Abstract
Background:
Clinicians typically estimate heart failure (HF) health status using the New York Heart Association (NYHA) class, which is often discordant with patient-reported health status. It is unknown if collecting patient-reported health status improves the accuracy of clinician assessments.
Methods:
The Patient-Reported Outcomes in Heart Failure Clinic (PRO-HF) trial is a randomized, non-blinded trial evaluating routine Kansas City Cardiomyopathy Questionnaire-12 (KCCQ-12) collection in HF clinic. Patients with a scheduled visit to Stanford HF clinic between August 30, 2021, and June 30, 2022 were enrolled and randomized to KCCQ-12 assessment or usual care. In this prespecified sub-study, we evaluated whether access to the KCCQ-12 improved the accuracy of clinicians’ NYHA assessment or patients’ perspectives on their clinician interaction. We surveyed clinicians regarding their patients’ NYHA class, quality of life, and symptom frequency. Clinician responses were compared with patients’ KCCQ-12 responses. We surveyed patients regarding their clinician interactions.
Results:
Of the 1,248 enrolled patients, 1,051 (84.2%) attended a visit during the sub-study. KCCQ-12 results were given to the clinicians treating the 528 patients in the KCCQ-12 arm; the 523 patients in the usual care arm completed the KCCQ-12 without the results being shared. The correlation between NYHA class and KCCQ-12 Overall Summary Score was stronger when clinicians had access to the KCCQ-12 (r=−0.73 vs. r=−0.61, p<0.001). More patients in the KCCQ-12 arm strongly agreed that their clinician understood their symptoms (95.2% vs. 89.7% of respondents; [OR 2.27; 95% CI: 1.32-3.87)]. However, patients in both arms reported similar quality of clinician communication and therapeutic alliance.
Conclusions:
Collecting the KCCQ-12 in HF clinic improved clinicians’ accuracy of health status assessment; correspondingly, patients believed their clinicians better understood their symptoms.
Registration:
URL: ClinicalTrials.gov; Unique Identifier: NCT04164004
Introduction
In heart failure (HF), the standard approach for summarizing patients’ health status is through the clinician-assigned New York Heart Association (NYHA) class. However, NYHA classification has poor reproducibility across clinicians and is often discordant with patients’ report of their health status.1-3 Patient-reported outcome (PRO) measures have been proposed to improve the patient-centeredness of care.4, 5 Patient-reported health status has been shown to better detect clinically meaningful changes in clinical practice and to better predict subsequent clinical events than NYHA class.6, 7 The reliability and prognostic value of PROs has led to their increasing inclusion as clinical trial endpoints,8-10 and there has been expanding interest in implementing PROs in routine care.4, 5 The 2022 American College of Cardiology/American Heart Association Heart Failure Clinical Practice Guidelines provided further momentum for this change by including a new 2A recommendation to incorporate patient-reported health status into routine heart failure (HF) care.5 However, the uptake of PROs as part of standard HF care has been slow and evidence of their value in routine practice is sparse.
For patients with HF, the Kansas City Cardiomyopathy Questionnaire (KCCQ) is the most widely used patient-reported measure of health status. The KCCQ is a well-validated instrument that summarizes the impact of HF on patients’ symptoms, function, and quality of life.11 An abbreviated version of the KCCQ, the KCCQ-12, was specifically developed for implementation in clinical care.12 While PROs have been shown to improve treatment and survival among patients with cancer, the clinical impact of PRO assessment has not been tested in HF.13, 14
The Patient-Reported Outcome Measurement in Heart Failure Clinic (PRO-HF) trial is a pragmatic randomized trial designed to evaluate whether the routine clinical assessment of the KCCQ-12 in HF clinic can increase the accuracy of clinicians’ assessments of their patients’ health status, improve patient experience, alter treatment, and impact patients’ 1-year health status. Herein, we present the findings of a prespecified sub-study investigating the effect of routine KCCQ-12 collection on the first of these goals: the accuracy of clinicians’ assessments of patients’ health status and patient experience.
Methods
Trial Design and Oversight
The PRO-HF (Patient-Reported Outcome Measurement in Heart Failure Clinic) trial was a pragmatic, single-center randomized clinical trial of routine KCCQ-12 assessment versus usual care. Details regarding the rationale and methods are described elsewhere.15 To protect participants’ personal health information, data for this study are available from the corresponding author only on reasonable request. The trial was approved by the Stanford Institutional Review Board and registered on ClinicalTrials.gov (NCT04164004).
The study team developed the protocol (available on clinicaltrials.gov) and oversaw the statistical analysis, patient recruitment, and data handling procedures. An independent safety monitor was notified of any safety concerns or adverse events related to KCCQ-12 assessment. All authors assume full responsibility for the accuracy and completeness of the analyses.
Participants
Adult patients with a scheduled visit (in-person or telehealth) in the Stanford HF clinic during the enrollment period (August 30, 2021 - June 30, 2022) were eligible. The Stanford HF clinic is a HF specialty clinic at a tertiary referral hospital. The clinic is staffed by advanced heart failure cardiologists and advanced practice providers. The clinic manages patients across the HF spectrum: asymptomatic patients with genetic cardiomyopathies, elderly patients with heart failure with preserved ejection fraction, and patients undergoing advanced HF therapy evaluation. Patients were included whether or not they had a HF diagnosis for two reasons. First, the trial enrolled patients and collected KCCQ-12 results pre-visit and some were invited to participate prior to diagnostic confirmation of HF. Second, this pragmatic trial aimed to represent future implementation efforts, which will likely be focused on all patients presenting to a HF specialty clinic, rather than requiring additional screening steps for whom should or should not get a survey. Patients were excluded, at the discretion of their treating cardiologist, if they were enrolled in alternate clinical trials focusing on patient-reported health status outcomes. There were no other clinical exclusions.
Patient-reported Health Status Assessment – The KCCQ-12
The KCCQ-12 is an abbreviated version of the original KCCQ. The KCCQ-12 is designed for use in clinical care;11 it includes 12 questions capturing the effect of HF on four domains—symptom frequency, physical limitations, social limitations, and quality of life—over the preceding 2 weeks. Individual responses are converted into those four domain scores, which each range from 0 (more severe symptoms/limitations/very poor quality of life) to 100 (no symptoms/limitations/excellent quality of life). The average of the four domain scores calculates the Overall Summary Score (OSS). The average of the symptom frequency, physical limitations, and social limitations domains is the Clinical Summary Score.
Enrollment and Randomization
Eligible patients were contacted via email approximately one week before their visit. The email included an online consent form with information regarding the trial and contact information for further inquiries. Patients electing to participate provided informed consent through a standardized online form or via telephone. They were then automatically randomized, using a REDCap electronic data capture tool and a randomly generated allocation sequence, to either routine KCCQ-12 assessment or usual care.16 Randomization was stratified by their treating clinician in the upcoming clinic visit with block sizes of 2 and 4. For patients with shared visits with both an advanced heart failure physician and an advanced practice provider or cardiology fellow, the stratification was based on the attending physician. Patients who declined participation received no further contact from the study team. A subset of patients without an email response were contacted by telephone and/or text message several days before the visit. Given the nature of the intervention, neither patients nor clinicians were blinded. Stanford HF clinicians also consented for this study.
