Patients report increased control and decreased impact of heart failure during pulmonary pressure‐guided management

Abstract

Aims

This study investigated Class III heart failure patient experience with pulmonary artery pressure (PAP) monitoring over 24 months including (1) changes in reported quality of life (QoL), (2) changes in HF illness perception and sense of control (control), (3) patient evaluation of PAP‐guided therapy as a modality and (4) behavioural adherence to remote monitoring.

Methods and results

A 272 patient subgroup from the N = 1200 patient CardioMEMS™ Post‐Approval Study completed 3 questionnaires including HF‐related QoL, illness perception and control, and PAP therapy evaluation. Patient measurement frequency was also collected as a behavioural adherence index. Patient perception of illness was assessed via a general linear mixed model to account for all data available at all time points (n = 272). A paired analysis of patients with 24‐month questionnaires (n = 135) was performed to further evaluate patients who completed follow‐up. Patients reported significant improvement in overall QoL scores from baseline (44.5 ± 24.0) to 6 months (58.8 ± 25.2) which was maintained through 2 years (all P < 0.0001). Patients described their sense of control as strong (4.3 ± 0.9, 5 is strongly agree), positively evaluated PAP‐guided therapy (4.1 ± 1.0, 5 is strongly agree), and 90% actively transmitted device data weekly with a significant reduction in patient perceived illness throughout 2 years of follow‐up.

Conclusions

Two‐year follow up of the patient experience of PAP‐guided therapy indicated that patients reported increased QoL, perceptions of control, device acceptability and continued remote monitoring over the 24 month follow‐up. These results suggest that knowledge of PA pressures is associated with positive patient experience and engagement in HF care.

Introduction

The day‐to‐day patient burden associated with the management of chronic heart failure (HF) spans multiple behaviours of self‐care, treatment adherence and emotional impacts. ¹ , ² Optimal self‐care involves health maintenance, monitoring for changes in HF signs and symptoms, and management of changes when they present. ³ Traditionally, providers have relied upon face‐to‐face evaluation during periodic clinic visits and sporadic patient monitoring of symptoms and weights at home. Engaging patients in the process of HF management is part of an optimal care plan, and understanding patient beliefs and their experiences may facilitate better outcomes. Modern disease management aspires to become increasingly pro‐active to prevent hospitalizations and improve broad health outcomes.

Early recognition of new or worsening symptoms with associated adjustment in a patient's treatment plan is key to preventing HF hospitalization and improving quality of life. Studies of implantable hemodynamic monitoring devices have demonstrated that cardiac filling pressures typically rise weeks before overt signs and symptoms of congestion. ⁴ HF symptoms and weight changes have not been a reliable indicator of worsening HF for patients or their care team to recognize and intervene to prevent escalating congestion. Treatment guided by pulmonary artery pressure (PAP)‐monitoring has reduced HF symptom burden as well as hospitalizations, with equal or greater value for HF with preserved ejection fraction and for HF with reduced ejection fraction. ⁵ , ⁶ , ⁷ , ⁸ The remarkable improvement in quality of life in the recent randomized MONITOR‐HF trial likely includes patient awareness of active monitoring to guide their therapy, compared to previous trials in which patients were kept blind to their participation. ⁹ Even when hospitalizations do not occur, unanticipated episodes of symptomatic congestion may impair patients' quality of life and undermine their sense of control over the illness. ¹⁰ Ambulatory monitoring of cardiac filling pressures offers patients a ‘tool’ to strengthen their confidence in the partnership with their care team to reduce symptom burden and promote best outcomes. The goal of this examination is to explore the impact of this care strategy to increase the patient sense of control over their HF in addition to decreasing their symptoms.

The analyses included here examined patient experiences from baseline to 24 months follow‐up including (1) changes in reported quality of life, (2) changes in HF illness perception and sense of control, (3) patient evaluation of PAP‐guided therapy as a modality and (4) behavioural adherence to remote monitoring and data transfer. We hypothesized that PAP‐guided therapy provides a ‘tangible’ metric of congestion that patients and their care team can share to foster increased perceptions of control to prevent and correct congestion.

Methods

The CardioMEMS Post‐Approval Study (PAS) (NCT02279888) is a multicentre, prospective, open‐label, single‐arm trial evaluating the use of PA pressure‐guided therapy in routine clinical practice. This study enrolled 1200 patients at 104 clinical sites across the United States. The study protocol was approved by all responsible institutional review boards and conducted according to Good Clinical Practice guidelines. Study design information and the positive results from the study have been reported previously. ⁶ , ⁸ All participants provided written informed consent and were followed for 24 months or until they exited the study.

