Association of Novel Ventricular Assist Device Self-report Measures With Overall Health-Related Quality of Life

Abstract

Background:

Few study authors examined factors influencing health-related quality of life (HRQOL) early after left ventricular assist device (LVAD) implantation.

Objective:

The purpose of this study was to determine whether 5 novel self-report measures and other variables were significantly associated with overall HRQOL at 3 months after LVAD surgery.

Methods:

Patients were recruited between October 26, 2016, and February 29, 2020, from 12 US sites. Data were collected before LVAD implantation and at 3 months post LVAD implantation. Overall HRQOL measures included the Kansas City Cardiomyopathy Questionnaire-12 (KCCQ-12) overall summary score (OSS) and EuroQol 5-dimension-3L visual analog scale. Potential factors associated with overall HRQOL included 5 novel self-report measures (Satisfaction with Treatment, Being Bothered by VAD Self-care and Limitations, VAD Team Communication, Self-efficacy regarding VAD Self-care, and Stigma), and demographic and clinical characteristics. Statistics included regression analyses.

Results:

Of enrollees, 242 completed self-report measures at baseline, and 142 completed measures 3 months postoperatively. Patients were 55 ± 13 years old, with 21% female, 24% non-White, 39% high school or lower educated, and 47% destination therapy. Using the KCCQ-12 OSS, higher Satisfaction with Treatment was associated with a higher KCCQ-12 OSS; Being Bothered by VAD Self-care and Limitations, high school or lower education, chest incision pain, cardiac dysrhythmias within 3 postoperative months, and peripheral edema were associated with a worse KCCQ-12 OSS (R² = 0.524). Factors associated with a worse 3-month EuroQol 5-dimension-3L visual analog scale were female sex, adverse events within 3 months post implantation (cardiac dysrhythmias, bleeding, and venous thrombosis), and chest incision pain (R² = 0.229). No factors were associated with a higher EuroQol 5-dimension-3L visual analog scale score at 3 months.

Conclusions:

Two novel measures, demographics, postimplantation adverse events, and symptoms were associated with post-LVAD KCCQ-12 OSS early after surgery.

Health-related quality of life (HRQOL) is an important patient-centered outcome for patients with heart failure who undergo advanced surgical therapies, including left ventricular assist device (LVAD) implantation. Studies have demonstrated significant and clinically important improvement in patient HRQOL from before to early and later after LVAD implantation as a bridge to heart transplantation and destination therapy (ie, long-term therapy if ineligible for heart transplantation).^1–8 A few studies have identified factors associated with HRQOL post implantation, including comorbidities, rehospitalization, and adverse events.^3,6,8,9 To our knowledge, authors of only 1 study, using the Interagency Registry for Mechanically Assisted Circulatory Support database, examined the relationship between overall HRQOL and adjustment to LVAD implantation, using exploratory single items (eg, patient satisfaction with an LVAD and confidence in LVAD self-care).⁹ Assessment of adjustment (an overarching multidimensional construct of adaptation to a chronic illness and its treatment)¹⁰ to LVAD implantation is important because it may provide targets for interventions to facilitate patient recovery early after implantation and later while living with an LVAD.

Grady et al¹¹ previously developed 5 novel ventricular assist device (VAD)-specific self-report measures of adjustment, which meet psychometric measurement standards, namely, Satisfaction with Treatment, Being Bothered by VAD Self-care and Limitations, VAD Team Communication, Self-efficacy regarding VAD Self-care, and 1 novel measure of the mental domain of HRQOL, that is, Stigma, using data from their National Institutes of Health/National Heart, Lung, and Blood Institute–funded study, “Mechanical Circulatory Support: Measures of Adjustment and Quality of Life.” Hahn et al¹² also established validity of 12 existing Patient-Reported Outcomes Measurement Information System self-report measures in patients who undergo LVAD implantation. The purpose of this report was to determine whether self-report measures, including the 5 novel measures and 3 Patient-Reported Outcomes Measurement Information System measures that assess support, as well as other factors (eg, patient characteristics and postoperative events), were significantly associated with patient overall HRQOL at 3 months after surgery. The framework for analyses was a previously developed conceptual model of potential factors associated with overall HRQOL (Figure 1).¹³ We chose the 3-month time point because it is a time when patients are completing recovery from surgery, finishing rehabilitation, and moving from the perioperative period back to a more usual lifestyle.