Comparator Arms
Participants randomized to the KCCQ-12 assessment arm were instructed to complete the KCCQ-12 prior to each clinic visit over the course of 12 months. The KCCQ-12 was collected via the electronic health record (EHR) patient portal 3 days prior to their visit. Patients who did not complete the KCCQ-12 via the patient portal were called by the research team or completed the KCCQ-12 during clinic check-in. Each visit, KCCQ-12 results were made available to the treating clinician. KCCQ-12 results, both the question responses and summary scores, were available in the EHR in both tabular and graphical form (for repeat assessments). The KCCQ-12 questions and summary scores could also be embedded within clinic notes using EHR dotphrases. In addition, the KCCQ-12 was printed out for in-person visits and sent via email to the clinician for virtual visits. The KCCQ-12 visualizations are displayed with the detailed trial methods.15
Patients randomized to the usual care arm were prompted to complete the KCCQ-12 at the time of enrollment, with the median interval being 6 days before their post-randomization clinic visit. Patients who did not complete the KCCQ-12 online during enrollment were called by the research team. These results were not visible to their treating clinicians. In both arms, treating clinicians were allowed to make all diagnostic and treatment decisions at their discretion.
Clinicians and clinic staff received education regarding the interpretation and potential utility of the KCCQ-12 score. This included presentations during group educational sessions and individual outreach from the principal investigator. Additionally, informational sheets regarding KCCQ-12 interpretation were posted throughout the HF clinic. Details regarding the education are available elsewhere.15
Outcomes
In this prespecified sub-study, we compared differences in the accuracy of clinician assessment of health status and patients’ experiences when KCCQ-12 scores were and were not available in clinic. We evaluated these outcomes using two separate surveys that were administered to clinicians and patients following their first post-randomization clinic visit between October 1, 2021, and June 30, 2022. We did not include clinic visits during the first month of the trial, so clinicians could have an opportunity to better understand the KCCQ-12 and its implementation in their practice.
To evaluate clinician health status assessment, we provided an 8-question paper survey to the treating clinicians following a patient’s first post-randomization clinic visit (Supplement Figure S1). This survey queried clinicians’ assessments of the following for each patient: (1) the New York Heart Association functional class, (2) the impact of HF on their patients’ quality of life, (3) HF symptom frequency (4) questions for edema, orthopnea, fatigue, and shortness of breath, (5) their expected health status trajectory, and (6) the primary factor impacting their quality of life. We mapped each response to a patient-reported answer on the KCCQ-12 or the patient survey and categorized each response as “concordant,” “discordant,” or “intermediate” (Supplement Table S1). For example, we considered NYHA Class III assessment by the clinician concordant with a KCCQ-OSS of ≥30 and <60, discordant with a KCCQ-OSS ≥70 or <20, and intermediate for values in between these ranges. As a secondary analysis, we mapped the NYHA classification to alternate domains—rather than the OSS—of the KCCQ-12.
Patients’ experiences of their interaction with their treating clinician were evaluated using a 10-question survey (Supplement Figure S2) that was emailed to participants after their first clinic visit. Those who did not respond to the patient experience survey via email were contacted by text message and phone call. The survey included a Likert scale with 5 levels of agreement for statements related to clinician-patient alliance and quality of communication.
Statistical Analysis
Baseline patient characteristics are described using median and interquartile range (IQR) for continuous variables. We compared patient characteristics between those included in this analysis and the overall trial population using standardized mean differences, Cohen’s d for continuous variables and Cramer’s V for categorical variables. Standardized mean differences (SMD) 0.2-0.5 were considered small, values of >0.5-0.8 were considered medium, and values > 0.8 were considered large.17
We assessed the relationship between clinician assessment of health status and patient-reported health status via multiple approaches. First, for clinician NYHA classification and quality of life assessment, we calculated the rank correlation. For NYHA classification, we calculated the correlation with the KCCQ-OSS; for the clinician quality-of-life assessment, we evaluated the correlation with the KCCQ-12 quality-of-life score. This analysis evaluated whether patients with lower KCCQ-OSS (lower patient-reported health status) were consistently classified as higher NYHA class. We compared the correlation across treatment arms using Fisher’s transformation.17 We also evaluated the correlation between NYHA classification and the KCCQ-12 Clinical Summary Score, which excludes the Quality of Life domain, and with individual KCCQ-12 domains. Second, within each NYHA classification, we compared the distribution of the KCCQ-OSS for each treatment arm to evaluate whether there was a difference in the consistency of NYHA classification. Within each NYHA class, lower dispersion of the KCCQ-OSS – measured as a smaller interquartile range or standard deviation - indicates more consistent NYHA classifications. We similarly evaluated the dispersion of the KCCQ-Quality of Life domain stratified by the clinician’s assessment of quality of life. Third, each of the 8 clinician assessment questions were classified as concordant, intermediate, and discordant. We compared the frequencies of concordant, intermediate, and discordant assessments using mixed-effects binomial logistic regression models while adjusting for the treating clinician as a random intercept. We computed odds ratios for concordant vs. intermediate/discordant and concordant/intermediate vs. discordant across trial arms.
For the patient experience survey responses, we conducted a nonparametric analysis using Wilcoxon rank-sum test to compare the level of agreement for each patient experience statement between trial arms. We also compared these responses across treatment arms using mixed-effects binomial logistic regression with a random intercept for the treating clinician. We computed odds ratios for “strongly agree” vs. all other options, “strongly/somewhat agree” vs. all other options, and “strongly/somewhat agree/not disagree” vs. “strongly/somewhat disagree”.
Prespecified subgroup analyses were performed according to age, sex, left ventricular ejection fraction (LVEF), Charlson Comorbidity Index, and baseline KCCQ-OSS. Continuous covariates were stratified at their median values to form subgroups. Modification of the intervention effect was assessed by adding a treatment arm × subgroup interaction term to the mixed-effects binomial logistic regression models. For both clinician assessment of NYHA and quality of life, we described the concordance classifications stratified by clinician and study arm.
Multiple sensitivity analyses were performed around missing data (Supplement Table S2). First, we used multiple imputation by chained equations to impute data for all patients missing responses to the KCCQ-12, the clinician assessment, or the patient experience, and for patients who did not attend a post-randomization clinic visit. Second, we imputed scenario analyses with all missing responses positive (concordant for clinician assessment and “strongly agree” for patient experience) or negative (discordant for clinician assessment and “strongly disagree” for patient experience). Details regarding the imputation models are provided in the Supplement Methods. We performed three additional sensitivity analyses. First, we excluded patients from the 3 clinicians with the highest missingness. Second, we excluded patients without a history of HF or cardiomyopathy. Third, we repeated our analyses with adjustment for prespecified baseline data: age, LVEF, and Charlson comorbidity index. For all analyses, the significance threshold was 0.05, and testing was 2-sided. Analyses were performed with Stata version 15.1 (StataCorp LLC).
