Quality of Life Outcomes Following Stroke for People with Aphasia: Perceptions of Patients, Caregivers, and Speech-Language Pathologists

Abstract

Introduction: Due to the inherent difficulty in obtaining information using language-based questionnaires, research on the quality of life (QOL) for persons with aphasia (PWAs) is lacking. The purpose of this study was to investigate which factors have an impact on a PWA and to investigate the reliability of using proxy responses from either spouse/caregivers or speech-language pathologists (SLPs). Methods: A QOL scale was developed with 11 factors identified in the literature as having an impact on the QOL of PWAs. Data were collected through interviews with 31 PWAs and 26 spouse/caregivers recruited from a community aphasia therapy center. Thirty-six SLPs responded to an online survey. Results: Vocation was identified by PWAs as having the most impact on their QOL and family as having the least impact. Cronbach’s alpha and intraclass correlation coefficients (ICCs) showed decreased internal consistency and reliability across ratings and respondents. Conclusions: Overall results suggest that proxy responses may not accurately reflect the factors that impact the QOL of a PWA but that spouse/caregiver responses may be more representative than those of SLPs.

1. Introduction

Stroke is one of the leading causes of death [1] and permanent impairment for adults [2]. Worldwide, nearly 200 of 100,000 people per year suffer a stroke [3]. Due to the improved emergency structure and the increased number of stroke units, there is an increased likelihood of surviving even severe stroke [4]. This increased survival, however, results in an increased number of people with severe disabilities. In fact, it has been estimated that 64% of persons who survive the first year post-stroke will need care for the rest of their lives [1]. Furthermore, 25% are still dependent on someone else for everyday activities after one year, and one third will have suffered a second stroke within five years [3]. The complexity of impairments impacts patients’ home, work, and family situations. Naturally, it is at this time when the patient, the families, and the caregivers fully experience the impact of the stroke [5].

One of the most obvious impairments following a stroke is reduced mobility [6], but non-visible impairments are frequent as well. These may include executive dysfunctions [7] and restrictions in memory, orientation, attention, and communication [8]. Healthcare professionals working with those who have had a stroke and their families should be mindful of these non-visible impairments during all discussions about rehabilitation, participation, and quality of life.

1.1. Impact of Stroke on QOL

Quality of life (QOL) is a multidimensional and subjective concept that is usually described by physical, functional, emotional, and social factors [9]. Research on QOL is important, as it is named as one of the main outcome factors for rehabilitants in measuring the success of their rehabilitation processes. With the introduction of the International Classification of Functioning, Disability and Health (ICF), the medical field has concluded that not only body structural and functional aspects have an influence on the health state of people but also social and personal aspects like the person’s character or their family and work situation. The primary components of the ICF include contextual factors, which act as either facilitators or barriers [10]. Due to this, a rehabilitation process should focus on the whole person and their situation in order to increase rehabilitative success by increasing the use of facilitators and decreasing factors identified as barriers to success.

Indeed, the research is replete with studies showing the impact of stroke on QOL as well as the importance of psychosocial issues during rehabilitation. For example, Abubakar and Isezuo [11] used the Modified Rankin Scale, measures of depression, and the Stroke Impact Scale to assess the health-related quality of life (HRQOL) of persons who had had a stroke. The results showed that functional status and depression influence the HRQOL [11]. In addition, Owolabi [12] studied the effect of stroke on HRQOL measures among diverse cultures by comparing outcomes in Germany and Nigeria. The results showed that persons who had experienced a stroke had poorer HRQOL scores across all domains in both countries, with the physical domain being impacted the most [12].

In addition to the above factors influencing HRQOL and QOL following a stroke, the ICF also identifies the family as a contextual factor that may act as either a facilitator or a barrier to rehabilitative success. Often, the family support role is relegated to the spouse of the person who had a stroke. Intuitively, a positive spousal relationship can serve as a facilitator to success in rehabilitation, while a negative situation may have a negative impact on the patients’ QOL as well as the entire rehabilitation process. For example, Visser-Meily et al. [13] investigated the spousal psychosocial functioning of patients with stroke from the onset of rehabilitation until three years post-stroke. The results included 211 couples and showed a significant effect of time for all four aspects of the psychosocial functioning of spouses [13].