FIGURE 1.

Conceptual model of adjustment to mechanical circulatory support and mechanical circulatory support health-related quality of life. BTT, bridge to transplant; DT, destination therapy; EQ-5D-3L, EuroQol 5-dimension-3L; HF, heart failure; HRQOL, health-related quality of life; KCCQ-12, Kansas City Cardiomyopathy Questionnaire-12; MCS, mechanical circulatory support; VAD, ventricular assist device.

We hypothesized that the 5 novel self-report measures, heart failure– and LVAD-related symptoms, clinical variables (ie, comorbidities, preimplant strategy, postimplant adverse events, and rehospitalization), and resources (ie, 3 Patient-Reported Outcomes Measurement Information System measures of support and healthcare-related financial difficulties) would be associated with patient overall HRQOL at 3 months after LVAD implantation.

Materials and Methods

Sites and Sample

Patients were recruited between October 26, 2016, and February 29, 2020, from 12 US VAD programs that participated in the Mechanical Circulatory Support: Measures of Adjustment and Quality of Life study. Enrollment occurred before primary implantation of a continuous flow LVAD (baseline, ie, accepted or scheduled for device surgery) with follow-up through 3 postoperative months; further follow-up occurred after 3 months but was not included in this report. Inclusion criteria were being older than 19 years, all implant strategies, sufficient cognitive ability to provide self-report data, and being willing and able to provide written informed consent. Exclusion criterion was being scheduled to receive a biventricular VAD, right VAD, or total artificial heart. Institutional review board approval was obtained by all sites. Participants provided written informed consent before study participation.

Measures and Procedures

Our measures of overall HRQOL included the EuroQol 5-dimension-3L (EQ-5D-3L)^14,15 visual analog scale (VAS), a generic measure, and the Kansas City Cardiomyopathy Questionnaire-12 (KCCQ-12)^16,17 overall summary score (OSS), a heart failure–specific measure, as described in Table 1. The 5 novel measures (ie, Satisfaction with Treatment, Being Bothered by VAD Self-care and Limitations, VAD Team Communication, Self-efficacy regarding VAD Self-care, and Stigma) have been described in detail elsewhere, including development and psychometric testing in accordance with Patient-Reported Outcomes Measurement Information System methodology (Table 1).^11,13,18,19 We also used individual items regarding heart failure– and LVAD-specific symptoms (Table 1), created by Grady et al.¹³ Resource measures included 3 types of support from the Patient-Reported Outcomes Measurement Information System (emotional, informational, and instrumental) (Table 1)^20,21 and financial strain, assessed using 1 item from the 30-item European Organization for the Research and Treatment of Cancer Quality of Life Questionnaire (Table 1).^22,23

TABLE 1.

Mechanical Circulatory Support: Measures of Adjustment and Quality of Life Measures