Results
There were 5,133 eligible patients with scheduled visits to Stanford HF clinic between August 30, 2021, and June 30, 2022. Among this cohort, 1,362 (24.7%) declined enrollment and 1,248 (22.7%) were enrolled in this trial, with the remainder not responding to outreach. The trial enrolled 624 patients in the KCCQ-12 arm and 624 in the usual care arm (Figure 1). Median age was 64 years. 39% were women. 5% were Black; 12% were Asian; and 8% were Hispanic (Table 1). 87% had a prior diagnosis of HF or cardiomyopathy or had an EF<50%. Of the 1,245 with an LVEF available, 28% had an LVEF≤40%, 19% with an LVEF 41-50%, and 53% with LVEF>50%. The median KCCQ-OSS was 82 (IQR: 58-95). The patients were distributed across 17 clinicians: 13 were advanced HF-trained physicians and 4 were advanced practice providers.
Figure 1. CONSORT Diagram.
Abbreviations: HF: heart failure; KCCQ-12: Kansas City Cardiomyopathy Questionnaire-12.
Table 1.
Patient Characteristics*
| Total | KCCQ-12 Arm | Usual Care Arm | |
|---|---|---|---|
| N=1,248§ | N=624 | N=624 | |
| Demographics | |||
| Age, years | 63.9 (51.8-72.8) | 64.1 (52.2-73.0) | 63.7 (51.4-72.4) |
| Female Sex | 485 (38.9%) | 248 (39.7%) | 237 (38.0%) |
| Race | |||
| American Indian or Alaska Native | 7 (0.6%) | 4 (0.6%) | 3 (0.5%) |
| Asian | 143 (11.5%) | 68 (10.9%) | 75 (12.0%) |
| Black | 57 (4.6%) | 30 (4.8%) | 27 (4.3%) |
| Pacific Islander | 15 (1.2%) | 6 (1.0%) | 9 (1.4%) |
| Unknown | 216 (17.3%) | 104 (16.7%) | 112 (17.9%) |
| White | 810 (64.9%) | 412 (66.0%) | 398 (63.8%) |
| Ethnicity | |||
| Hispanic/Latino | 101 (8.1%) | 45 (7.2%) | 56 (9.0%) |
| Non-Hispanic | 1,077 (86.3%) | 551 (88.3%) | 526 (84.3%) |
| Unknown | 70 (5.6%) | 28 (4.5%) | 42 (6.7%) |
| Baseline HF Characteristics † | |||
| Heart Failure or Cardiomyopathy Diagnosis | 1,089 (87.3%) | 542 (86.9%) | 547 (87.7%) |
| Prior HF Clinic Encounter | 1,046 (83.8%) | 520 (83.3%) | 526 (84.3%) |
| Left ventricular ejection fraction | 52 (39-60) | 51 (39-60) | 53 (38-60) |
| Left ventricular ejection fraction ≤40% | 348 (27.9%) | 173 (27.7%) | 175 (28.0%) |
| Left ventricular ejection fraction 41-50% | 230 (18.4%) | 126 (20.2%) | 104 (16.7%) |
| Left ventricular ejection fraction >50% | 667 (53.4%) | 322 (51.6%) | 345 (55.3%) |
| Missing left ventricular ejection fraction | 3 (0.2%) | 3 (0.5%) | 0 (0.0%) |
| KCCQ-12 Overall Summary Score | 82 (58-95) | 82 (57-96) | 82 (58-95) |
| KCCQ-12 Physical Limitation Score | 83 (58-100) | 83 (58-100) | 83 (58-100) |
| KCCQ-12 Symptom Frequency Score | 88 (67-100) | 89 (67-100) | 88 (65-100) |
| KCCQ-12 Quality of Life Score | 75 (50-100) | 75 (50-88) | 75 (50-100) |
| KCCQ-12 Social Limitations Score | 83 (58-100) | 83 (50-100) | 83 (58-100) |
| Comorbidities, % ‡ | |||
| Atrial Fibrillation/Atrial Flutter | 437 (35.0%) | 229 (36.7%) | 208 (33.3%) |
| Coronary Artery Disease | 500 (40.1%) | 258 (41.3%) | 242 (38.8%) |
| Cancer | 169 (13.5%) | 89 (14.3%) | 80 (12.8%) |
| Chronic Kidney Disease | 276 (22.1%) | 143 (22.9%) | 133 (21.3%) |
| Chronic Obstructive Pulmonary Disease | 176 (14.1%) | 87 (13.9%) | 89 (14.3%) |
| Depression | 159 (12.7%) | 86 (13.8%) | 73 (11.7%) |
| Diabetes Mellitus | 235 (18.8%) | 116 (18.6%) | 119 (19.1%) |
| Hypertension | 658 (52.7%) | 329 (52.7%) | 329 (52.7%) |
| Peripheral Vascular Disease | 481 (38.5%) | 243 (38.9%) | 238 (38.1%) |
| Vitals and Labs § | |||
| Systolic Blood Pressure, mmHg | 120 (109-132) | 121 (109-132) | 120 (109-133) |
| Diastolic Blood Pressure, mmHg | 69 (61-77) | 70 (62-78) | 69 (61-76) |
| Heart Rate, bpm | 72 (64-82) | 72 (63-82) | 73 (64-83) |
| Body Mass Index, kg/m2 | 27 (24-31) | 27 (24-31) | 27 (24-31) |
| Creatinine, mmol/L | 1.0 (0.8-1.2) | 1.0 (0.8-1.2) | 1.0 (0.8-1.2) |
| Potassium, mEq/dL | 4.4 (4.1-4.7) | 4.4 (4.1-4.7) | 4.4 (4.1-4.7) |
| Sodium, mmol/L | 139 (137-140) | 139 (137-140) | 139 (137-140) |
| Baseline Medication Therapies, % | |||
| ACEI/ARB/ARNI | 499 (40.0%) | 245 (39.3%) | 254 (40.7%) |
| Beta-blocker | 823 (65.9%) | 410 (65.7%) | 413 (66.2%) |
| Loop Diuretics | 352 (28.2%) | 174 (27.9%) | 178 (28.5%) |
| MRA | 472 (37.8%) | 223 (35.7%) | 249 (39.9%) |
| SGLT2I | 115 (9.2%) | 62 (9.9%) | 53 (8.5%) |
Abbreviations: ACEI: angiotensin-converting-enzyme inhibitor; ARB: angiotensin receptor blocker; ARNI: angiotensin receptor-neprilysin inhibitor; MRA: mineralocorticoid receptor antagonist; SGLT2I: sodium/glucose cotransporter-2 inhibitor.
Continuous variables displayed as median (interquartile range); binary variables displayed as outcome (percentage).
Patients were managed by one of 13 advanced HF physicians or one of 4 advanced practice providers.
Comorbidities based on electronic health record diagnoses within 2 years preceding index visit (detailed in Supplemental Methods).
Vitals and laboratory values based on most recent values within 2 years preceding index visit. Missing frequency detailed in Supplement Table 2.
Among this cohort, 1,221 patients (97.8%) had at least 1 HF clinic encounter between randomization and June 30, 2022. 1,051 participants (84.2% of total) completed their first post-randomization clinic visit between October 1, 2021, and June 30, 2022, and were included in this analysis. There were 528 in the KCCQ-12 arm and 523 in the usual care arm. Patient characteristics were generally similar among those in this analysis and the overall PRO-HF population except for KCCQ-12 scores being higher among those in this analysis (Supplement Table S3).
