Development and Initial Evaluation of the Psychometric Properties of Self-Efficacy and Adherence Scales for Patients with a Left Ventricular Assist Device

Abstract

Context—

No tools exist to measure patients' self-efficacy for and adherence to the complex home-care regimen after having a left ventricular device (LVAD) implanted.

Objective—

To develop 2 new instruments, the LVAD Patient Self-Efficacy Scale (LPSES) and the LVAD Patient Home Management Adherence Scale (LPHMAS), and evaluate their psychometric properties.

Methods—

This multistage instrumentation study recruited 102 patients (77.5% men and 22.5% women) aged 20 to 82 years, predominantly from the Midwest (34.3%) and the Northeast (26.5%) regions of the United States. Main indications for LVAD were bridge-to-transplant (69.6%) and destination therapy (21.6%), with mean implant duration of 19.9 (SD, 15.5) months. Study participants completed the following instruments: LPSES, LPHMAS, General Self-Efficacy Scale (GSES), Medical Outcomes Study General Adherence (MOSGA), and Self-Care Heart Failure Index (SCHFI) confidence and maintenance subscales. Item analyses, psychometric properties including factorial and convergent validities, and internal consistency reliability were tested.

Results—

Factor analyses showed that the variances for the 20-item LPSES and 9-item LPHMAS were 60.2% and 53.6%, respectively. Convergent validity of the newly developed instruments was supported by the following correlations: LPSES and GSES (r = 0.34); LPSES and SCHFI-confidence (r = 0.60); LPHMAS and MOSGA (r = 0.33); LPHMAS and SCHFI-maintenance (r = 0.40). Internal consistency reliability coefficients were 0.94 (LPSES) and 0.84 (LPHMAS). Based on these data, the LPSES and LPHMAS are valid and reliable measures of self-efficacy and adherence specific for LVAD patients. Confirmatory testing is needed to further support the validity of these instruments for use in research and clinical practice.

Self-efficacy and adherence are widely known behavioral factors that influence treatment and/or health outcomes. Higher levels of self-efficacy and consistent adherence to therapies have been associated with adequate self-care ability, which in turn often results in higher level of functioning, health maintenance, and well-being.^1–4 Self-efficacy is defined as the patient's ability to complete skills or tasks successfully and confidently through a variety of learning experiences, which is enhanced by reinforcing the learned experiences (eg, self-care) through feedback.¹ The World Health Organization defined adherence as “the extent to which a person's behavior, [such as] taking medications, following a diet, and/or executing lifestyle changes, corresponds with agreed recommendations from a healthcare provider.”⁵

The linkages among self-efficacy, adherence, and outcomes in patients with a left ventricular assist device (LVAD) are yet to be explored. However, one can argue that the patient's self-efficacy is acquired before LVAD implant and during the preparation for hospital discharge. While in the hospital, patients are required to complete a comprehensive LVAD education/training program and are expected to demonstrate the essential competencies (knowledge and skills) needed for LVAD care at home. These competencies include, but are not limited to, monitoring and maintaining the workings of the LVAD components, changing sterile dressings, taking medications, and engaging in appropriate physical activities. Patients are also educated about the importance of consistent adherence to the LVAD home-care regimen to prevent complications and maintain health and quality of life.^6–9 Because of the complexity of monitoring and maintaining the LVAD components in the home environment, it is important to be able to measure self-efficacy and adherence specifically in LVAD patients.

The paucity of research on self-efficacy and adherence in LVAD population is due in part to the lack of instruments specifically designed to measure these concepts in patients with LVADs. To address this deficit, 2 instruments were developed for measuring patients' self-efficacy for and adherence to the complex LVAD home-care regimen. This study was aimed to develop and then evaluate the psychometric properties of the newly developed instruments suitable for use in research and clinical practice.

Methods

In this instrumentation study, we employed a 5-stage design, namely item generation, content validation, preliminary analysis, instrument revision, and a final stage focusing on evaluating the validity and reliability of the revised instruments. Approval was obtained from the institutional review board before the implementation of the study.