Construct	Name of Instrument	Definition/Description	No. of Items (if Appropriate)/Instrument Scoring/Meaning of Higher Score
Outcome measures
Generic HRQOL	EQ-5D-3L	VAS assesses overall HRQOL.	1 VAS VAS, 0–100 points Higher score = better overall HRQOL
Heart failure-specific HRQOL	KCCQ-12	Measures health status in 4 domains, ie, physical limitations, symptom frequency, quality of life, and social limitations, and has an OSS	12 items Domain scores are summed and represented on a 0–100 point scale. Higher score = better overall heart failure-specific health status
Newly developed MCS A-QOL measures
Satisfaction with treatment	Satisfaction with Treatment	The degree to which patients perceive that a treatment fulfills their health needs and therapeutic outcome (ie, treatment benefit). Regarding VADs, items include effectiveness of the VAD, getting used to having a VAD, and satisfaction with having a VAD.	IRT T score Higher score = greater Satisfaction with VAD treatment
	VAD Team Communication	Communication refers to a process by which information is exchanged between individuals. This measure asks patients about communication from the VAD team about VADs. Items in this measure include communication by the VAD team regarding daily activities, personal relationships, and emotions.	IRT T score Higher score = better VAD Team Communication
	Being Bothered by VAD Self-care and Limitations	Bothersome refers to something being annoying or aggravating. This measure assesses the level of being bothered by having a VAD, driveline, and wearables. Examples of items are caring for VAD equipment and limitations in activities of daily living.	IRT T score Higher score = being more bothered by VAD self-care/having more limitations
Self-efficacy	Self-efficacy Regarding VAD Self-care	Self-efficacy refers to belief in one's ability to execute behaviors necessary for performance attainment. This measure assesses self-efficacy regarding incorporating pump management into one's daily life (eg, responding to alarms, taking care of VAD equipment, and contacting a VAD clinician).	IRT T score Higher score = better VAD self-care self-efficacy
Mental HRQOL	Stigma	Defined as perceptions of self and publicly enacted negativity, prejudice, and discrimination as a result of disease- or treatment-related manifestations. This measure assesses stigma attributed to having a VAD and includes being embarrassed about having a VAD, feeling different from others, and unhappiness about how the VAD affects one's appearance.	IRT T score Higher score = more stigma
Symptoms	HF and MCS symptoms	Heart failure symptoms include frequency of peripheral edema, abdominal bloating, and poor appetite. MCS symptoms include severity of pain (chest incision and driveline exit site) and discomfort wearing VAD equipment, and frequency of dizziness and fluid drainage from the driveline exit site.	8 items HF symptoms: ordinal scale, 1 (never) to 5 (almost constantly) MCS symptoms (pain and discomfort items): ordinal scale, 1 (none) to 5 (very severe) MCS symptoms (dizziness and fluid drainage from driveline exit site items): ordinal scale, 1 (never) to 5 (almost constantly) Higher score = more frequent or more severe symptoms, depending on the item
Resources
Social support	PROMIS emotional support, v2.0	Perceived feelings of being cared for and valued as a person; having confident relationships	IRT T score Higher score = more emotional support
	PROMIS informational support, v2.0	Perceived availability of helpful information or advice	IRT T score Higher score = more informational support
	PROMIS instrumental support, v2.0	Perceived availability of assistance with material, cognitive, or task performance	IRT T score Higher score = more instrumental support
Financial	EORTC QLQ-C30	Measure of financial strain: Has your physical condition or medical treatment caused you financial difficulties?	1 item Ordinal scale: 1, not at all, to 4, very much Higher score = greater difficulty

T score: mean (SD), 50 (10).

Abbreviations: EORTC QLQ-C30, 30-item European Organization for the Research and Treatment of Cancer Quality of Life Questionnaire; EQ-5D-3L, EuroQol 5-d imensi on-3L; HF, heart failure; HRQOL, health-related quality of life; IRT, item response theory; KCCQ-12, Kansas City Cardiomyopathy Questionnaire-12; MCS, mechanical circulatory support; MCS A-QOL, Mechanical Circulatory Support: Measures of Adjustment and Quality of Life; OSS, overall summary score; PROMIS, Patient-Reported Outcomes Measurement Information System; VAD, ventricular assist device; VAS, visual analog scale.

The original conceptual model¹³ to be evaluated was expanded to include the specific variables under headings that may be associated with overall HRQOL. Variables included (1) the 5 novel self-report measures (under the heading “Impact of Disease and Treatment”), (2) the 3 Patient-Reported Outcomes Measurement Information System measures of support and 1 financial strain item (under the heading “Resources”), and (3) demographic and preimplantation and postimplantation clinical variables (Figure 1).