Clinician Assessment of NYHA
The clinician survey of patient health status was completed for 479 of 528 patients (90.7%) in the KCCQ-12 arm and 467 of 523 patients (89.3%) in the usual care arm (p=0.44). Completion rates varied substantially across clinicians; the overall completion rate was 95.6% excluding the three clinicians with lowest survey completion rates. Baseline KCCQ-12 completion rates were high for both arms (517/528 [97.9%] for KCCQ-12 vs. 510/523 [97.5%] for usual care, p=0.66).
There was a stronger correlation between clinician NYHA classification and patient-reported KCCQ-OSS in the KCCQ-12 arm than in the usual care arm (r=−0.73 [n=474] vs. −0.61 [n=457], p<0.001 for difference in correlations). After excluding patients without a HF/cardiomyopathy diagnosis, there remained a stronger correlation between NYHA class and KCCQ-OSS when clinicians had access to the KCCQ-12 (r=−0.73 [n=413] vs. r=−0.61 [n=398], p=0.002).
Table 2 demonstrates concordance was more frequent in the KCCQ-12 arm. The results were similar after excluding patients without a HF/cardiomyopathy diagnosis (Supplement Table S4). Figure 2 displays the KCCQ-OSS for patients in both arms stratified by their NYHA classification, with green indicating concordance, red discordance, and white intermediate (displayed as a histogram in Supplement Figure S3). For any given NYHA class, the interquartile range of the KCCQ-OSS is smaller when clinicians have access to the KCCQ-OSS, indicating a decrease in the variability of KCCQ-OSS within each NYHA classification (Figure 3, Supplement Table S5). A greater decrease in variability of clinician assessment was noted among patients classified as NYHA Class III/IV HF than those classified as NYHA Class I/II.
Table 2.
Concordance Between Clinician Health Status Assessment and Patient Reporting
| KCCQ-12 Arm | Usual Care Arm |
Concordant OR† (95% CI) |
Concordant/ Intermediate OR† (95% CI) |
|
|---|---|---|---|---|
| n=528* | n=523* | |||
| NYHA and KCCQ-OSS | 1.47 (1.12-1.91) | 1.74 (1.28-2.38) | ||
| Concordant | 296 (56.1%) | 244 (46.7%) | ||
| Intermediate | 91 (17.2%) | 85 (16.3%) | ||
| Discordant | 87 (16.5%) | 128 (24.5%) | ||
| Missing | 54 (10.2%) | 66 (12.6%) | ||
| Quality of Life | 1.34 (1.04-1.74) | 2.17 (1.47-3.21) | ||
| Concordant | 244 (46.2%) | 202 (38.6%) | ||
| Intermediate | 186 (35.2%) | 172 (32.9%) | ||
| Discordant | 44 (8.3%) | 83 (15.9%) | ||
| Missing | 54 (10.2%) | 66 (12.6%) | ||
| Edema Frequency | 1.71 (1.27-2.31) | 1.84 (1.23-2.76) | ||
| Concordant | 375 (71.0%) | 315 (60.2%) | ||
| Intermediate | 56 (10.6%) | 71 (13.6%) | ||
| Discordant | 43 (8.1%) | 71 (13.6%) | ||
| Missing | 54 (10.2%) | 66 (12.6%) | ||
| Orthopnea Frequency | 1.52 (1.07-2.16) | 1.78 (1.11-2.85) | ||
| Concordant | 406 (76.9%) | 364 (69.6%) | ||
| Intermediate | 38 (7.2%) | 44 (8.4%) | ||
| Discordant | 30 (5.7%) | 49 (9.4%) | ||
| Missing | 54 (10.2%) | 66 (12.6%) | ||
| Fatigue Frequency | 1.83 (1.41-2.38) | 1.64 (1.22-2.19) | ||
| Concordant | 290 (54.9%) | 212 (40.5%) | ||
| Intermediate | 74 (14.0%) | 94 (18.0%) | ||
| Discordant | 110 (20.8%) | 151 (28.9%) | ||
| Missing | 54 (10.2%) | 66 (12.6%) | ||
| Shortness of Breath Frequency | 1.36 (1.05-1.77) | 1.49 (1.11-2.02) | ||
| Concordant | 301 (57.0%) | 256 (48.9%) | ||
| Intermediate | 74 (14.0%) | 71 (13.6%) | ||
| Discordant | 99 (18.8%) | 129 (24.7%) | ||
| Missing | 54 (10.2%) | 67 (12.8%) | ||
| Expected Health Status Stability Over Next Year | 1.51 (1.14-1.99) | 1.54 (1.10-2.15) | ||
| Concordant | 212 (40.2%) | 173 (33.1%) | ||
| Intermediate | 118 (22.3%) | 131 (25.0%) | ||
| Discordant | 75 (14.2%) | 105 (20.1%) | ||
| Missing | 123 (23.3%) | 114 (21.8%) | ||
| Biggest Factors Impacting Quality of Life | 1.00 (0.75-1.32) | 1.44 (1.07-1.95) | ||
| Concordant | 162 (30.7%) | 163 (31.2%) | ||
| Intermediate | 137 (25.9%) | 107 (20.5%) | ||
| Discordant | 106 (20.1%) | 138 (26.4%) | ||
| Missing | 123 (23.3%) | 115 (22.0%) | ||
Abbreviations: NYHA: New York Heart Association; KCCQ-OSS: Kansas City Cardiomyopathy Questionnaire- Overall Summary Score. OR: Odds Ratio.
Number of patients based on patients in the sub-study with a first post-randomization clinic visit between October 1, 2021, and June 30, 2022.
Based on mixed-effects binomial logistic regression models with a random intercept for treating clinician.
Figure 2. Clinician NYHA Assessment and Patient-reported KCCQ-12 Overall Summary Score.
Figure 2 displays the concordance between clinician assessment of health status via the New York Heart Association (NYHA) class and patient-reported health status based on the KCCQ-12 Overall Summary Score (OSS) for each treatment arm. For the KCCQ-OSS, a higher score indicates better HF health status. Patients are divided based on clinician NYHA Class. The distribution of KCCQ-OSS is plotted via kernel density plots. The shaded green area represents concordance between patient and clinician assessment and the shaded red area represents discordance. Intermediate areas are white. Supplement Figure S3 displays a histogram version of this plot with patient counts.
Figure 3. Variation in KCCQ-12 Scores Stratified by Clinician Assessment: KCCQ-OSS and NYHA Class and KCCQ-12 Quality of Life Score and Clinician Quality of Life Assessment.
The upper panel displays the distribution of Kansas City Cardiomyopathy Questionnaire-12 Overall Summary Score (KCCQ-OSS) scores stratified by clinician New York Heart Association (NYHA) assessment for each treatment arm as a box and whiskers plot. The dark horizontal line is the median and the box is the interquartile range (IQR). The vertical lines extending from the box (“the whiskers”) represents values outside of the IQR within the IQR*1.5. Outliers are displayed as individual dots. The lower panel displays the distribution of Kanas City Cardiomyopathy Questionnaire-12 (KCCQ-12) Quality of Life domain scores stratified by the clinician’s assessment of HF’s impact on patient quality of life. Supplement Table S5 evaluates the distributions using the mean and standard deviation for each group.