Stage 1: Item Generation

In this stage, items (questions) for the instruments were generated by 2 different approaches. First, we reviewed and synthesized the pertinent literature about self-efficacy and adherence in the general patient population,^{1–4,10–12} best practices for long-term care of patients with LVADs,^9,13,14 and literature related to self-care management^6,7 and lifestyle adjustments with the LVAD.^15,16 Through this approach, we identified several themes that guided the development of preliminary items for the instruments. For example, we found that we needed to develop items for the instruments such as daily monitoring and recording of LVAD parameters and watching for signs and symptoms of infection at the LVAD exit site, which are specific to this population.^9,13,14 Second, we interviewed clinical experts, which consisted of 2 LVAD nurse practitioners and 2 LVAD nurse coordinators from large VAD centers in the Northeast and Midwest regions of the United States. Each clinical expert had at least 15 years of experience in educating, training, and managing patients with long-term LVADs. This approach (interview) was aimed (a) to elicit the clinical experts' perspectives on the need for developing self-efficacy and adherence measures specific for the patient's self-management of LVAD regimen in their homes, and (b) to confirm the relevance of the items we generated from the literature with the clinical experts' “real-world” experiences. All clinical experts had agreed that we were “asking correct and important questions.” Also, they had indicated that the need for developing instruments for measuring the patient's self-efficacy and adherence is long overdue. Therefore, the research and clinical literature and the clinical experts' appraisal of the relevance and clarity of the items were the basis for constructing preliminary items for the 2 instruments: the LVAD Recipient Self-Efficacy Scale (LRSES) and the LVAD Recipient Home Management Adherence Scale (LRHMAS) version 1.0 (V1.0).

The LRSES-V1.0 (18 items) is a self-administered questionnaire designed for measuring a patient's belief in his/her knowledge and confidence in managing the LVAD home-care regimen. Each item is rated by using a 4-point rating scale ranging from 0 (no confidence) to 3 (high confidence). A summative score is calculated, with higher scores indicating greater knowledge and confidence in the ability to manage the LVAD regimen. The LRHMAS-V1.0 (20 items) is a self-administered questionnaire designed for measuring the degree to which the patient follows the LVAD home-care regimen. Each item is measured by using a 6-point rating scale ranging from 0 (never) to 5 (once a day). A summative score is also obtained, with higher scores indicating greater adherence to the LVAD regimen.

Stage 2: Content Validation

Content of the LRSES-V1.0 and LRHMAS-V1.0 was judged by patients from an LVAD online support group. The judges consisted of 8 men and 2 women, 35 to 76 years old, with educational levels that ranged from high school to master's degree. Seven judges had second-generation LVADs (HeartMate II, Thoratec Corp), and 3 judges had third-generation LVADs (HeartWare Inc). All were discharged from the hospital 3 to 30 months before their participation as judges. Items for LRSES-V1.0 and LRHMAS-V1.0 were evaluated by the judges for relevance and clarity. Judges were asked by the principal investigator (J. Casida) to make comments and/or revise the items (questions) for clarity purposes. Item relevance was appraised with a content validity index (CVI), defined as the proportion of the judges' agreements with each item.¹⁷ Based on the judges' evaluations, items were refined and the number of items was increased from 18 to 19 for the LRSES and 20 to 22 for the LRHMAS. The individual-item CVI for both instruments ranged from 0.80 to 1.00, and the total scale CVI was 0.96 for LRSES and 0.97 for LRHMAS. These CVI values are above the minimum requirement of 0.78 for individual and total scale items evaluated by 10 judges.¹⁷ The readability for content-validated LRSES-V1.0 and LRHMAS-V1.0 was assessed by using the Simple Measure of Gobbledygook readability criteria,¹⁸ which indicated that they could be understood at a fifth grade level of education.

Stage 3: Preliminary Analysis

The aim of this stage was to conduct a preliminary analysis of the LRSES-V1.0 and LRHMAS-V1.0. This aim was accomplished by using the data provided by the patients from 2 LVAD online support groups, closed to public access. Membership in these groups is restricted to patients and family caregivers; each group had more than 900 members as of May 2014. The principal investigator was given permission by the group leaders to access study participants by using a recruitment flyer approved by the institutional review board that was posted on the groups' websites. Inclusion criteria for this stage of the study were the following: (1) male or female sex, all races/ethnic backgrounds, and ages 18 years or older with a minimum education of fifth grade; (2) recipients of a second- or third-generation LVAD as a bridge to transplant, myocardial recovery, or as destination therapy; and (3) have been discharged from the hospital for at least 1 month. Patients were excluded if they did not speak, write, and/or comprehend the English language, or if they had any type of condition (eg, stroke) that could alter their ability to understand instructions and complete a questionnaire.

Potential participants were screened by the principal investigator through telephone interviews. Their names, LVAD type, implant date, and contact information were cross-checked by the principal investigator by using the rosters of the support groups. Out of the 129 screened, 125 met the inclusion criteria and agreed to receive research packets mailed via the US Postal Service. The packets contained instructions, informed consent, measurement instruments, other data collection tools (eg, demographics), and a self-addressed envelope for return within 30 days of receipt. Two telephone calls within a 1-month period were conducted as reminders for participants who had not returned the packets by 30 days after the mailing. Two months later, 102 (81.6 %) of the 125 patients had returned the packets with completed data.