Clinical data were collected directly from site medical records or securely downloaded from the Society of Thoracic Surgeons Interagency Registry for Mechanically Assisted Circulatory Support database. Demographics, outcome measures, and impact of disease and treatment measures were collected from patients via self-administration (computer or paper) or interviewer (in-person) administration. Data were collected at baseline and 3-month follow-up.

Statistical Analyses

Variables were summarized using means and standard deviation, or counts and percentages, as appropriate. Scoring for the 5 novel measures and 3 Patient-Reported Outcomes Measurement Information System support measures was based on the T-score metric (mean [SD], 50 [10])²⁴ and item response theory calibrations.^25,26 A higher T score represents more of the concept being measured. For the 5 novel measures, a T score of 50 represents the average score for the US general population with LVADs.¹¹ For the Patient-Reported Outcomes Measurement Information System, an item response theory-based T score of 50 represents the average score for the US general population.²⁷

Regarding missing data, per analyses from another set of patient-reported outcomes for this patient cohort, Hahn et al²⁸ determined that they could use the missingness at random assumption. Six participant groups were created, based on the last completion of measures and reason for missing data, to examine the extent and reasons for missing data.²⁸ An evaluation was made as to whether data were missing completely at random, missing at random, or missing not at random. Comparisons across participant groups, using the χ² test, Fisher exact test, or analysis of variance, were conducted by sociodemographic and clinical characteristics. Hahn et al²⁸ reported that sociodemographic and clinical characteristics were comparable across the 6 participant groups categorized by the time of the last assessment and reason for withdrawal/dropout. In addition, longitudinal models adjusted for missing not at random²⁹ data showed no significant difference compared with missing at random models.

For these analyses, we conducted linear regression analyses using the EQ-5D-3L VAS and KCCQ-12 OSS separately, as dependent variables, within a structural equation modeling framework. With full-information maximum likelihood estimation,^30,31 and on the basis of previous assessment of missingness in this cohort,²⁸ we were able to use all available data and the missingness at random assumption. To assess factors associated with overall HRQOL outcomes, we conducted regressions for each outcome measure (ie, the EQ-5D-3L VAS and KCCQ-12 OSS) at 3 months. Individual factors (ie, independent variables) were first evaluated, controlling for the outcome measure at baseline. Demographic variables included age, race and ethnicity, sex, marital status/domestic partner, and education. Clinical variables included number of comorbidities (eg, chronic kidney disease, pulmonary hypertension, major stroke, peripheral vascular disease, and liver dysfunction), number of rehospitalizations between baseline and 3 months, occurrences of adverse events (bleeding, cardiac dysrhythmia, infection, neurological dysfunction, respiratory problems, venous thrombosis, renal dysfunction, and other events) between baseline and 3 months, New York Heart Association class closest to the date of surgery, and preimplant Interagency Registry for Mechanically Assisted Circulatory Support profile, which further characterizes the severity of heart failure for patients with New York Heart Association class III or IV symptoms being considered for LVAD implantation. There are 7 profiles ranging from 1 = critical cardiogenic shock to 7 = advanced New York Heart Association class III symptoms. Left ventricular assist device preimplant strategy included bridge to transplant, possible bridge to transplant, and destination therapy. Disease- and treatment-related factors included the 5 novel measures and self-reported individual symptoms, administered at 3 postoperative months. Heart failure– and LVAD-related symptoms included any reported occurrence in the past 7 days of (1) swelling in the feet, ankles, or legs; (2) abdominal bloating; (3) poor appetite; (4) pain at the driveline exit site; (5) pain at the chest incision; (6) discomfort from wearing the LVAD equipment; (7) dizziness; and (8) fluid leakage around the driveline exit site. Symptom responses were dichotomized regarding frequency and severity (ie, never vs rarely to almost constantly or none vs mild to very severe), depending on the symptom. Resource variables, measured at baseline, included Patient-Reported Outcomes Measurement Information System instrumental, informational, and emotional support measures and whether financial difficulties resulted from the medical condition or treatment. Individual variables were retained for further evaluation when the P value by Wald test was less than 0.25.³² The retained variables were then included in a final backward stepwise selection regression for each of the 2 outcomes at 3 months. Variables in the final multivariable models were retained when the P value by Wald test was less than 0.05. Mplus version 8.8 was used for all regression modeling. The data sets generated and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.