This analysis reports on a prospective sub‐study focused on the patient experience of PAP monitor use. The post‐approval study was already well underway when the need for this sub‐study was recognized by the Steering Committee and investigators as they observed positive changes in patient understanding and health‐related behaviours. The sub‐study was incorporated subsequently through an amendment of the study protocol and consent forms, by which time only 272 of the 1200 patients had baseline questionnaires administered.

Patient perception was assessed at each time point and changes in patient perception of their illness were assessed by comparing the follow‐up values to the baseline values in two ways. First, in order to account for all data available at all time points, a general linear mixed model (least squares means) with factors for subject ID and time point was used to assess all subjects who completed a questionnaire at baseline and at least one subsequent time (n = 272). To further evaluate subjects who completed 24‐month follow‐up, a paired analysis for subjects with questionnaires at the 24‐month follow‐up visit (n = 135) was performed.

Baseline characteristics

A comparison of baseline characteristics, assessments, medical histories and medications of patients who were administered baseline questionnaire and completed questionnaires at the 24‐month follow‐up visit (n = 135) and patients who did not have 24‐month questionnaires, for any reason (e.g., may not have had baseline questionnaire), was performed. The significance of each variable was assessed using a two‐sided test and a significance level of 5%.

Quality of life

Quality of life (QoL) was assessed via the Kansas City Cardiomyopathy Questionnaire. The Kansas City Cardiomyopathy Questionnaire (KCCQ) is a 12‐item self‐administered questionnaire that has been developed to independently measure patient perceptions of their health status. Analysis of the KCCQ includes both overall score types such as the Clinical Summary Score (CSS) and Overall Summary Score (OSS) which assess varying domains to gain a holistic view of QoL, as well as analysis of individual domains (e.g., HF related symptoms, physical limitations and social limitations). A clinically meaningful change in QoL as assess by the KCCQ starts with an improvement of 5 score points. A 5 score point improvement is currently considered a small clinically meaningful change in QoL, whereas an improvement in 10 score points is considered moderate and a 20 point improvement is considered large. ¹¹ As part of the research presented here, an analysis of patients with baseline questionnaires who completed 24‐month questionnaires was performed.

Measurement of illness perception and control

Patients were administered the 9‐item Brief Illness Perception Questionnaire (BIPQ) and a set of user‐experience survey items at baseline and every 6 months for 2 years. ¹² The BIPQ consists of nine questions that patient's rate on a scale from 0 (minimum) to 10 (maximum). The first five questions assess patients' cognitive perceptions of their illness including the amount their illness affects their life (consequences); how long their illness will continue (duration); the amount of control over their illness (personal control); assessment of treatment effectiveness (treatment control); and the extent to which patients experience symptoms (identity). The next three questions assess the emotional impact and patient understanding of their illness and included patients' level of concern (concern) and understanding of their illness (illness comprehensibility), as well as the emotional effect of their illness (emotional representations). The final question is open‐ended and asks patients to rank the three most important factors causing their illness. A composite score can be formed via summation, as well as individual item interpretation with the higher scores indicating more negative perception of patient illness.

Patient evaluation of PAP therapy

User experience questions were investigator‐derived and asked participants to rate their user satisfaction and perceived benefits related to the PAP monitoring system. Items involved both positively and negatively worded items to avoid bias in the response set. Questions were rated on a scale of 1 to 5 with 1 being strongly disagree and 5 being strongly agree. Each question was scored individually at each time point.

Adherence to PAP monitoring

Patient engagement in regular PAP readings was calculated as the total number of readings divided by the number of days a patient was in the study. Weekly adherence was calculated for each patient and used to derive mean weekly compliance for the entire population.

Significance level was 5% using two‐sided tests unless stated otherwise. SAS version 9.4 was used for the analyses. Descriptive statistics are presented as number of observations, mean and standard deviation, or median/interquartile range (as applicable) for continuous variables, and as counts and percentages/rates for categorical variables.