Additionally, Carlsson and colleagues [14] studied the impact of stroke on life satisfaction from the perspective of a patient with a mild stroke, as well as the perspective of a spouse, one year post-stroke. Results showed that patients were generally less satisfied than spouses in several domains [14]. Conversely, the spouses were significantly less satisfied with their relationship than the patients [14]. Spouses who worked were significantly more satisfied with their ability in self-care and their vocational/occupational situation, and the most common symptom named by both groups was mental fatigue [14].

1.2. Impact of Aphasia on QOL

Aphasia is known as one of the most common symptoms following stroke [15] and results from damage to those parts of the brain related to language; it can cause problems with any or all modalities of language [16]. Currently, there are an estimated 2 million people in the United States living with aphasia, and it has been estimated that 38% of people who have a stroke will also have aphasia [17]. In addition, it has also been estimated that 1 in 250 people in the United States currently has aphasia [18].

While most recover from aphasia [15], there is a great deal of variability. In fact, nearly one third of people who have had a stroke recover from aphasia within 12 to 18 months post-onset, while about 60% still suffer from aphasia [19]. Furthermore, the first weeks may have prognostic value, as this time frame has been associated with extent of recovery and degree of long-term disability [20]. Clearly, therapy provided by qualified healthcare professionals can also have an impact on the recovery pattern, but a stroke that results in aphasia as a concomitant disorder does have a negative impact on a person’s QOL [21,22,23]. The problem with the bulk of this research, however, is that persons with aphasia, in particular those with severe aphasia, were typically excluded (e.g., [11,12,24,25]) or that responses were completed via a proxy (e.g., [26,27]).

1.3. Difficulties of QOL Studies and Aphasia

As described, aphasia often results from damages of the left cerebral hemisphere, which have an influence on a person’s language [16]. As a consequence of these damages, all or some linguistic levels may be affected, including lexical, phonological, morphological, syntactical, and pragmatic [15,28]. Additionally, PWAs often present with symptoms such as literal paraphasias, deletion or substitutions of phonemes, and verbal paraphasias. PWAs who have word-finding problems (anomia) often have omissions and substitutions in their continuous speech. Others, however, may unconsciously use neologisms or perseverations, like the constant repetition of words or syllables, to compensate for lost language skills. Finally, other symptoms associated with aphasia are reduced auditory comprehension and/or reduced orthographic comprehension, among others [28].

Quality of life, as a construct, is clearly subjective. Because of this, it is imperative that QOL measures be collected from the person or the patient [9]. However, due to the concomitant language impairments often associated with stroke, PWAs are often excluded from QOL studies. In cases that they are included, the data collection is often incomplete or based on a questionnaire completed by their significant others (proxy responses) [29]. In cases where PWAs are included in QOL studies, the design of the studies often does not mention the complexity of the questionnaires and how the PWAs are able to handle them [30]. Based on the estimation of Cella [9] regarding the particularity of QOL as a highly subjective concept, it is recommended that these studies be done by self-reports from PWAs, which may be more valid than proxy reports [31]. Clearly, there are inherent difficulties in obtaining QOL information from persons who have impaired language, yet most existing questionnaires focusing on QOL or HRQOL were not developed with PWAs. This lack of inclusion may ultimately introduce biased estimates for PWAs [32]. As an alternative, assessments such as the Stroke and Aphasia Quality of Life Scale-39 (SAQOL39) have been developed for PWAs and been proved valid and reliable [33].

Additionally, the Assessment for Living with Aphasia (ALA) assesses aspects that affect QOL and is suitable for PWAs [34]. The ALA has been found to be a valid and reliable measurement tool; it that was specially developed to determine the impact of aphasia on a person’s overall QOL, including the ability of a PWA to participate in life [34]. The ALA utilizes a pictographic approach, which is appropriate for PWAs across a wide range of severity levels and was designed to include real-life issues such as the impact of aphasia on relationships with spouses, sex, finances, and physical problems, among others [34]. Such issues germane to QOL and life participation for PWAs were procured from the Living with Aphasia: Framework for Outcome Measurement (A-FROM) and the International Classification of Functioning (ICF) from the World Health Organization [10,35].