Results

Sample Characteristics

Of 359 patients with advanced heart failure approached, 272 (76%) enrolled in the Mechanical Circulatory Support: Measures of Adjustment and Quality of Life study (Figure 2). The most frequent reasons for nonenrollment were too anxious, stressed, or depressed (n = 29); too tired (n = 13); too sick (n = 12); and too much trouble (n = 11). Of those who enrolled, 242 (89%) completed self-report measures at baseline, and 142 (52%) completed measures at 3 months postoperatively. The most frequent reasons for noncompletion of measures at baseline were withdrawal (n = 9), not physically well (n = 6), and death (n = 5), and the most frequent (cumulative including baseline) reasons at 3 months were death (n = 37), passive refusal (n = 32), and withdrawal (n = 17).

FIGURE 2.

Flow diagram of cohort. EQ-5D-3L, EuroQol 5-dimension-3L; KCCQ-12, Kansas City Cardiomyopathy Questionnaire-12; VAS, visual analog scale.

For this report, at 3 months, n = 111 completed the KCCQ-12, and of these, n = 108 completed the EQ-5D-3L VAS. Some means, variances, and/or covariances among regressors were excluded from estimation, which resulted in a decrease of sample size for the KCCQ-12 OSS from n = 111 to n = 109, whereas a sample size of n = 108 was retained for the EQ-5D-3L VAS.

Patients were, on average, 55 years old, with 21% female, 24% non-White, 39% high school or lower educated, and 55% married/partnered (Table 2). Baseline clinical characteristics included 73% dilated cardiomyopathy, 97% New York Heart Association class III or IV, 47% destination therapy implantation strategy, and 63% Interagency Registry for Mechanically Assisted Circulatory Support profiles 2 to 3. By 3 months after implantation, 68% of patients had been rehospitalized (including planned and unplanned occurrences), and the most frequent adverse events were venous thrombosis, major infection, cardiac dysrhythmias, and major bleeding (Table 2).

TABLE 2.

Characteristics of Participants

Variable	N Available
Demographic characteristics
Age, mean (SD), y	111	55 (13)
Gender (female), n (%)	111	23 (20.7)
Race (non-White), n (%)	109	26 (23.9)
Ethnicity (Hispanic or Latino), n (%)	110	6 (5.5)
Marital status (married/domestic partner), n (%)	109	60 (55.0)
Education (high school or lower), n (%)	109	43 (39.4)
Baseline clinical characteristics, n (%)
Heart failure etiology	111
Ischemic cardiomyopathy		23 (20.7)
Dilated cardiomyopathy		81 (73.0)
Other		7 (6.3)
New York Heart Association class	111
I		0 (0.0)
II		3 (2.7)
III		25 (22.5)
IV		83 (74.8)
Preimplant strategy	111
Bridge to transplant		35 (31.5)
Long-term implant (ie, destination therapy)		52 (46.8)
Possible bridge to transplant		24 (21.6)
INTERMACS profile	109
Profile 1		5 (4.6)
Profiles 2–3		69 (63.3)
Profiles 4–7		35 (32.1)
Left ventricular ejection fraction	100
>50 (normal)		1 (1.0)
40–49 (mild)		1 (1.0)
30–39 (moderate)		2 (2.0)
20–29 (moderate/severe)		20 (20.0)
<20 (severe)		76 (76.0)
Postimplant clinical characteristics
Adverse events,a mean (SD)	111	1.46 (1.13)
Rehospitalization,b n (%)	111	75 (67.6)
Major infection, n (%)	111	23 (20.7)
Major bleeding, n (%)	111	12 (10.8)
Cardiac dysrhythmia, n (%)	111	13 (11.7)
Renal dysfunction, n (%)	111	4 (3.6)
Hepatic dysfunction, n (%)	111	0 (0.0)
Neuro dysfunction: stroke, n (%)	111	2 (1.8)
Neuro dysfunction: nonstroke, n (%)	111	6 (5.4)
Respiratoryfailure, n (%)	111	4 (3.6)
MCS device malfunction, n (%)	111	1 (0.9)
Venous thromboembolism, n (%)	111	66 (59.5)
Arterial non-CNS thromboembolism, n (%)	111	0 (0.0)
Wound dehiscence, n (%)	111	1 (0.9)
Psychiatric episode, n (%)	111	1 (0.9)