We performed subgroup analyses by stratifying the cohort based on median age, sex, Charlson comorbidity index, LVEF, and KCCQ-OSS. None of these variables significantly modified the association between treatment arm and the concordance between NYHA and KCCQ-OSS (Figure 4, Supplement Table S6). The magnitude of improvement in concordance between clinician NYHA class and patient-reported KCCQ-OSS varied across clinicians (Supplement Figure S4).
Figure 4. Subgroup Analysis of Clinician and Patient Concordance on Health Status and Quality of Life Assessments.
Abbreviations: Charlson: Charlson Comorbidity Index score; EF: ejection fraction; KCCQ-OSS: Kansas City Cardiomyopathy Questionnaire-12 Overall Summary Score. The listed p-values are tests for heterogeneity of treatment effect based on the interaction between the subgroup identifier and the treatment arm in a mixed effects binomial logistic regression model. Subgroup results for each clinician question listed in Supplement Table S6.
We performed multiple sensitivity analyses. Our results were similar after excluding patients from the 3 clinicians with lowest survey response rates (Supplement Table S4). Our results remained consistent across each of the three imputation models (Supplement Table S7), adjustment for prespecified variables (Supplement Table S8), and alternate KCCQ-OSS and NYHA mapping (Supplement Table S9). Finally, we found the KCCQ-12 arm had a stronger correlation between NYHA and individual domains of the KCCQ-12 than the usual care arm (Supplement Table S10).
Clinician Assessment of Quality of Life, Symptom Frequency, and Health Status Trajectory
There was also a stronger correlation between clinician quality-of-life classification and the patient-reported KCCQ-12 Quality of Life domain score in the KCCQ-12 arm than in the usual care arm (r=−0.69 vs. −0.52, p<0.001). Figure 5 illustrates the greater agreement between clinician and patient reporting of quality of life in the KCCQ-12 arm (displayed as a histogram in Supplement Figure S5). Supplement Figure S6 displays the concordance of clinician and patient quality of life assessment stratified by clinician. There was no significant effect modification based on baseline KCCQ-OSS (Supplement Table S6).
Figure 5. Clinician and Patient Assessment of Impact of Heart Failure on Quality of Life (QoL).
Figure 5 displays the concordance between clinician assessment of quality of life and patient-reported quality of life based on the Kansas City Cardiomyopathy Questionnaire-12 (KCCQ-12) Quality of Life score for each treatment arm. With the KCCQ-12 Quality of Life score, a higher number indicates less of an impact on quality of life secondary to heart failure. Patients are divided based on clinician New York Heart Association (NYHA) Class. The distribution of KCCQ-12 Quality of Life score is plotted via kernel density plots. The shaded green area represents concordance between patient and clinician assessment and the shaded red area represents discordance. Intermediate areas are white. Supplement Figure S5 displays a histogram version of this plot with patient counts.
Table 2 displays the levels of concordance for clinician and patient reporting of symptom frequency, health status trajectory, and the biggest factor impacting patient quality of life. The KCCQ-12 arm had higher levels of concordance between clinician and patient assessment of frequency of each of the four assessed symptoms: edema, orthopnea, fatigue, and shortness of breath. The KCCQ-12 arm also had increased odds of greater agreement between the clinician and patient regarding expected health status trajectory (OR for concordance: 1.51 [95% CI: 1.14-1.99]). Both arms had similar levels of agreement between patient and clinician identification of the biggest factor impacting quality of life. These analyses were generally consistent across subgroups (Supplement Table S6) and with imputation for missing data (Supplement Table S7).
Patient Experience
Survey completion was similar across the KCCQ-12 and usual care arms (436/528 [82.6%] vs. 438/523 [83.7%], p=0.61). Overall, patients noted high agreement with the stated questions (Table 3). A higher proportion of patients in the KCCQ-12 arm had strong agreement with the statements “My clinician understood symptoms related to my heart” (95.2% [415/436] vs. 89.7% [393/438] of respondents [OR 2.27 of strong agreement; 95% CI: 1.32-3.87]) and “My clinician and I agreed on how I was doing overall” (93.3% [407/436] vs. 86.5% [379/438] of respondents [OR 2.18; 95% CI: 1.37-3.48]) compared with the usual care arm (Supplement Table S11). These effect sizes were attenuated when evaluating the odds of “Strongly/Somewhat Agree” across patient experience statements (Supplement Table S11). Additionally, the difference in patient experience across arms did not remain significant across imputation approaches (Supplement Table S12). There were no significant differences for the other 6 questions related to clinician-patient alliance and quality of communication. There was no evidence of heterogeneity across subgroups (Supplement Table S13).
Table 3.
Results of Patient Experience Survey
| KCCQ-12 Arm | Usual Care Arm | p-value† | |
|---|---|---|---|
| n=528* | n=523* | ||
| "My clinician listened to me carefully" | |||
| Strongly Agree | 412 (78.0%) | 400 (76.5%) | 0.072 |
| Somewhat Agree | 11 (2.1%) | 21 (4.0%) | |
| Neither Agree Nor Disagree | 4 (0.8%) | 3 (0.6%) | |
| Somewhat Disagree | 1 (0.2%) | 5 (1.0%) | |
| Strongly Disagree | 8 (1.5%) | 9 (1.7%) | |
| Missing | 92 (17.4%) | 85 (16.3%) | |
| "My clinician understood symptoms related to my heart" | |||
| Strongly Agree | 415 (78.6%) | 393 (75.1%) | 0.002 |
| Somewhat Agree | 10 (1.9%) | 21 (4.0%) | |
| Neither Agree Nor Disagree | 3 (0.6%) | 12 (2.3%) | |
| Somewhat Disagree | 2 (0.4%) | 4 (0.8%) | |
| Strongly Disagree | 6 (1.1%) | 8 (1.5%) | |
| Missing | 92 (17.4%) | 85 (16.3%) | |
| "My clinician showed respect for what I had to say" | |||
| Strongly Agree | 416 (78.8%) | 411 (78.6%) | 0.364 |
| Somewhat Agree | 8 (1.5%) | 11 (2.1%) | |
| Neither Agree Nor Disagree | 6 (1.1%) | 3 (0.6%) | |
| Somewhat Disagree | 0 (0.0%) | 3 (0.6%) | |
| Strongly Disagree | 6 (1.1%) | 9 (1.7%) | |
| Missing | 92 (17.4%) | 86 (16.4%) | |
| "My clinician and I agreed on how I was doing overall" | |||
| Strongly Agree | 407 (77.1%) | 379 (72.5%) | 0.001 |
| Somewhat Agree | 16 (3.0%) | 37 (7.1%) | |
| Neither Agree Nor Disagree | 6 (1.1%) | 9 (1.7%) | |
| Somewhat Disagree | 2 (0.4%) | 6 (1.1%) | |
| Strongly Disagree | 5 (0.9%) | 7 (1.3%) | |
| Missing | 92 (17.4%) | 85 (16.3%) | |
| "My clinician spent enough time with me" | |||
| Strongly Agree | 403 (76.3%) | 401 (76.7%) | 0.607 |
| Somewhat Agree | 22 (4.2%) | 21 (4.0%) | |
| Neither Agree Nor Disagree | 0 (0.0%) | 2 (0.4%) | |
| Somewhat Disagree | 6 (1.1%) | 5 (1.0%) | |
| Strongly Disagree | 5 (0.9%) | 9 (1.7%) | |
| Missing | 92 (17.4%) | 85 (16.3%) | |
| "My clinician explained things in a way that was easy to understand" | |||
| Strongly Agree | 409 (77.5%) | 401 (76.7%) | 0.201 |
| Somewhat Agree | 14 (2.7%) | 21 (4.0%) | |
| Neither Agree Nor Disagree | 6 (1.1%) | 4 (0.8%) | |
| Somewhat Disagree | 2 (0.4%) | 2 (0.4%) | |
| Strongly Disagree | 5 (0.9%) | 10 (1.9%) | |
| Missing | 92 (17.4%) | 85 (16.3%) | |
| "I understand the importance of taking the treatments recommended by my clinician" | |||
| Strongly Agree | 410 (77.7%) | 408 (78.0%) | 0.668 |
| Somewhat Agree | 17 (3.2%) | 17 (3.3%) | |
| Neither Agree Nor Disagree | 3 (0.6%) | 2 (0.4%) | |
| Somewhat Disagree | 0 (0.0%) | 1 (0.2%) | |
| Strongly Disagree | 6 (1.1%) | 9 (1.7%) | |
| Missing | 92 (17.4%) | 86 (16.4%) | |
| "My goals regarding treatment and those of my clinician are aligned" | |||
| Strongly Agree | 404 (76.5%) | 399 (76.3%) | 0.390 |
| Somewhat Agree | 20 (3.8%) | 23 (4.4%) | |
| Neither Agree Nor Disagree | 3 (0.6%) | 3 (0.6%) | |
| Somewhat Disagree | 3 (0.6%) | 5 (1.0%) | |
| Strongly Disagree | 6 (1.1%) | 8 (1.5%) | |
| Missing | 92 (17.4%) | 85 (16.3%) |
Abbreviations: KCCQ-12: Kansas City Cardiomyopathy Questionnaire-12.