1.4. Proxy Responses

Overall, research suggests that proxies tend to overestimate the impairments experienced by a partner or spouse following a stroke, and thus, proxy responses may not be reliable [36]. However, a study conducted by Carod-Artal, Ferreira, Trizotto, and Moreira [37], which estimated the agreement between self-report and proxy report using the Stroke Impact Scale (SIS), showed no significant difference between proxy and patient scores in six of eight domains. The proxy scored significantly lower (worse) strength in ADL domains and composite physical domain, but the estimated effect size was small [37]. In addition, a strong non-significant trend was observed in the memory and thinking domains [37]. It seems as if less observable (subjective) domains of the SIS (memory, communication, emotion, and social participation) had lower values. The level of agreement may differ depending on the type of construct measured, since there is adequate agreement reported for the assessment of physical abilities and ADL but poor agreement for the psycho-social domains of the HRQOL measures [37]. In another study by Hilari, Owen, and Farrelly [33], people with chronic aphasia and their proxies were interviewed using the SAQOL-39. On the whole, the data for 50 pairs were included.

The results showed that PWAs see themselves as less affected than their proxies on three of the four SAQOL-39 domains (communication, psychosocial, and physical domains) as well as the overall mean [33]. Cranfill and Wright investigated the importance of HRQOL for persons with aphasia and their significant others, including SLPs in the study [38]. The Stroke and Aphasia Quality of Life Scale-39 (SAQOL-39) and the Quality of Communication Life scale were given to 24 PWAs, seven SLPs, and 24 significant others (SO). Overall results provide further evidence of an incongruity between how SLPs and SOs view the impact of a stroke and its resultant aphasia for PWAs [38].

1.5. Problems in Using Proxy Response

Clearly, the use of proxy responses when attempting to determine the impact of both a stroke and the presence of aphasia for persons with severe aphasia is problematic. To help provide insight, Oczkowski and O’Donnell [27] completed a systematic review on the reliability of proxy respondents for people who have experienced a severe stroke with aphasia. Results showed that ADL, sphincter control, self-care, and defined activities (e.g., preparing meals or shopping) were found to have the highest agreement levels. Conversely, heavy housework, social occasions, and hobbies had to the lowest levels of agreement [27]. With regard to QOL, physical functioning and self-care had the greatest agreement, whereas pain, emotion, and social functioning had the lowest agreement [27]. Finally, increased stroke severity was shown to be the most accurate and consistent predictor between the proxy and proxy respondent [27]. Often the proxy respondents overestimated the ADL and QOL impairments of patients with stroke. The systematic review demonstrates that further research is needed to determine the reliability of proxy respondents for measuring the QOL of PWAs [27]. Carod et al. [37] also reported poorer results when the domains measured were more subjective and the potential for differing results depending on the type of construct measured. Specifically, there was adequate agreement for the assessment of physical abilities and ADL but poor agreement for the psycho-social domains of the HRQOL measures [37]. Finally, Hilari and colleagues investigated proxy and self-report agreements and found that proxies rated PWAs as more severely affected than PWAs rated themselves [33]. Proxy scores were significantly different from the PWA scores on three of the four SAQOL-39 domains (communication, physical, and psychosocial) and the overall mean [33].

1.6. Rational for Current Study

Clearly, a person’s QOL is an important domain for healthcare professionals to monitor following a stroke, and difficulties arise when attempting to obtain those measures from a PWA or their proxy. Even with the inherent problems associated with proxy responses, the use of those responses is still superior to excluding PWAs from QOL studies. Thus, the overall purpose of the current study was to investigate the reliability of proxy responses to the responses of PWAs, including those persons with severe aphasia. Specific research questions include the following:

• Which factors related to QOL are most important for PWAs?
• How reliable are proxy respondents (i.e., spouse/caregivers) in identifying those factors that have a significant impact on the QOL of PWAs?
• How reliable are speech-language pathologists (SLPs) in identifying those factors that have a significant impact on the QOL of PWAs?