All n (%) are patient counts.

Abbreviations: CNS, central nervous system; INTERMACS, Interagency Registry for Mechanically Assisted Circulatory Support; MCS, mechanical circulatory support.

^aAdverse event within a specified time before completion of the study questionnaires.

^bRehospitalization for any reason within a specified time before completion of the study questionnaires.

Multiple Regression Analyses

The initial disease- and treatment-related factors, LVAD preimplantation strategy, resources, demographics, and clinical variables that were identified (P < 0.25) for each outcome variable (KCCQ-12 OSS, n = 109, and EQ-5D-3L VAS, n = 108) are shown in supplemental materials (see Supplemental Digital Content Tables 1 and 2, http://links.lww.com/JCN/A299, respectively, which describe these potentially significant initial factors). In the multivariable model for the KCCQ-12 OSS, after controlling for baseline KCCQ-12 OSS, higher Satisfaction with Treatment was associated with a higher KCCQ-12 OSS; Being Bothered by VAD Self-care and Limitations, cardiac dysrhythmias within 3 postoperative months, high school or lower education, chest incision pain, and peripheral edema were associated with a lower KCCQ-12 OSS (R² = 0.524) (Table 3). Factors in the multivariable model associated with a lower 3-month EQ-5D-3L VAS, after controlling for baseline VAS, were female sex, adverse events within 3 months post implantation (cardiac dysrhythmias, bleeding, and venous thrombosis), and chest incision pain (R² = 0.229) (Table 4). None of the novel self-report measures were independently associated with the 3-month VAS. Severity of chest incisional pain, a significant factor in both models, was reported as none (57%), mild-moderate (35%), and severe–very severe (8%).

TABLE 3.

Final Multivariable Model With the Kansas City Cardiomyopathy Questionnaire-12 Overall Summary Score as the Outcome

	Estimate	SE	Two-Tailed P
Intercept	51.86	13.54	.000
Baseline
KCCQ-12 summary score	0.40	0.07	.000
VAD bother T score at 3 mo	−0.48	0.16	.003
Satisfaction with Treatment	0.58	0.15	.000
T score at 3 mo
Occurrence of adverse events between baseline and 3 mo
Cardiac dysrhythmia	−11.91	4.24	.005
Education
More than high school	0.00
High school or lower	−10.88	2.68	.000
In the past 7 d (month 3): How often did you have swelling in your feet, ankles, or legs?
Never	0.00
Rarely to almost constantly	−8.14	2.61	.002
In the past 7 d (month 3): What was the severity of pain at your chest incision, at its worst?
None	0.00
Mild to very severe	−7.48	2.75	.006
R2 = 0.524, n = 109

Abbreviations: KCCQ-12, Kansas City Cardiomyopathy Questionnaire-12; SE, standard error; VAD, ventricular assist device.