Number of patients based on patients in the sub-study with a first post-randomization clinic visit between October 1, 2021, and June 30, 2022.
Based on Wilcoxon rank-sum test. Further details of statistical testing in Supplemental Methods. Odds ratios from mixed-effects binomial logistic regression models listed in Supplement Table S11.
Discussion
For chronic conditions such as HF, a clinician’s assessment of patients’ symptoms, function, and quality of life serves as the foundation for recommending treatment. In this randomized, non-blinded trial, we found that the routine collection and sharing of the KCCQ-12 improved the accuracy of clinicians’ assessments of patients’ health status at a tertiary HF clinic. Additionally, more patients whose physicians had access to their KCCQ-12 scores felt their clinicians understood their symptoms and health status as compared with the usual care arm. These results support the implementation of patient-reported health status as part of routine clinical care for HF to improve the accuracy of patients’ health status assessments and to improve patients’ experiences with care.
Our results support prior work demonstrating the potential of incorporating PROs into routine care. In oncology, Basch and colleagues evaluated weekly electronic monitoring of patient-reported symptoms among patients with metastatic cancer receiving outpatient chemotherapy. They found symptom monitoring led to improved quality of life, decreased frequency of ED visits, and increased survival.13, 14 A key difference with the PRO-HF trial is Basch and colleagues monitored symptoms remotely between clinic visits and alerted nurses regarding severe or worsening symptoms whereas this trial only includes PRO assessment during clinic visits. Evaluating the use of validated patient-reported health status measures within remote monitoring programs for patients with HF is an important area of future research.
In cardiology, the challenge clinicians face in accurately determining health status has also been documented with angina assessment.18 Across 25 outpatient practices, Arnold and colleagues found 42% of patients with angina had more frequent angina than recognized by their clinician.19 As expected, under-recognition of angina was associated with a decreased likelihood of treatment intensification.20 Our results build on this literature by demonstrating that the use of a structured, reproducible patient-reported health status assessment, which is less subject to bias or inter-observer variation, can improve the accuracy of clinician health status assessment.
The impact of a patient’s cardiomyopathy on their health status is critical to understand for clinical decision-making. Multiple HF therapies can substantially improve health status when symptomatic patients are accurately identified.21 This is increasingly important for treatments in which the primary expected benefit is a positive effect on quality of life, as opposed to a treatment indicated for mortality reduction.22-24 Understanding which patients have poor health status may facilitate identifying those most likely to derive substantial benefit from interventions such as cardiac rehabilitation or behavioral health referral. Finally, determining the appropriate timing for advanced HF therapies for ambulatory patients requires accurate detection of worsening health status.25 As the HF armamentarium continues to grow, identifying patients with the most to gain in terms of health status will be important for delivering high-quality efficient care.
There are several reasons why the PRO-HF trial may have underestimated the potential impact of routine patient-reported health status assessment on patient experience and clinician assessment. First, incorporating the KCCQ-12 into routine clinical workflows is a learned behavior that likely requires time and repetition. Clinicians were new to the KCCQ-12 and only had KCCQ-12 results for a small proportion of their patients. Over time, they may be more likely to incorporate the data into treatment decisions. Prior work has demonstrated that patient valuation of patient-reported outcomes is highly dependent on clinician engagement with the data.26 As part of the PRO-HF trial, we are formally evaluating the perspectives of HF clinicians on KCCQ-12 implementation.15 These interviews will further elucidate the potential utility and barriers of patient-reported health status assessment from the clinician perspective. Second, our KCCQ-OSS was higher than that of other HF studies, suggesting a healthier study population.6, 27 This may be explained by our remote enrollment and inclusion of patients seen in the HF clinic without a HF diagnosis, although the results were consistent after excluding patients without HF. The impact of routine KCCQ-12 evaluation may be greater among more symptomatic patients. Third, the KCCQ-12 may be especially valuable at tracking health status changes over time. The current analysis only focuses on health status assessment following a single visit. Finally, we only tested the routine collection of KCCQ-12 during clinic visits. Ambulatory KCCQ-12 measurements may further enhance remote monitoring of HF populations and minimize clinical inertia.
Incorporating patient-reported outcome measures into clinical care may have inherent value by increasing the patient’s voice in their care, independent of the impact on clinical care or outcomes. However, even when there is such a strong theoretical benefit, it remains valuable to formally evaluate the implementation of novel care strategies with quantitative and qualitative assessment. Large-scale implementation of PROs requires substantial effort and may shift resources away from other important healthcare delivery strategies, so understanding the benefits and costs before scaling broadly is important. Moreover, evaluation helps identify opportunities to adapt and improve the effectiveness of the intervention. Several questions regarding PRO implementation would benefit from further evaluation.
While we demonstrated routine assessment of the KCCQ-12 improved clinician assessment, the downstream clinical impact of our findings is unclear. Follow-up data from the PRO-HF trial will better evaluate whether an improved understanding of health status impacts longitudinal health status and treatment patterns. A larger trial incorporating patient-reported health status into routine care will be needed to evaluate the impact on clinical outcomes.