Study participants at this stage of the study (n = 102) were predominantly white (78%), male sex (66%), mean age of 53.0 (SD, 13.8) years, married (75%), and had “some college” education (45%). About 87% of the 102 patients had a second-generation LVAD as a bridge to transplant (69%) or destination therapy (22%) with a mean implant duration of 19.5 (SD, 14.6) months. Results of the item analyses for LRSES-V1.0 and LRHMAS-V1.0 showed overwhelmingly low and/or negative correlations between the items, which warranted revisions before the next stage of psychometric testing.

Stage 4: Instrument Revision

Review of the instruments consisted of 3 steps: editing questions, changing response scales, and cognitive interviewing. A careful review of all the instruments revealed that some questions were vague and/or potentially had multiple answers. Accordingly, questions were revised for clarity and some were split into 2 items or questions. The directions for completing the instruments and the response scales were revised. For the LRSES-V1.0, the intensity of each item was changed from a 4- to a 6-point response scale using extreme opposite anchors ranging from 0 (not confident at all) to 5 (extremely confident). Although the response scale for LRHMAS-V1.0 was retained, we found that many of the original items could not be appropriately answered by “never” or “once a day.” Thus, extreme opposite anchors were also used, ranging from 0 (never) to 5 (all the time).

Once the items and response scales were revised, the principal investigator used a cognitive interviewing technique¹⁹ to refine the instruments further. Ten randomly selected participants (6 males and 4 females, aged 21 to 76 years) from stage 3 of the study were invited by the principal investigator to review and assess the revised instruments via telephone interviews. Participants were asked to read the directions and questions carefully, at least twice, before selecting their answers. After completion of the 2 instruments, the principal investigator interviewed the participants and asked them to explain their understanding of the questions and their answers that were randomly presented to them. Finally, they were asked to edit the questions and make recommendations.

The participants' general appraisal of the revised instruments were positive, including comments such as “much clearer,” “easier to read,” and “better questions” compared with the first version. Also, the participants recommended changing the word “recipient” to “patient.” The content was revised and the titles of the 2 instruments were changed to LPSES and LPHMAS version 2.0 (V2.0). Evaluations of the reliability and validity of the LPSES-V2.0 (26 items) and LPHMAS-V2.0 (24 items) were conducted in the final stage of this study.

Stage 5: Psychometric Testing

Participants in stage 3 of the study were invited to participate in this stage of psychometric testing. Of the 102 patients, 87 (85.3%) agreed to participate and provided informed consent. The 15 patients who did not participate either received a heart transplant (n = 7), had a stroke (n = 3), died (n = 2), or refused to participate again (n = 3). Therefore, 15 new patients were recruited from the same online support groups by using the same study procedures described in stage 3.

The specific aims for this stage of the study were to perform an item analysis and to estimate the validity and internal consistency reliability of the revised scales (LPSES-V2.0 and LPHMAS-V2.0). Item analysis of each scale (interitem and item-total correlations) was evaluated by Pearson correlations. Exploratory factor analyses, principal axis factoring with oblimin rotation, were performed to evaluate the construct validity of the instruments. We proposed that each instrument would produce a 1-factor measurement model. Convergent validity was evaluated by the correlations among the revised scales and established scales that measure similar concepts of self-efficacy and adherence in other populations of patients.

The established scales included the General Self-Efficacy Scale (GSES),^10,20 Medical Outcomes Study General Adherence (MOSGA),¹¹ and Self-Care Heart Failure Index (SCHFI)-V6.2, confidence and maintenance subscales.^21,22 The GSES is a general measure of self-efficacy used to evaluate a sense of the person's competence including problem solving in stressful situations. The GSES consists of 10 items using a 4-point rating scale ranging from 1 (not at all true) to 4 (exactly true) with possible sum scores of 10 to 40. Higher scores indicate greater self-efficacy. The GSES is a reliable (α > 0.75) and valid measure of general self-efficacy that has been used in research involving thousands of patients with a variety of health conditions or illnesses from different countries.^10,20

The MOSGA is used to evaluate the person's tendency to adhere to medical treatments. It consists of 5 items and uses a 6-point rating scale ranging from 1 (none of the time) to 6 (all of the time). Higher scores indicate greater adherence to medical recommendations. Test-retest reliability of the MOSGA ranged from α = 0.40 to 0.81. The validity of MOSGA was supported by factor analytics in patients with chronic medical conditions.¹¹ The SCHFI-V6.2 is used to measure the adequacy or inadequacy of self-care in patients with heart failure. For the purposes of this study, we used the SCHFI confidence (6 items) and maintenance (10 items) subscales as another convergent validity measure for self-efficacy and adherence, respectively. Higher self-care confidence and maintenance scores imply adequate patient's self-care capability. The SCHFI V6.2 is a sufficiently reliable (α = 0.76) and valid (comparative fit index = 0.73)^21,22 instrument used extensively in heart failure research.