Results

A total of 1200 patients were enrolled into the PAS. Patient‐reported outcomes (PRO) including patient sense of control over illness were not originally part of the study design but were added later as an exploratory sub‐study with strong encouragement from the investigators. After amendments and re‐consenting, this sub‐study began almost 2 years from the first patient enrolment. A total of 329 subjects were enrolled after PRO questionnaires were incorporated into the PAS. Of the possible 329 subjects, 272 subjects completed baseline questionnaires and 216 subjects completed a baseline questionnaire and at least one follow‐up questionnaire. Of the 272 patients who completed a baseline questionnaire, 177 completed a questionnaire at the 1‐year follow‐up visit and 135 completed a questionnaire at both baseline and the 24‐month visit. Attrition of subjects was due to 73 subjects who expired prior to 24‐month follow‐up, 39 subjects who withdrew, 18 subjects who missed their 24‐month visit or did not fill out a questionnaire, and 7 subjects who were lost to follow‐up.

Demographic data for the complete CardioMEMS™ US PAS cohort has been previously published for both 1 year ⁶ and 2‐year follow‐up. ⁸ For this analysis, subject demographic data was stratified by subjects with questionnaires paired at baseline and at the 24‐month follow‐up visit (n = 135) versus the remaining subjects in the study cohort (n = 1065) (Table  1 ). When compared, these two groups were generally similar, although subjects with 24‐month follow‐up data tended to be slightly younger (67.8 years) compared to subjects without paired 2‐year data (69.6 years, P = 0.05), with higher BMI (33.8 vs. 31.4, P = 0.002) and ejection fraction (42.0% vs. 38.4%, P = 0.03).

Table 1.

Subject baseline characteristics stratified by presence of KCCQ data at 2 years of follow‐up

	KCCQ data (N = 135)	No KCCQ data (N = 1065)	All subjects (N = 1200)	P‐value
Age (year) [mean ± SD (n)]	67.8 ± 10.3 (135)	69.6 ± 11.9 (1065)	69.4 ± 11.7 (1200)	0.0543
Female	41.5% (56/135)	37.2% (396/1065)	37.7% (452/1200)	0.3465
Body mass index [mean ± SD (n)]	33.8 ± 8.2 (135)	31.4 ± 7.8 (1065)	31.7 ± 7.9 (1200)	0.0015
GFR [mean ± SD (n)]	55.0 ± 18.8 (135)	53.0 ± 21.4 (1065)	53.2 ± 21.1 (1200)	0.2536
Ejection fraction [mean ± SD (n)]	42.0 ± 17.2 (135)	38.4 ± 17.4 (1063)	38.8 ± 17.4 (1198)	0.0259
Cardiac index [mean ± SD (n)]	2.26 ± 0.63 (135)	2.23 ± 0.72 (1061)	2.23 ± 0.71 (1196)	0.5387
Ischaemic cardiomyopathy	37.8% (51/135)	41.8% (445/1065)	41.3% (496/1200)	0.4044
CRT/CRT‐D or ICD	43.7% (59/135)	51.0% (543/1065)	50.2% (602/1200)	0.1206
Hypertension	91.1% (123/135)	86.9% (926/1065)	87.4% (1049/1200)	0.2141
Coronary artery disease	63.7% (86/135)	66.3% (706/1065)	66.0% (792/1200)	0.5635
Diabetes mellitus	58.5% (79/135)	54.1% (576/1065)	54.6% (655/1200)	0.3593
Chronic obstructive pulmonary disease	34.1% (46/135)	35.3% (376/1065)	35.2% (422/1200)	0.8484

Quality of life

We examined the 135 patients who completed 2‐year follow‐up in a paired analysis with data at both baseline and 24 months. A statistically significant and clinically meaningful improvement from baseline was observed in both the overall summary score and individual domains at all time points (all P < 0.0001 except physical limitation score at 24 months P < 0.01) (Figure 1 and Table S1). Improvements in quality of life were observed in patients with preserved and reduced ejection fraction (all P < 0.0001). Both men and women reported improvement in quality of life (all P < 0.0001).

Figure 1.

Changes in KCCQ scores from baseline for 24 months paired cohort. Changes in KCCQ Scores over study follow‐up. A statistically significant and clinically meaningful improvement from baseline was observed in both the overall summary score and individual domains at all time points.

Illness perception

Initially, the data was assessed using the least squares means approach so that all patient data including repeated questionnaires could be included in the analysis. This mixed model analysis showed a significant improvement in all questions measured by the Brief Illness Perception Questionnaire (BIPQ) and a significant reduction of 5.4 ± 0.9 points from baseline (P < 0.0001) in the BIPQ overall summary score for perceived impact of illness.