Understanding the generalizability of our results in alternate clinical settings will also be important. We evaluated the KCCQ-12 in a specialized HF center with a modest number of clinicians; the impact of using PROs as part of HF care may differ across clinic setting - general cardiology clinics or telehealth practices – and across clinicians. Alternate clinical settings will likely have less of a primary focus on HF management, less specialized HF training, and differences in patient case mix. In such settings, there may be greater benefit to HF PRO assessment, but effective implementation may also require additional training.
Our trial does not capture all the challenges of a real-world implementation of patient-reported outcomes. We ensured high rates of baseline KCCQ-12 completion with the use of clinical research support staff. Barriers related to collecting the data and sharing it with clinicians have contributed to the slow clinical adoption of PROs.28 Future research should test implementation strategies to ensure PROs can be efficiently collected and integrated into the clinical workflow. It is especially critical to identify approaches to ensure PRO engagement among patients with lower levels of technology literacy and historically disadvantaged groups.
Incorporating the KCCQ-12 into routine care improved patient perception of clinician health status assessment. This is noteworthy not only because it supports our findings that KCCQ-12 assessment improved the accuracy of clinician health status assessment, but also because patient experience is a fundamental dimension of quality of care. However, there are limitations to this finding. The survey has not been previously validated, although the questions capture similar themes as common outpatient experience surveys.29 Additionally, while access to the KCCQ-12 improved patient perception of clinician understanding of their disease, the effect size was modest, and we did not find an impact on the other domains of patient experience that we tested. However, improving patient experience poses a substantial challenge when baseline satisfaction is high as seen in this study.
There are important limitations to our trial. First, there was no blinding of patients, clinicians, or investigators. This may have impacted completion of the surveys. Second, the clinician assessment occurred at the time of the visit; it is unclear if improvements in the accuracy of clinician health status assessment are sustained. Third, there is moderate missing data around patient and clinician assessment. While our patient experience results were sensitive to the imputation approach, our clinician survey results were generally consistent across imputation approach. Fourth, our enrollment strategy may have also selected for younger patients with more digital literacy, likely resulting in higher levels of patient engagement.
In this pragmatic randomized trial, routine assessment of patient-reported health status in HF clinic using the KCCQ-12 led to improved accuracy of clinician assessments of patients’ health status and improved patient perception of clinicians’ understanding of their disease. The downstream impact of patient-reported health status assessment on clinical decision-making and outcomes remains to be seen and should be further investigated.
Supplementary Material
What is new?
In this pragmatic, single-center, randomized clinical trial, routine KCCQ-12 assessments in heart failure clinics improved the accuracy of clinicians’ assessments of patients’ health status.
Patients in the KCCQ-12 arm reported higher levels of understanding by their HF clinicians regarding their symptoms and overall health status compared with usual care.
What are the clinical implications?
Incorporating patient-reported health status assessment into routine clinical care can improve the accuracy of clinician assessments, thereby reducing bias and inter-observer variation.
The routine use of patient-reported health status assessments can improve the quality of care for heart failure by improving patient experience.
Acknowledgments
The authors would like to thank the patients participating in the PRO-HF trial, the study safety monitor, the entire Stanford HF team including the clinicians, nurses, and clinic staff, and Stanford research information technology.
Sources of Funding
The PRO-HF trial is supported by the NHLBI (1K23HL151672-01) and Stanford institutional funding. The data collection is supported by the National Institute of Health (UL1 TR001085).
Non-standard Abbreviations and Acronyms
- EHR
electronic health record
- HF
heart failure
- IQR
interquartile range
- KCCQ-12
Kansas City Cardiomyopathy Questionnaire-12
- KCCQ-OSS
Kansas City Cardiomyopathy Questionnaire-12 Overall Summary Score
- LVEF
left ventricular ejection fraction
- NYHA
New York Heart Association
- SMD
standardized mean difference
- PRO
patient-reported outcome
- PRO-HF trial
Patient-Reported Outcome Measurement in Heart Failure Clinic trial
Footnotes
Publisher's Disclaimer: This article is published in its accepted form, it has not been copyedited and has not appeared in an issue of the journal. Preparation for inclusion in an issue of Circulation: Heart Failure involves copyediting, typesetting, proofreading, and author review, which may lead to differences between this accepted version of the manuscript and the final, published version.
This work was presented as an abstract at the American Heart Association Scientific Sessions, November 5 to November 7, 2022.
Disclosures
Dr. Spertus owns the copyright to the Kansas City Cardiomyopathy Questionnaire. The interpretation and conclusions contained herein are those of the authors and do not necessarily represent the views of the NIH.
References
- 1.Bennett JA, Riegel B, Bittner V and Nichols J. Validity and reliability of the NYHA classes for measuring research outcomes in patients with cardiac disease. Heart Lung. 2002;31:262–70. [DOI] [PubMed] [Google Scholar]
- 2.Raphael C, Briscoe C, Davies J, Ian Whinnett Z, Manisty C, Sutton R, Mayet J and Francis DP. Limitations of the New York Heart Association functional classification system and self-reported walking distances in chronic heart failure. Heart. 2007;93:476–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Tran AT, Chan PS, Jones PG and Spertus JA. Comparison of Patient Self-reported Health Status With Clinician-Assigned New York Heart Association Classification. JAMA Netw Open. 2020;3:e2014319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Heidenreich PA. The Growing Case for Routine Collection of Patient-Reported Outcomes. JAMA Cardiol. 2021;6:497–498. [DOI] [PubMed] [Google Scholar]
- 5.Heidenreich PA, Bozkurt B, Aguilar D, Allen LA, Byun JJ, Colvin MM, Deswal A, Drazner MH, Dunlay SM, Evers LR, Fang JC, Fedson SE, Fonarow GC, Hayek SS, Hernandez AF, Khazanie P, Kittleson MM, Lee CS, Link MS, Milano CA, Nnacheta LC, Sandhu AT, Stevenson LW, Vardeny O, Vest AR and Yancy CW. 2022 AHA/ACC/HFSA Guideline for the Management of Heart Failure: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation. 2022;145:e895–e1032. [DOI] [PubMed] [Google Scholar]
- 6.Greene SJ, Butler J, Spertus JA, Hellkamp AS, Vaduganathan M, DeVore AD, Albert NM, Duffy CI, Patterson JH, Thomas L, Williams FB, Hernandez AF and Fonarow GC. Comparison of New York Heart Association Class and Patient-Reported Outcomes for Heart Failure With Reduced Ejection Fraction. JAMA Cardiol. 2021;6:522–531. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hawwa N, Vest AR, Kumar R, Lahoud R, Young JB, Wu Y, Gorodeski EZ and Cho L. Comparison Between the Kansas City Cardiomyopathy Questionnaire and New York Heart Association in Assessing Functional Capacity and Clinical Outcomes. J Card Fail. 2017;23:280–285. [DOI] [PubMed] [Google Scholar]
- 8.Anker SD, Agewall S, Borggrefe M, Calvert M, Jaime Caro J, Cowie MR, Ford I, Paty JA, Riley JP, Swedberg K, Tavazzi L, Wiklund I and Kirchhof P. The importance of patient-reported outcomes: a call for their comprehensive integration in cardiovascular clinical trials. Eur Heart J. 2014;35:2001–9. [DOI] [PubMed] [Google Scholar]
- 9.Spertus JA, Birmingham MC, Nassif M, Damaraju CV, Abbate A, Butler J, Lanfear DE, Lingvay I, Kosiborod MN and Januzzi JL. The SGLT2 inhibitor canagliflozin in heart failure: the CHIEF-HF remote, patient-centered randomized trial. Nat Med. 2022;28:809–813. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Food and Drug Administration. Guidance for Industry Patient-Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims. December 2009. Accessed September 10, 2022. https://www.fda.gov/media/77832/download.