The internal consistency reliability of the LPSES-V2.0 and LPHMAS-V2.0 was estimated with the Cronbach α. Study participants' demographics, clinical characteristics, and information about their LVAD education/training also were collected. Descriptive and psychometrics data were analyzed by using IBM SPSS 20.0 software. Results of the final stage of this study are presented and discussed in the remaining sections of this article.

Results

Study Participants

Table 1 summarizes the demographic and clinical characteristics of the 102 patients who participated in the final stage of the study. Participants were predominantly white (77.5%), male (77.5%), aged 20 to 82 years (mean [SD], 51.4 [13.8]), from the Midwest (34.3%) or Northeast (26.5%) regions of the United States. Almost 72% of the participants were educated beyond high school, and they were retired (20.6%), on disability (53.9%), or unemployed (5.9%) with a mean annual household income of $55 750 (SD, $35 205.11). Slightly more than 85% of them had a second-generation LVADs as a bridge to transplant (69.6%) or destination therapy (21.6%) and were living with someone at home (87.2%) for at least 2 months after hospital discharge. About half (47%) had nonischemic cardiomyopathy and 31.4% had ischemic cardiomyopathy. All of them had at least 1 coexisting health condition (Table 1).

Table 1

Patients' demographics and clinical profile (n = 102)

Characteristic	Value^a
Age, mean (SD), y	51.4 (13.8)

Sex
Male	79 (77.5)
Female	23 (22.5)

Marital status
Married	74 (72.5)
Single	16 (15.7)
Divorced	8 (7.8)

Race/ethnicity:
White	79 (77.5)
Black	16 (15.7)
Mixed	3 (2.9)
Asian	2 (2.0)
Hispanic (nonwhite)	1 (1.0)
Native American	1 (1.0)

Education
Some college	44 (43.1)
College and beyond	29 (28.4)
High school	24 (23.5)
Less than high school	5 (4.9)

Regional origin
Midwest	35 (34.3)
Northeast	27 (26.5)
Southeast	17 (16.7)
Southwest	12 (11.8)
West	11 (10.8)

Employment status
On disability	55 (53.9)
Retired	21 (20.6)
Full-time	15 (14.7)
Unemployed	6 (5.9)
Part-time	5 (4.9)

Type of left ventricular assist device (LVAD)
Second-generation (HeartMate II)	87 (85.3)
Third-generation (HeartWare)	15 (14.7)

Duration of LVAD, mean (SD), months	19.9 (15.5)

Indications for LVAD therapy
Bridge to transplant	71 (69.6)
Destination therapy	22 (21.6)
Bridge-to-myocardial recovery	9 (8.8)

Durations of heart failure before LVAD, mean (SD), years	8.0 (6.6)

Cause of heart failure
Nonischemic cardiomyopathy	48 (47.1)
Ischemic cardiomyopathy	32 (31.4)
Unknown	22 (21.6)

Implantable cardiac defibrillator/pacemaker	82 (80.4)

Comorbid conditions
Hypertension	50 (49.0)
Cardiac arrhythmias	38 (37.3)
Obesity	34 (33.3)
Diabetes	31 (30.4)
Hypercholesterolemia	33 (32.4)
Sleep apnea	27 (26.5)
Gout	26 (25.5)
Coronary artery disease	22 (21.6)
Anemia	20 (19.6)
Arthritis	15 (14.7)
Cancer	10 (9.8)
Heart valve disease	10 (9.8)
Kidney disease	10 (9.8)
Pulmonary hypertension	8 (7.8)
Chronic obstructive pulmonary disease	7 (6.9)
Asthma	4 (3.9)

Values are number (percentage) of patients unless otherwise indicated. Because of rounding, missing or “not applicable” data, or multiple responses, not all percentages total 100.

Most participants reported that they were required to complete an LVAD home-care education/training program before discharge (Table 2). Hands-on training (93.1%) and skill demonstration (82.4%) were commonly used methods of instruction. The LVAD nurse practitioner or coordinator evaluated patients' readiness for discharge via skill demonstration (90.2%) or a quiz (61.8%). The mean duration for the entire program and evaluation for discharge was 13.6 (SD, 14.3) hours. On the day of discharge, 91.2% were provided with binders containing videos and manuals, logs for medications, vital signs, and LVAD parameters, among other materials (Table 2).

Table 2

Education and training of 102 patients about left ventricular assist device (LVAD)

Education and training	No. (%)^a
Education materials/methods
Hands-on experience with LVAD equipment	95 (93.1)
Skill demonstration	84 (82.4)
Watched LVAD video	55 (53.9)
Attended lecture	31 (30.4)

Evaluation of learning
Return demonstration of skill	92 (90.2)
Quiz	63 (61.8)
Quiz and return demonstration of skill	29 (28.4)

Discharge education materials (take home)
LVAD manual	91 (89.2)
LVAD discharge binder and video	93 (91.2)
LVAD video	52 (51.0)

Because of rounding, missing or “not applicable” data, or multiple responses, not all percentages total 100.