Patients reported significant improvement in their cognitive and emotional perception of their illness, Figure 2 . Their overall summary score improved at each timeframe measured and was maintained out to 2 years. The improvement in perception of illness and adherence to monitoring was the same regardless of the patient's ejection fraction or gender. This was accompanied by a remarkably high level of adherence with hemodynamic data transmissions (mean weekly compliance with taking readings 90%) through 2 years of follow‐up (Figure  3 ).

Figure 2.

BIPQ individual scores for 24 months paired cohort. Changes in individual BIPQ scores over study follow‐up. Panel (A) depicts significant decreases in patient perception of how much a concern their illness is and the severity of their illness, as well as the amount of severe symptoms and the effects of their illness on their emotional well‐being. In Panel (B), a statistically significant improvement in understanding and feeling of control over illness is observed. A high score for the belief that treatment can help illness was also observed. This was scored high at baseline and this score was maintained with a small improvement, not significant, over the duration of the 2‐year follow‐up.

Figure 3.

Adherence to at‐home reading and BIPQ summary score over study duration. Adherence to at‐home reading and BIPQ summary score over study duration for patients with paired baseline and 24‐month questionnaires. The overall BIPQ summary score (black) is significantly reduced signifying a reduction in illness perception while patient weekly adherence to taking monitor readings remains high throughout study follow‐up. The red line in the figure depicts locally weighted smoothing (LOESS) of the CardioMEMS™ readings.

Patients reported a high level of understanding of their condition and a realistic outlook that this was a lifelong condition. Patients reported significantly reduced symptoms, improved control, and less impact of their illness both emotionally and on their life at each timeframe measured.

PAP‐guided therapy

In addition to understanding patient perception of illness, this study employed a survey which would allow us to understand the patients' satisfaction and burden with remote monitoring. The survey demonstrated that patients had a high level of patient satisfaction with the monitoring device, improved sense of control, and strongly agreed that they would get the ambulatory hemodynamic monitoring system again (Figure  4 ). When patients were asked if their monitoring system gave them a sense of control, this was high on the Likert scale (1, strongly disagree to 5, strongly agree) at the first‐time patients were eligible to answer this question (4.1 ± 1.0 at 6 months), which was maintained with a small but significant increase at 24 months to 4.3 ± 0.9. Similarly, patients reported response to ‘pressure monitoring system provides me a sense of security’ was high at 6 months, 4.2 ± 0.9 with a slight increase at 24 months, 4.3 ± 1.0.

Figure 4.

Patient evaluation of haemodynamic (PAP) monitoring. Patient evaluation of the monitoring system was assessed on the Likert scale (1, strongly disagree to 5, strongly agree) at baseline (when appropriate), 6, 12, 18, and 24 months post sensor implant. Patients reported strong agreement that they were satisfied with their monitoring system, that they would agree to get their monitoring system again, and that their monitoring system gave them a sense of control and a sense of security.

Adherence to PAP monitoring

In this sub‐group, patient adherence to transmitting pressures was high throughout the 24‐month follow‐up period [80.0% ± 19.9% (89.0% median) daily, 97.0% ± 6.7% (100.0% median) weekly at 1 year; and 74.7% ± 22.5% (81.5% median) daily, 94.9% ± 9.5% (99.0% median) weekly at 2 years], which was comparable to the entire PAS study cohort [76.3% ± 23.9% (85.0% median) daily, 92.8% ± 15.8% (100.0% median) weekly at 1 year; and 71.9% ± 25.4% (80.0% median) daily, 90.0% ± 18.2% (98.8% median) weekly at 2 years].

Discussion

HF patients using PAP‐guided therapy reported clinically and statistically significant improvements in quality of life as measured by the KCCQ scales of overall summary and each of the subscales. For the first time in a management strategy trial, patients also reported marked decreases in perception of their HF illness, from pre‐treatment to all follow up time points, with particular effects on increased perceived control of HF. This patient‐centric measure was used to examine how PAP‐guided therapy may impact patients' experience of HF. Changes in disease perception have not previously been described for physiologic HF interventions, patient beliefs about their illness can predict both emotional and behavioural responses to therapy. ¹²

Insights into how HF patients interpret, feel, act and derive benefit from PAP‐guided therapy may help to clarify the value proposition for patients considering implantation of PAP systems for HF management. HF is known to be a challenging condition requiring patients to manage and adapt their lives continuously. ³ The Common Sense Model of Health has been postulated to address the key elements of how patients come to understand their condition including illness identity (diagnosis), illness cause, illness timeline (expected course), illness controllability and illness consequences. ¹³ Patient perceptions of their illness have been consistently linked to self‐reported health and adjustment to illness. Specifically in a Danish HF sample, patient perception of illness, particularly levels of control, identity of illness and illness consequences, was associated with general health ratings, even after controlling for NYHA class, mood and other illness perceptions. ¹⁴ Recent studies of PAP‐guided therapy have demonstrated key benefits in terms of depressive symptoms, cardiac anxiety, quality of life and specific health behaviours. ⁹ , ¹⁵ , ¹⁶ The overall patient experience with PAP guided therapy may serve to ‘unmask and de‐mystify’ HF management by creating a tangible treatment target (i.e., pressures) and give patients a more direct sense of control over their condition and health.