TABLE 4.

Final Multivariable Model With the EuroQol 5-Dimension-3L Visual Analog Scale as the Outcome

	Estimate	SE	Two-Tailed P
Intercept	74.85	5.66	.000
Baseline
Your own health state today	0.17	0.07	.021
Gender
Male	0.00
Female	−8.39	4.21	.046
Occurrence of adverse events between baseline and 3 mo
Bleeding	−12.49	6.10	.041
Cardiac dysrhythmia	−12.08	5.99	.044
Venous thrombosis	−10.50	4.24	.013
In the past 7 d (month 3): What was the severity of pain at your chest incision, at its worst?
None	0.00
Mild to very severe	−11.13	3.41	.001
R2 = 0.229, n = 108

Abbreviations: SE, standard error.

Discussion

Our findings extend the existing literature assessing factors associated with HRQOL early after implantation, confirm the significance of several variables contributing to HRQOL in our conceptual model, and partially supported our hypothesis. Two novel measures, Satisfaction with Treatment and Being Bothered by VAD Self-care and Limitations, were associated with overall HRQOL at 3 months postoperatively using the KCCQ-12 OSS, thus partially supporting our hypothesis. However, no associations with overall HRQOL were found for 3 of our novel measures (VAD Team Communication, Self-efficacy regarding VAD Self-care, and Stigma). In addition, baseline HRQOL, demographic characteristics, symptoms, and adverse events were significantly associated with 3-month HRQOL using both the EQ-5D-3L VAS and KCCQ-12 OSS, also providing support for our hypothesis. There was no evidence for our hypothesis that LVAD preimplantation strategy, resources (ie, support and financial strain), and rehospitalization were associated with overall HRQOL at 3 months of follow-up.

Association of the 2 novel measures (Satisfaction with Treatment, and Being Bothered by VAD Self-care and Limitations) with overall HRQOL, using the KCCQ-12, provides support for their use as measures of adjustment early after LVAD implantation. Satisfaction with treatment is an important patient-centric cardiac surgical construct.^33,34 The multi-item measure of Satisfaction with Treatment is an improvement over previous measures using only single items or items focused exclusively on LVAD equipment.^9,35 To our knowledge, no measure similar to “Being Bothered by VAD Self-care and Limitations” has been reported in the literature. Grady et al¹³ developed this measure to gain insight into challenges in managing LVAD equipment and limitations in usual activities. For example, one of the items in the “Being Bothered by VAD Self-care and Limitations” measure is being bothered by having to take VAD equipment with you when you leave home for a short time. Regarding patients who respond that they are bothered “quite a bit” or “very much,” clinicians can discuss this hassle regarding ways to reduce being bothered and, at the same time, reinforce safety risks when equipment is not immediately available. The lack of association of all 5 new measures with the EQ-5D-3L VAS may be because the VAS is a single-item generic scale of health state rather than a composite of items within domains as used to create the KCCQ-12 OSS. In addition, the KCCQ-12 is heart failure–specific and may be more sensitive as an outcome than a generic measure. Patients considering LVAD implantation may benefit from shared decision-making discussions that include treatment satisfaction and hassles of living with an LVAD.¹¹ After implantation, use of the 5 new measures, including the 3 that were not associated with HRQOL, may identify individual issues and concerns and guide treatment-related strategies.

Lack of association of overall HRQOL with support is somewhat counterintuitive, because support has been demonstrated by others to be related to HRQOL after LVAD implantation.^36–38 Hahn et al¹² previously reported that correlations of Patient-Reported Outcomes Measurement Information System informational, emotional, and instrumental support with the KCCQ-12 OSS and EQ-5D-3L VAS after LVAD implantation were low, yet mean scores for all 3 support measures were similar to mean scores for the US general population.