Item Analysis

Analyses of the 26-item LPSES-V2.0 and the 24-item LPHMAS-V2.0 showed that 7 items of the LPSES had interitem correlations less than 0.04 and/or equal to −0.04 and 13 items of the LPHMAS had interitem correlations less than 0.02 and/or equal to −0.02. These items with very low correlations were subsequently removed from the item pool, yielding a total of 19 items for LPSES-V2.0 and 11 items for LPHMAS-V2.0.

Validity and Reliability

LPSES-V2.0. Results of factor analyses revealed a 2-factor solution accounting for 60.2% of the variance of the 19 items on the LPSES. Eigenvalues were 10.48 (factor 1) and 1.67 (factor 2), with factor loadings ranging from 0.31 to 0.94. The small eigenvalue of factor 2 suggested a potential 19-item 1-factor measurement model. Item 1 was re-entered into the item pool, resulting in a total of 20 items for the next analysis. Item 1 was re-entered into the pool because of its importance with respect to infection control such as the procedure of changing sterile dressings at the LVAD exit site.

A principal-axis factoring analysis was performed to force a 1-factor solution as originally proposed. This resulted in a 1-factor solution accounting for 51.0% of the variance of the 20 items with an eigenvalue of 10.66. The 1-factor model of the 20 items is further supported by factor loadings of 0.41 to 0.82 (Table 3). Item-total correlations of the 20-item LPSES ranged from 0.40 to 0.75. The correlation between LPSES and GSES was r = 0.34, P = .001, and the correlation between LPSES and SCHFI-confidence was r = 0.60, P < .001. The internal consistency reliability coefficient of the LPSES-V2.0 was α = 0.94.

Table 3

One-factor pattern matrix with factor loadings of the 20-item Left Ventricular Assist Device (LVAD) Patients Self-Efficacy Scale (version 2.0)^a

Item number	Item description	Factor 1
23	Knowing the prescribed diet (eg, low sodium, low fat, foods rich in vitamin K)	0.82
25	Knowing what to do if I cut myself, or I am bleeding from my nose, mouth, ears, and/or have blood in my stools	0.80
19	Knowing how to communicate with the VAD team about problems related to heart failure and/or LVAD function	0.79
20	Knowing what to do during emergency situations related to LVAD (eg, controller malfunction, pump failure)	0.78
13	Understanding the meaning of yellow and red alarms	0.78
14	Knowing what to do with yellow and red alarms	0.78
26	Knowing what household things I should avoid such as being close to computer screens, using vacuum cleaners, or things that create static	0.78
15	Responding appropriately to the absence of LVAD flow on the screen	0.76
11^b	Understanding the meaning of an abnormal LVAD speed, flow, power, and PI displayed on the screen	0.75
9	Knowing how to check the proper operation of all LVAD equipment	0.74
12^b	Knowing what to do with abnormal LVAD speed, flow, and power, and PI displayed on the screen	0.74
24	Knowing when and how to do exercise and other types of physical activity	0.74
7	Using the power module (HeartMate) or AC/DC adapter (HeartWare)	0.73
18	Knowing how to secure the driveline	0.69
22	Engaging in sexual activity	0.66
21	Knowing what to do during extreme weather conditions	0.65
6	Recognizing signs of infection on the driveline	0.63
16	Knowing the physical activity limitations related to LVAD (eg, contact sports, swimming)	0.59
5	Knowing the side effects of blood thinners (eg, warfarin [Coumadin] or other blood thinners)	0.53
1	Changing sterile dressings	0.41

Abbreviation: AC/DC, alternating current/direct current.

Extraction method: principal axis.

Patient judges for content validity of the instrument recommended by using the “PI,” which is the abbreviation for pulsatility index.

LPHMAS-V2.0. Factor analytic procedures showed a 3-factor solution of the 11 items of LPHMAS accounting for 55.4% of the variance with eigenvalues of 4.81 (factor 1), 1.28 (factor 2), and 1.24 (factor 3). Because of the smaller eigenvalues for factors 2 (1.28) and 3 (1.24), a 2-factor solution was suspected. Thus, a principal-axis factoring was performed to force a 2-factor solution and to examine the possibility of a 2-factor measurement model. The resultant 2-factor solution accounted for 48.3% of the variance with eigenvalues of 4.81 (factor 1) and 1.28 (factor 2). Factor loadings of the 2-factor solution ranged from 0.43 to 0.97, except for the 2 items showing factor loadings of 0.19 and 0.20 that were subsequently removed from the item pool. Although an item about adherence to sterile dressing change procedure was not included in the first round of factor analyses, because of its low interitem correlation, we reentered such an item in the pool because of its importance with respect to infection control. A total of 10 items was deemed to be appropriate for further analysis by using principal-axis factoring with oblimin rotation. This analysis revealed a 2-factor solution accounting for 50.5% of the variance of the 10 items of LPHMAS with eigenvalues of 4.49 (factor 1) and 1.28 (factor 2).