- 11.Spertus JA, Jones PG, Sandhu AT and Arnold SV. Interpreting the Kansas City Cardiomyopathy Questionnaire in Clinical Trials and Clinical Care: JACC State-of-the-Art Review. J Am Coll Cardiol. 2020;76:2379–2390. [DOI] [PubMed] [Google Scholar]
- 12.Spertus JA and Jones PG. Development and Validation of a Short Version of the Kansas City Cardiomyopathy Questionnaire. Circ Cardiovasc Qual Outcomes. 2015;8:469–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Basch E, Deal AM, Dueck AC, Scher HI, Kris MG, Hudis C and Schrag D. Overall Survival Results of a Trial Assessing Patient-Reported Outcomes for Symptom Monitoring During Routine Cancer Treatment. JAMA. 2017;318:197–198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Basch E, Deal AM, Kris MG, Scher HI, Hudis CA, Sabbatini P, Rogak L, Bennett AV, Dueck AC, Atkinson TM, Chou JF, Dulko D, Sit L, Barz A, Novotny P, Fruscione M, Sloan JA and Schrag D. Symptom Monitoring With Patient-Reported Outcomes During Routine Cancer Treatment: A Randomized Controlled Trial. J Clin Oncol. 2016;34:557–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kalwani N CJ, Verghese G, Gupta A, Zheng J, Brown-Johnson C, Amano A, Villender S, Winget M, Asch S, Heidenreich P, Sandhu AT. The Patient-Reported Outome Measurement in Heart Failure Clinic Trial: Methods and Rationale. Am Heart J. 2022. 10.1016/j.ahj.2022.10.08. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N and Conde JG. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42:377–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Cohen J Statistical Power Analysis for the Behavioral Sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates; 1988. [Google Scholar]
- 18.Shafiq A, Arnold SV, Gosch K, Kureshi F, Breeding T, Jones PG, Beltrame J and Spertus JA. Patient and physician discordance in reporting symptoms of angina among stable coronary artery disease patients: Insights from the Angina Prevalence and Provider Evaluation of Angina Relief (APPEAR) study. Am Heart J. 2016;175:94–100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Arnold SV, Grodzinsky A, Gosch KL, Kosiborod M, Jones PG, Breeding T, Towheed A, Beltrame J, Alexander KP and Spertus JA. Predictors of Physician Under-Recognition of Angina in Outpatients With Stable Coronary Artery Disease. Circ Cardiovasc Qual Outcomes. 2016;9:554–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Qintar M, Spertus JA, Gosch KL, Beltrame J, Kureshi F, Shafiq A, Breeding T, Alexander KP and Arnold SV. Effect of angina under-recognition on treatment in outpatients with stable ischaemic heart disease. Eur Heart J Qual Care Clin Outcomes. 2016;2:208–214. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Turgeon RD, Barry AR, Hawkins NM and Ellis UM. Pharmacotherapy for heart failure with reduced ejection fraction and health-related quality of life: a systematic review and meta-analysis. Eur J Heart Fail. 2021;23:578–589. [DOI] [PubMed] [Google Scholar]
- 22.Butler J, Filippatos G, Jamal Siddiqi T, Brueckmann M, Bohm M, Chopra VK, Pedro Ferreira J, Januzzi JL, Kaul S, Pina IL, Ponikowski P, Shah SJ, Senni M, Vedin O, Verma S, Peil B, Pocock SJ, Zannad F, Packer M and Anker SD. Empagliflozin, Health Status, and Quality of Life in Patients With Heart Failure and Preserved Ejection Fraction: The EMPEROR-Preserved Trial. Circulation. 2022;145:184–193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Jankowska EA, Kirwan BA, Kosiborod M, Butler J, Anker SD, McDonagh T, Dorobantu M, Drozdz J, Filippatos G, Keren A, Khintibidze I, Kragten H, Martinez FA, Metra M, Milicic D, Nicolau JC, Ohlsson M, Parkhomenko A, Pascual-Figal DA, Ruschitzka F, Sim D, Skouri H, van der Meer P, Lewis BS, Comin-Colet J, von Haehling S, Cohen-Solal A, Danchin N, Doehner W, Dargie HJ, Motro M, Friede T, Fabien V, Dorigotti F, Pocock S and Ponikowski P. The effect of intravenous ferric carboxymaltose on health-related quality of life in iron-deficient patients with acute heart failure: the results of the AFFIRM-AHF study. Eur Heart J. 2021. Jun 3;42(31):3011–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Nassif ME, Windsor SL, Borlaug BA, Kitzman DW, Shah SJ, Tang F, Khariton Y, Malik AO, Khumri T, Umpierrez G, Lamba S, Sharma K, Khan SS, Chandra L, Gordon RA, Ryan JJ, Chaudhry SP, Joseph SM, Chow CH, Kanwar MK, Pursley M, Siraj ES, Lewis GD, Clemson BS, Fong M and Kosiborod MN. The SGLT2 inhibitor dapagliflozin in heart failure with preserved ejection fraction: a multicenter randomized trial. Nat Med. 2021;27:1954–1960. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Morris AA, Khazanie P, Drazner MH, Albert NM, Breathett K, Cooper LB, Eisen HJ, O’Gara P, Russell SD, American Heart Association Heart F, Transplantation Committee of the Council on Clinical C, Council on Arteriosclerosis T, Vascular B, Council on Cardiovascular R, Intervention and Council on H. Guidance for Timely and Appropriate Referral of Patients With Advanced Heart Failure: A Scientific Statement From the American Heart Association. Circulation. 2021;144:e238–e250. [DOI] [PubMed] [Google Scholar]
- 26.Mondesir FL, Zickmund SL, Yang S, Perry G, Galyean P, Nativi-Nicolau J, Kemeyou L, Spertus JA and Stehlik J. Patient Perspectives on the Completion and Use of Patient-Reported Outcome Surveys in Routine Clinical Care for Heart Failure. Circ Cardiovasc Qual Outcomes. 2020;13:e007027. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Spertus J, Peterson E, Conard MW, Heidenreich PA, Krumholz HM, Jones P, McCullough PA, Pina I, Tooley J, Weintraub WS, Rumsfeld JS and Cardiovascular Outcomes Research C. Monitoring clinical changes in patients with heart failure: a comparison of methods. Am Heart J. 2005;150:707–15. [DOI] [PubMed] [Google Scholar]
- 28.Spertus J Barriers to the use of patient-reported outcomes in clinical care. Circ Cardiovasc Qual Outcomes. 2014;7:2–4. [DOI] [PubMed] [Google Scholar]
- 29.CAHPS Clinician and Group Survey. Accessed March 15, 2021. https://www.ahrq.gov/cahps/surveys-guidance/cg/index.html.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.