Speech-language pathologists were included since SLPs frequently work with PWAs during the rehabilitative process and are thus heavily invested in in their patients’ overall QOL [38]. Furthermore, SLPs are instrumental in developing therapy goals following stroke, and for those PWAs who may have limited ability to communicate, it is essential that SLPs have an accurate picture of the areas of QOL that are important [38]. Limited evidence suggests agreement between the factors that SLPs and significant others deem important but reduced agreement between SLP perception and those of the PWAs [39]. Given the reduced language abilities of a PWA, clearly, more research is needed to help educate SLPs and other healthcare professionals on which QOL domains are most important.

2. Materials and Methods

This study was reviewed and approved by the Institutional Review Board at Towson University. Consent for PWAs included pictographic icons in order to increase understanding. Signed consents were obtained from each proxy, and implied consent was obtained from SLPs.

2.1. Participants

Three groups of participants were involved: PWAs, spouse/caregivers, and SLPs. Each group was asked to rank eleven different aspects commonly found in various QOL questionnaires. PWAs were asked to answer the questions related to their individual situation during an interview session, while the spouse/caregivers were directed to respond as proxies and answer the questions as if they were answering for their partner. SLPs were asked to answer the questions based on how much of an impact they feel each aspect has on a PWA. It should be noted that while the spouse/caregivers were ranking the factors based on how they felt their partner would respond, the SLPs were ranking factors based on PWAs in general. As such, the SLPs were not necessarily responding as a proxy.

PWAs and their spouse/caregivers were recruited from the Snyder Center for Aphasia Life Enhancement (SCALE) in Baltimore, MD, USA, a local aphasia therapy center. All PWAs at SCALE who were diagnosed with aphasia following a stroke and who were 18 years or older were asked to take part in the study. All PWAs were initially assessed by an SLP upon admission to SCALE. Proxies were also required to be 18 years or older, and the identification of a proxy was the responsibility of the PWA. However, guidance was provided to PWAs in that a proxy may be a spouse, caregiver, family member, or a close friend.

2.2. Questionnaire

Data collection was completed by four examiners, all of whom received training in Supported Conversation for Adults with Aphasia [40] from a certified SLP with over 20 years of experience in aphasia. The questionnaire included eleven factors of QOL derived from other QOL questionnaires and surveys [14,27,41]. To increase ease of administration and understanding for the PWA, the questionnaire was modelled after The Assessment for Living with Aphasia (ALA) [42]. The ALA is a psychometrically sound self-report measure of QOL for PWAs with moderate to high test–test reliability with high internal consistency [42]. The ALA uses pictographs and icons to represent different aspects believed to have an impact on the QOL for a PWA [42]. For instance, rather than a Likert-type response scale, the ALA utilizes a rating scale card with a drawing of a person with a frown and with their thumbs down, and a person with a smile and thumbs up on the other end, following eight equal increments.

Ultimately, the questionnaire consisted of eleven aspects: (1) Leisure time; (2) Sex life; (3) Vocation/Occupation; (4) Social contacts; (5) Relationship with partner; (6) Family life; (7) Ability to self-care; (8) Finances; (9) Overall ability to communicate; (10) Physical problems associated with stroke; and (11) Life as a whole. Due to the limited language skills of the participants, each aspect was accompanied by a picture/icon or series of pictures/icons that provided a visual image or representation of that aspect/concept. Visual icons were adapted from either the ALA or PECS (Picture Exchange Communication System; Pyramid Educational Consultants, New Castle, DE).

Prior to administration of the 11-item questionnaire, each PWA was administered the ALA. This was completed for two reasons. First, the ALA is a relatively new QOL assessment tool that was specifically designed for PWAs and can be administered to all severity levels of aphasia, including those with severe aphasia. Given the paucity of QOL information for people with severe aphasia, it was determined that such data may make an important contribution to the literature. Second, by completing the ALA prior to the 11-item questionnaire, administrators were more confident in obtaining accurate, reliable data, as the PWA would have demonstrated understanding of the rating scale and use of pictographic icons.