Upon examination of the meaning of the items loaded in each factor, we found that an item loaded in factor 2 appeared to be problematic because it was not consistent with the theme related to the technical aspect of LVAD care. Thus, the item was removed, yielding a total of 9 items for LPHMAS-V2.0. A factor analysis of the 9 items resulted in a 2-factor solution accounting for 53.6% of the variance with eigenvalues of 4.31 (factor 1) and 1.13 (factor 2) and with factor loadings ranging from 0.32 to 0.96 (Table 4). The correlation between factor 1 and factor 2 was r = −0.57. Interitem correlations among the 9 items of LPHMAS ranged from r = 0.31 to r = 0.72. The correlation between LPHMAS and MOSGA was r = 0.40, P < .001, and the correlation between LPHMAS and the SCHFI maintenance was r = 0.33, P = .002. Internal consistency reliability of the LPHMAS-V2.0 was α = 0.84.

Table 4

Two-factor pattern matrix with factor loadings of the 9-item Left Ventricular Assist Device (LVAD) Patient Home Management Adherence Scale (version 2.0)^a

		Factor
Item number	Item description	1	2
10	I check the driveline for signs of damage like cuts, holes, or tears daily	0.85
11	I inspect all cable connectors for dirt or damage daily	0.83
9	I avoid bending, kinking, or twisting the driveline	0.79
7	I avoid pulling or moving the driveline	0.56
1^b	I check the LVAD speed, flow, and PI (HeartMate) or power (HeartWare) daily	0.52
6	I keep the driveline clean and dry	0.49
3	I follow the steps in changing a sterile dressing on the exit site of the driveline	0.32
14	I follow the steps in performing controller self-test (skip this item if you have a HeartWare)		−0.96
13	I follow the steps in performing a power module (HeartMate) or AC/DC adapter (HeartWare) test to make sure it's working properly		−0.96

Abbreviation: AC/DC, alternating current/direct current.

Extraction method: principal axis; rotation method: Oblimin with Kaiser normalization.

Patient judges for content validity of the instrument recommended by using the “PI,” which is the abbreviation for pulsatility index.

Discussion

This study is the first to report the psychometric properties of 2 self-administered instruments specifically designed for measuring self-efficacy and adherence concepts in LVAD patients. Significant findings consisted of instruments producing a 1-factor (20-item LPSES-V2.0) and a 2-factor (9-item LPHMAS-V2.0) measurement models. The data suggest that these instruments are both sufficiently valid and reliable measures of the patient's self-efficacy and adherence within the context of LVAD care management in home settings.

Validity and Reliability of the LPSES-V2.0

The LPSES-V2.0 (20 items) is a 1-dimensional instrument that can be used for measuring the level of self-efficacy in patients with second- or third-generation LVADs. Factor loadings of each of the 20 items of the LPSES exceeded the minimum requirement for a new instrument (Table 3).^23,24 Notably, the Kaiser-Meyer-Olkin (KMO) of 0.89 with a significant χ² value of the Bartlett test of sphericity are suggestive of sample adequacy and appropriateness of use of factor analyses.²³ Correlations (r = 0.34 to r = 0.60) among the LPSES, GSES, and SCHFI-confidence are evidence of convergent validity. These low to moderate correlations are desirable coefficients for new instruments,^24–26 suggesting that the LPSES is not duplicating the other measures of self-efficacy (GSES and SCHFI). Collectively, the findings of factorial construct validity and convergent validity analytics and the high internal consistency reliability (α = 0.94)²⁵ supported the validity and reliability of the LPSES. Overall, the data showed that the 20-item LPSES-V2.0 measures the concept of self-efficacy specifically for patients managing LVADs in home settings.

Validity and Reliability of the LPHMAS-V2.0

The LPHMAS-V2.0 (9 items) is a 2-dimensional instrument that can be used for measuring the degree of patients' adherence to the management of second- or third-generation LVADs. The 2-factor solution of the LPHMAS-V2.0 is supported by sample adequacy and appropriateness of factor analyses, demonstrated by a KMO of 0.78 and a significant χ² value of the Bartlett test of sphericity.²³ Notably, the factor loadings of the items of the LPHMAS (Table 4) have met the required factor loading values for new instruments.^23,24 Items loaded in factor 1 imply an absolute task factor, whereas the items loaded in factor 2 imply a relative task factor (Table 4). The label “absolute” for factor 1 was based on the best practices, manufacturer recommendations, and clinical experience that the tasks listed under factor 1 must be followed by the patient whether he/she has a second- or third-generation LVAD.^9,13,14 The label “relative” for factor 2 is determined by the type of the LVAD that the patient has. For example, item 14 does not apply to third-generation LVAD, and the terms used for item 13 are specific to the type of LVAD.^13,14 Thus, the optional item 14 and device-specific item 13 may have explained the negative factor loading values for factor 2. The 2 subscales of the LPHMAS-V2.0 should be scored, interpreted, and reported separately.