Authors of few studies have reported an association of demographic characteristics (ie, sex and education) with HRQOL early after LVAD implantation. In support of our findings, female sex was related to worse HRQOL at 3 months after LVAD implantation in a small international single-center study.³⁹ Contrary to our findings, a report from Interagency Registry for Mechanically Assisted Circulatory Support noted that both men and women experienced significant improvement in overall HRQOL from before to 6 months after LVAD implantation using the EQ-5D-3L VAS, and sex was not associated with change in overall HRQOL from pre-operatively to 6 months postoperatively.³ Yet, in this same article, women reported significantly more problems with some EQ-5D-3L dimensions (ie, usual activities, pain/discomfort, and anxiety/depression) than men both before and after implantation.³ Differences in HRQOL by sex after LVAD implantation are in need of further study.

We found no literature on the association of education with HRQOL after LVAD implantation. Education is associated with other outcomes after advanced surgical therapies. Findings from the United Network for Organ Sharing database and other studies identified worse survival in heart transplantation recipients with lower education.^40–42 There is also some literature, although contradictory, regarding associations of level of education with resource utilization.^43,44 Notably, findings from a recent pilot randomized trial of LVAD self-care found no associations between sociodemographic characteristics and self-care skills performance.⁴⁵ Reasons for the association of lower education with worse outcomes after advanced surgical therapies are unknown. Additional research regarding the association of education and outcomes after advanced surgical therapies is needed.

Our focus was on early recovery after LVAD implantation. Some factors (eg, incisional chest pain) associated with HRQOL may not be relevant later after implantation. Forty-seven percent of patients reported having incisional chest pain. It is no surprise that incisional pain is associated with post-LVAD HRQOL during the early period of surgical recovery and has been reported for other cardiac surgical patients.⁴⁶ Assessments of incisional chest pain and pain control are standard practices in cardiac surgical patient care, including monitoring for persistent incisional pain, which may be related to complications of wound healing.

Associations of adverse events with HRQOL have been previously reported both early and later after LVAD implantation.^3,47 Using the Society of Thoracic Surgeons Interagency Registry for Mechanically Assisted Circulatory Support database, adverse events related to HRQOL at 6 months and 2 to 3 years after LVAD implantation include renal dysfunction, respiratory failure, neurologic dysfunction, and infection.^3,47 Our report increases knowledge of associations of adverse events with HRQOL at a time (ie, 3 months) earlier than reported previously, wherein we found that cardiac dysrhythmias, bleeding, and venous thrombosis were related to HRQOL after LVAD implantation. The association of rehospitalization with overall HRQOL has been previously reported²; our lack of finding an association may be due to inclusion of both planned and unplanned rehospitalizations, or collinearity with adverse events in the models.

Our study has limitations. Most patients who participated in our study were male and White, although these characteristics are similar to Society of Thoracic Surgeons Interagency Registry for Mechanically Assisted Circulatory Support enrollees.³ In addition, our study is observational, and there may be bias due to unmeasured confounding variables. We acknowledge having missing data, including missingness due to death, which may bias our findings, yet previous analyses of patient-reported outcomes using this cohort and our analytic strategy allowed use of all available data.

Conclusions

In addition to other variables, 2 novel patient-reported outcome measures (Satisfaction with Treatment, and Being Bothered by VAD Self-care and Limitations) were associated with overall HRQOL at 3 months after implantation. Our study supports the use of these 2 measures, which may inform shared decision making when patients consider treatment options and, after LVAD implantation, may guide treatment strategies, such as counseling, support, and education. Thus, our novel treatment-specific measures assess potential areas of concern unique to LVAD implantation, which may provide actionable patient care targets.

Supplementary Material

Supplemental digital content is available for this article.

What’s New and Important.

• Two novel patient-reported outcome measures regarding satisfaction with being on a ventricular assist device and how the ventricular assist device negatively impacts patient lives were associated with overall HRQOL early after device implantation.

• Early after ventricular assist device implantation, demographic variables and postoperative symptoms and adverse events were associated with overall HRQOL.