2.3. Administering the Questionnaire

The testing began by showing the 11 aspects of QOL and placing four cards with pictographic icons on the table. These cards showed a range that included “severely impacted”, “somewhat impacted”, “mildly impacted”, and “not impacted at all”. The participants were asked to use the four cards to show the impact of each item. The examiners randomized the presentation order of the 11 factors. As a result of this initial categorization, the 11 aspects (represented on individual cards) were now grouped in four (i.e., “severely impacted”, “somewhat impacted”, “mildly impacted”, and “not impacted at all”) categories or piles. The participants were then asked to rank each aspect in each pile from “most impacted” (number 1) to “least impacted” (number 11) by starting with the group “severely impacted” and ending with the group “not impacted at all”. Based on these rankings from the PWA, results were entered onto the score sheet. All efforts were made (e.g., use of gestures, additional speech cues, additional instructions) to ensure the PWA understood the task.

The spouse/caregiver proxy questionnaire included the same eleven aspects of QOL as well as the same pictures/icons in order to make both testing situations as similar as possible. Proxy respondents were provided both verbal and written instructions and asked to complete the ranking based on how they thought the PWA would complete the scale.

Finally, a web-based survey was created and distributed on the listservs of two different special interest divisions (swallowing [SIG 13] and gerontology [SIG 15]) of the American Speech-Language and Hearing Association (ASHA). On each listserv, a notice was submitted with a request to participate in the research study. In addition to an implied consent and basic demographic data, the online questionnaire presented each of the 11 aspects and asked respondents to rate each item on a scale of 1–11, with 1 having the most impact on the person and 11 as having the least impact (“Please rank the following categories according to how much of an impact you feel that it has on the patient following a stroke”).

2.4. Data Analysis

All data were recorded in an Excel document and transferred to a statistical analysis software program (IBM SPSS Statistics Version 21) for further analysis. Since the first research question focused on PWAs ranking the 11 factors, descriptive analyses were used. The second and third research questions were addressed by calculating the relationship among the groups of participants using Intra-class Correlation Coefficients (ICCs). ICCs less than 0.40 indicate poor reliability, 0.40 to 0.59 indicates fair reliability, 0.6 to 0.74 indicates good reliability, and excellent reliability is considered as those ICC values greater than 0.75 [43]. To investigate the reliability and consistency across all three participant groups, Cronbach’s alpha was calculated. Alpha values range from 0 to 1, with values >0.7 considered to have acceptable reliability across rating groups and values <0.6 to have poor reliability [44].

3. Results

Participants

The final data set consisted of 31 PWAs (mean age 63.3 years, SD = 9.7), including 24 men and 7 women. The average time since stroke-onset was 7.5 years (SD = 6.5), with a range from 0.8–19.3 years. Twenty-six participants served as proxies for the PWAs, with five PWAs either not indicating a spouse/caregiver or the caregiver declining to participate. The group of caregivers included spouses as well as brothers/sisters, fiancés/fiancées, and sons/daughters to the PWA. Finally, 36 speech-language pathologists responded to the online survey. Demographics include 97% female, 89% working full time, 94% working in a hospital/medical setting, 42% working primarily with adults, and 53% primarily working with the geriatric population. All SLPs responded that they had their certificates of clinical competence in speech-language pathology (CCC-SLP) and were in the United States.

As indicated earlier, all PWAs were originally administered the ALA to ensure understanding of the task. Results are provided in

Table 1. Individual results of the Assessment of Living with Aphasia (ALA).