Low correlations (r = 0.33 and r = 0.40) among the LPHMAS, MOSGA, and SCHFI-maintenance are indicative of convergent validity. These low correlations are desirable coefficients for new instruments,^24–26 suggesting that the LPHMAS does not duplicate the other measures of adherence (MOSGA and SCHFI) in other patient populations. Findings of the factorial and convergent validity and the relatively high internal consistency reliability of the 9 items (α = 0.84)²⁵ collectively support the validity and reliability of the LPHMAS. Overall, the data showed that the 2-factor LPHMAS-V2.0 (9 items) measures the concept of adherence of patients managing the LVAD in home settings.

Limitations and Implications

The limitations of the present study are related to methodological issues inherent to any type of self-report administered via surveys.^27,28 We consider using the Internet for recruitment, despite its becoming a common medium for recruiting participants in social sciences research,²⁷ to be another limitation. True to any self-report,²⁷ participants' responses may have had some biased responses, for example, those patients with “live-in” family caregivers who may have helped patients respond to the questions. These limitations will be addressed in future studies using face-to-face recruitment in several VAD centers across the United States. In-person interviews will be audiotaped, and the influence of the participant's social desirability on an item(s) will be examined. Additionally, a random sampling and hypothesis-based instrumentation testing using confirmatory factor analysis and other types of validity testing such as criterion-related validity (using external measurement criteria including LVAD knowledge assessment quiz and skill demonstration) will be implemented in future studies. These studies are paramount to further support the validity of the LPSES-V2.0 and LPHMAS-V2.0 as 1- and 2-dimensional measures, respectively.

Given the data supporting the validity and reliability of the LPSES-V2.0 and LPHMAS-V2.0, these instruments are ready for use in research involving patients living with a second or a third generation of LVAD as a bridge to transplant or as destination therapy in the United States. The instruments were developed on the basis of LVAD education/training, discharge preparations, LVAD self-care management, and long-term follow-up that are reflective of the customary care provided for these patients in VAD centers across the United States.^7,8 Notably, the demographics and clinical characteristics of the patients who participated in this study are consistent with the LVAD population in the United States.²⁹ Therefore, the study findings are not applicable outside the United States with different health care systems, cultures, and/or languages. Furthermore, the LPSES-V2.0 and LPHMAS-V2.0 are subject to revisions in concert with the changes in LVAD technology and the data that will be generated from future psychometric studies. Despite the limitations, these instruments offer investigators the tools needed for empirical research to explore the influence of self-efficacy and adherence on outcomes of patients' self-care management of LVADs in home settings. Investigators are also encouraged to examine the difference, if any, in the self-efficacy and adherence measurement outcomes between patients using the devices as a bridge to transplant and patients using an LVAD as destination therapy; the LPSES and the LPHMAS were developed on the basis of empirical and experiential knowledge that the knowledge and skill required for LVAD patient home-care management are the same for both types of patients.^{7–9,13–15}

Conclusion

The data obtained from this research demonstrated the first evidence of the psychometric properties of LPSES-V2.0 and LPHMAS-V2.0. These instruments provide researchers with sufficiently valid and reliable measures to scrutinize the impact of certain patients' behaviors (self-efficacy and adherence) on self-care capability and LVAD outcomes. Confirmatory testing is the next step to support the validity of the measures further. These instruments are crucial for advancing the science in LVAD self-care management by helping clinicians identify critical behavioral issues (eg, low self-efficacy and/or poor adherence) related to LVAD home care so that interventions can be tailored to improve outcomes.

Footnotes

This research was supported by funding from the University of Michigan Office of the Vice-President for Research faculty grant and a Sigma Theta Tau International Rho Chapter research grant.

Calendar

References

Zulkosky

. Self-efficacy: a concept analysis. Nurs Forum. 2009;44(2):93–102. doi:10.1111/j.1744-6198.2009.00132.x.

Jones

Razi

. Self-efficacy and self-management after stroke: a systematic review. Disabil Rehabil. 2011;33(10):797–810. doi:10.3109/09638288.2010.511415.

Berben

Dobbels

Engberg

Hill

De Geest

. An ecological perspective on medication adherence. West J Nurs Res. 2012;34(5):635–653. doi:10.1177/0193945911434518.