Multiple contemporary HF therapies have been shown to improve quality of life or decrease the decline in quality of life, as measured in standard questionnaires about symptoms and activities in KCCQ and MLHF. However, PAP‐guided therapy is the first heart failure therapy that has been shown to improve patient sense of control and perception of their illness. Patients reported positive appraisal and satisfaction with PAP‐guided therapy in follow‐up survey items. In addition, there was a high level of patient adherence to regular monitoring from device implant to 24 months of follow‐up. These results suggest that PAP‐guided therapy may allow HF patients to experience greater levels of confidence and behavioural engagement in self‐care of HF and with their care team. This level of adherence has not been sustained even over shorter periods of time in other large‐scale telemonitoring trials. ¹⁷ In the recently published CONNECT‐HF trial, adherence with a mobile application for HF self‐monitoring fell to less than 30% within 6 weeks. ¹⁸

Limitations

The results of this study should be interpreted in the context of several potential limitations. First, the CardioMEMS™ US PAS is a prospective, open‐label, single‐arm trial evaluating the use of PA pressure‐guided therapy in routine clinical practice, where patients served as their own control. Even in the absence of a control arm, it is unlikely that a placebo effect of device implantation would endure for more than 6 months and in this study results from hemodynamic‐guided care were sustained through 2 years. However, we are unable to solely assign the beneficial effects to the care strategy without a control group comparison. All data was collected before COVID‐19 and focused only on Class III HF patients with a prior HF hospitalization, so generalization of our data to the current broader indications of Class II and III should be done cautiously. As the decision to assess patient reported outcomes arose after experience with the monitoring strategy, the questionnaires were not incorporated until the trial was well underway. Thus, the sample presented here is a sub‐study of a larger 1200 patient study. As the difference was only date of enrolment, no specific differences were found between the sub‐study and the larger study. However, we cannot rule out the effects of a survivor bias in our long‐term follow up data and/or the effects of demand characteristics that might cue participants about the purpose of expectations of the research. Nonetheless, these results suggest that the potential to improve quality of life and patients' confidence of control over their HF. These data provide more information about a significant potential benefit when discussing the benefits of ambulatory pulmonary pressure monitoring during shared‐decision making.

Conclusions

The management of HF from a patient perspective remains a complicated and demanding commitment that now includes multiple treatment modalities. Heart failure patients in the PAS study managed using PAP‐guided therapy reported increased quality of life, perceptions of control, device acceptability and continued remote monitoring over 24‐month follow‐up. These results suggest that knowledge of PA pressures could be associated with positive patient experience and engagement in HF care in the future. PAP‐guided therapy may be increasingly seen by patients as a route to increased empowerment and success in managing HF.

Funding

The United States CardioMEMS™ Post‐Approval Study (NCT 02279888) was funded by Abbott.

Conflict of interest

L.D.R. received honoraria and/or serves as a consultant to Abbott, Alnylam, Abiomed, Boehringer Ingelheim, Impulse Dynamics, Medtronic, Boston Scientific, Merck, Novo Nordisk. S.F.S. serves as a consultant to Medtronic, Abbott and Milestone Pharmaceutical and has received honorarium from Medtronic and Zoll Medical. M.‐E.B. is an Abbott employee and shareholder. J.D.H. is an Abbott employee and shareholder. N.R.C. no disclosures. P.B.A. is an Abbott employee and shareholder. L.W.S. received research support Abbott.

Supporting information

Table S1. Summary of Kansas City Cardiomyopathy Questionnaire ‐ paired at 24 months.

Rathman, L. D. , Sears, S. F. , Brett, M.‐E. , Henderson, J. D. , Castillo, N. R. , Adamson, P. B. , and Stevenson, L. W. (2025) Patients report increased control and decreased impact of heart failure during pulmonary pressure‐guided management. ESC Heart Failure, 12: 3381–3387. 10.1002/ehf2.15364.