the visit, rather than just clicking “mark as reviewed.”
The Future of PROs
Optimal care for HF fills a complex matrix of interventions to relieve symptoms, minimize medication side effects, empower patients to manage their disease, and improve future outcomes through modification of cardiac injury and maladaptive responses. To justify the considerable time asked of both patients and providers, PRO displays will need to prompt actions. This will require parsing symptoms to highlight those that can be relieved and revealing concerns about roles and goals that warrant conversation. Thoughtful responses to PROs and optimization of recommended therapies should be viewed as complementary but distinct aspects of how well we keep our promises to care for patients with symptomatic heart disease.
↵∗ Editorials published in JACC: Heart Failure reflect the views of the authors and do not necessarily represent the views of JACC: Heart Failure or the American College of Cardiology.
Dr. Stevenson has no financial conflicts and provides unreimbursed consultation to Abbott Medical and Biotronik. Dr. Brinkley has reported that he has no relationships relevant to the contents of this paper to disclose.
- 2019 American College of Cardiology Foundation
- ↵Frost, Robert. Stopping by Woods on a Snowy Evening. Poetry Foundation. Available at: https://www.poetryfoundation.org/poems/42891/stopping-by-woods-on-a-snowy-evening. Accessed May 13, 2019.
- Rumsfeld J.S.,
- Alexander K.P.,
- Goff D.C.,
- et al.
- Kelkar A.A.,
- Spertus J.,
- Pang P.,
- et al.
- Thomas M.,
- Khariton Y.,
- Fonarow G.C.,
- et al.
- Stehlik J.,
- Rodriguez-Correa C.,
- Spertus J.A.,
- et al.
- Joyce E.,
- Chung C.,
- Badloe S.,
- et al.