Wouters

Bouvy

Van Geffen

Gardarsdottir

Stiggelbout

Van Dijk

. Antidepressants in primary care: patients' experiences, perceptions, self-efficacy beliefs, and nonadherence. Patient Prefer Adherence. 2014;8:179–190. doi:10.2147/PPA.S53748.

World Health Organization. Setting the scene: Chapter 1 defining adherence. 2003:17–19. http://www.who.int/chp/knowledge/publications/adherence_Section1.pdf. Accessed October 6, 2014.

Casida

Peters

Magnan

. Self-care demands of persons living with an implantable left-ventricular assist device. Res Theory Nurs Pract. 2009;23(4):279–293.

Widmar

Dietrich

Minnick

. How self-care education in ventricular assist device programs is organized and provided: a national study. Heart Lung. 2014;43(1):25–31. doi:10.1016/j.hrtlng.2013.09.003.

Casida

Ilaqua

. A survey of nurses in the mechanical circulatory support programs in the United States. Heart Lung. 2011;40(4):e103–e111. doi:10.1016/j.hrtlng.2010.12.001.

Feldman

Pamboukian

Teuteberg

. The 2013 International Society for Heart and Lung Transplantation Guidelines for mechanical circulatory support: executive summary. J Heart Lung Transplant. 2013;32(2):157–187. doi:10.1016/j.healun.2012.09.013.

10.

Luszczynska

Scholz

Schwarzer

. The general self-efficacy scale: multicultural validation studies. J Psychol. 2005;139(5):439–457.

11.

Hays

. The medical outcomes study (MOS) measures of patient adherence. http://www.rand.org/content/dam/rand/www/external/health/surveys_tools/mos/mos_adherence_survey.pdf. Accessed October 6, 2014.

12.

Raebel

Schmittdiel

Karter

Konieczny

Steiner

. Standardizing terminology and definitions of medication adherence and persistence in research employing electronic databases. Med Care. 2013;51(8 suppl 3):S11–S21. doi:10.1097/MLR.0b013e31829b1d2a.

13.

Thoratec Corporation. Heart Mate II Left Ventricular Assist System. Patient Handbook. Pleasanton, CA: Thoratec Corporation; 2012.

14.

HeartWare, Inc. HeartWare Ventricular Assist System. Instruction for Use. Miami Lakes, FL: HeartWare, Inc; 2012.

15.

Casida

Marcuccilli

Peters

Wright

. Lifestyle adjustments of adults with long-term implantable left ventricular assist devices: a phenomenologic inquiry. Heart Lung. 2011;40(6):511–520. doi:10.1016/j.hrtlng.2011.05.002.

16.

Sandau

Hoglund

Weaver

Boisjolie

Feldman

. A conceptual definition of quality of life with a left ventricular assist device: results from a qualitative study. Heart Lung. 2014;43(1):32–40. doi:10.1016/j.hrtlng.2013.09.004.

17.

Lynn

. Determination and quantification of content validity. Nurs Res. 1986;35(6):382–385.

18.

Readability Formulas. SMOG readability formula. http://www.readabilityforumulas.com/smog-readability-formula.php. Accessed October 6, 2014.

19.

Jobe

Mingay

. Cognitive research improves questionnaires. Am J Public Health. 1989;79(8):1053–1055.

20.

Schwarzer

Jerusalem

. Generalized self-efficacy scale. In: Weinman

Wright

Johnston

, eds. Measures in Health Psychology: A User's Portfolio. Causal and Control Beliefs. Windsor, United Kingdom: Nfer-Nelson; 1995:35–37.

21.

Vellone

Riegel

Cocchieri

. Psychometric testing of the Self-Care of Heart Failure Index Version 6.2. Res Nurs Health. 2013;36(5):500–511. doi:10.1002/nur.21554.

22.

Reigel

. SCHFI V 6.2 self-care heart failure index. http://www.self-careofheartfailureindex.com. Accessed October 6, 2014.

23.

Pett

Lacey

Sullivan

. Making Sense of Factor Analysis. Thousand Oaks, CA: Sage; 2003.

24.

DeVellis

. Scale Development: Theory and Applications. Los Angeles, CA: Sage; 2012.

25.

Nunnally

Bernstein

. Psychometric Theory. McGraw Hill Series in Psychology. New York, NY: McGrawHill; 1994.

26.

Frank-Stromborg

Olsen

. Instruments for Clinical Health-Care Research. Sudbury, MA: Jones and Bartlett; 1997.

27.

Waltz

Strickland

Lenz

. Measurement in Nursing and Health Research. New York, NY: Springer; 2010.

28.

Van Selm

Jankowski

. Conducting on-line surveys. Qual Quant. 2006;40:435–456.

29.

Interagency Registry for Mechanically Assisted Circulatory Support. ITERMACS statistical summaries. https://www.uab.edu/medicine/intermacs/research/statistical-summaries. Accessed October 6, 2014.