Time Post-Onset (Years/Months)	Aphasia Domain	Participation Domain	Environment Domain	Personal Domain	Wall Question	Total
6.6	1.2	1.9	2.1	2.0	2.5	1.9
3.5	2.9	3.7	4.0	3.3	2.0	3.5
6	3.3	3.0	4.0	2.5	2.0	2.7
6	3.0	2.3	0.3	3.3	4.0	2.5
12	1.4	2.9	2.4	2.8	2.0	2.6
4.5	2.6	2.9	3.0	3.4	3.0	3.0
8	2.2	3.5	4.0	3.9	3.5	3.5
1.8	2.3	3.3	2.6	3.2	4.0	3.1
3.3	1.8	2.8	3.6	2.0	1.0	2.5
14.3	2.2	3.8	3.0	3.6	4.0	3.5
3.4	1.5	3.0	3.1	3.4	3.0	2.9
4	1.8	2.8	2.4	3.2	3.5	2.7
3.4	0.0	0.5	1.0	2.5	1.5	1.1
13.2	3.3	3.1	3.8	3.6	4.0	3.4
8	2.1	2.3	3.0	2.6	2.0	2.4
19	3.0	3.7	4.0	3.9	3.0	3.7
19.3	2.0	2.6	1.6	2.2	4.0	2.3
7.4	2.0	2.2	2.5	2.4	2.0	2.3
10	1.7	2.0	1.6	2.1	2.0	2.0
4.9	2.2	2.6	3.0	3.1	4.0	2.8
2.6	2.4	3.5	3.4	4.0	3.5	3.5
27	1.9	2.2	3.3	2.6	3.0	2.4
0.8	3.2	3.8	3.3	3.2	3.5	3.5
4.6	0.7	1.6	1.1	2.8	2.0	1.8
1.1	1.5	2.5	2.0	2.2	2.5	2.3
1	2.2	2.3	3.4	2.2	3.0	2.4
2	2.0	2.6	2.0	1.8	2.0	2.2
8.2	3.7	3.5	3.9	3.8	3.5	3.7
18.9	3.1	3.4	3.6	3.7	3.5	3.5
5.6	2.2	2.9	3.1	3.4	3.0	3.0
Average	2.18	2.77	2.80	2.95	2.88	2.75

Note: Average domain values represent the total sum of ratings divided by the number of questions, with a possible range of 0–4 (0 representing the most negative (high impact on QOL) and 4 representing the most positive (low impact on QOL)).

. Scoring of the ALA was purposely designed to be similar to the World Health Organization (WHO) International Classification of Functioning (ICF) [10], thus allowing for conversion of the ALA scoring to the ICF. The “wall question” was included on the ALA as a means to gain an overall impression of “how much aphasia gets in the way of life” for the PWA. While the results of the ALA do not specifically address the research questions, the results are seen as a valuable addition to the growing body of QOL research that has used the ALA and is based on PWAs who attend an aphasia therapy center. Overall, average ALA scores across domains suggest that aphasia has had a neutral-minimal impact on the QOL of the PWAs (average scores = 2.18–2.95).

The first research question addressed which areas had the most significant impact on PWAs. Mean rankings for PWAs indicate that vocation had the most impact on the PWAs, followed by overall impact on communication and life as a whole (see

Table 2. Reliability statistics.

				ICC
Factor	PWA	Proxy	SLP	PWA and Proxy	PWA and SLP	Cronbach’s Alpha
Leisure	5.3 (2.6)	6.7 (2.6)	8.8 (1.8)	0.361	−0.152	0.099
Sex	5.5 (3.5)	6.0 (2.9)	9.5 (1.6)	0.525	0.112	0.496
Vocation	3.5 (2.6)	3.7 (2.5)	6.5 (2.7)	0.304	−0.23	0.16
Life as whole	4.9 (2.4)	3.8 (2.5)	2.9 (2.3)	0.347	0.137	0.299
Social	6.4 (3.2)	6.9 (2.1)	5.4 (2.1)	0.481	−0.102	0.087
Relationship with spouse	7.7 (2.8)	7.4 (2.9)	4.1 (2.0)	0.45	−0.254	0.119
Family	8.1 (2.7)	7.6 (2.5)	5.0 (1.7)	0.523	−0.058	0.354
Ability to care for self	7.3 (3.0)	8.1 (2.7)	4.9 (2.1)	0.601	0.088	0.463
Finances	6.7 (3.1)	5.3 (2.9)	6.7 (2.3)	0.548	0.129	0.404
Overall impact on communication	4.6 (2.8)	3.4 (2.9)	2.1 (1.1)	0.276	0.007	0.147
Physical problems	5.3 (2.6)	5.3 (3.3)	3.6 (2.3)	0.333	−0.376	0.027

Note: PWA = person with aphasia; Proxy = spouse/caregiver; SLP = speech-language pathologist; ICC = intra-class correlation coefficient; Cronbach’s alpha calculated across three participant groups.

). Relationship with spouse, family, and ability to care for self were ranked as having the least amount of impact. Variability in the rankings ranged from multiple factors having a standard deviation of 2.6 to sex, which had a standard deviation of 3.5.

The second research question focused on the reliability of proxy responses with regard to the responses of PWAs. Recall that proxies were instructed to rank the factors as they thought their partner would rank them. Table 2 shows that the spouse/caregivers of the PWAs ranked overall impact on communication as having the most impact (i.e., lower ranking) on QOL, followed by vocation and life as a whole. Relationship with spouse, family, and ability to care for self were felt to have the least impact (i.e., higher ranking). Variability in the rankings ranged from 2.1 (social) to 3.3 (physical problems). Factors ranked as having the most impact on the QOL of a PWA by SLPs were overall impact on communication and life as a whole. Factors ranked as having the least impact by SLPs were leisure and sex. Variability ranged from a standard deviation of 1.1 (overall impact on communication) to 2.7 (vocation).

Table 2 also shows the results of the ICCs, which represent the reliability or internal consistency of proxy respondents (spouse/caregiver) to PWAs. Five of the eleven factors (45%) had ICC values indicating poor reliability (leisure, vocation, life as a whole, overall impact of communication, and physical problems); five of the eleven factors (45%) had ICC values indicating fair reliability (sex, social, relationship with spouse, family, and finances); and one item showed good reliability (ability to care for self).

The third research question addressed the relationship between those factors considered to have the most impact on QOL by SLPs as compared to PWAs (see Table 2). Results of the reliability analysis showed that five of the eleven factors (45%) had ICC values below 0.2, indicating poor reliability, and six of the eleven (55%) had negative ICC values, indicating systematic disagreement [45]. Thus, none of the SLP ratings indicated acceptable reliability with regard to the PWA ratings.

Finally, it was also of interest to determine the overall consistency and reliability across all three participant groups (PWA, spouse/caregiver, and SLP). None of the 11 Cronbach’s alpha values met the criteria for acceptable reliability (>0.7). Alpha values ranged from 0.027 (physical problems) to 0.496 (sex), with three of the values being greater than 0.4 (finances [0.404], ability to care for self [0.463], and sex).

4. Discussion

The overall purpose of this study was to investigate the reliability of proxy responses in QOL studies using PWAs. The study is one of only a few that not only compare the responses of the spouse or caregiver but also seek to provide insight on which QOL factors SLPs believe to be relevant. Overall, the results show that PWAs ranked vocation, overall impact on communication, and life as a whole as having the most impact on their QOL and that aphasia had a neutral-minimal impact on their overall QOL in general. In addition, the results also showed that proxy responses on QOL-related questionnaires may not be entirely reliable or consistent with those of the PWA. Finally, the results suggest that those factors that SLPs believe to have a significant impact on the QOL of PWAs may not be consistent or reliable with those identified by their patients. Specific information by research question is provided below.

4.1. QOL and PWA

In the current study, PWAs were asked to rank 11 pre-selected and different factors on how much each impacted their QOL. Vocation was identified as having the most impact, and family was identified as having the least impact. Overall, aphasia had a minimal impact on the QOL of the PWAs involved in the study, as indicated by a standardized aphasia QOL questionnaire (ALA). Factors with the greatest variability were finances (SD = 3.1), social (SD = 3.2), and sex (SD = 3.5), potentially indicating a lack of consensus among the PWAs, at least in these three domains.

While there are no existing studies that have used the same 11 factors to investigate QOL for PWAs, the results are in concert with other studies in the literature. For example, Hilari and Byng [26] reported the PWAs in their study had a severe communication disability, which affected their QOL. In the current study, PWAs ranked communication as having the second greatest impact on their QOL. Other studies also support the finding that PWAs rank communication as having a major impact on their QOL [33,46,47]. Given the fact that stroke is often accompanied by a language deficit, the negative impact on communication is understandable. Additionally, the current results indicate that vocation had the most impact on the QOL of PWAs included in this study. This is in contrast to Hilari and colleagues, who reported that employed status was not significantly associated with overall HRQOL [46].

It may be possible that the discrepant results are attributed to the fact that the participants in the current study were recruited from and members of an aphasia therapy center. While impaired communication is routinely cited as having the most impact on a PWA following stroke, the fact that vocation was cited as having the most impact might suggest that the current participants intend to or hope to return to work in some form. In fact, it has been suggested that healthcare professionals recommend PWAs to find others with aphasia in order to develop a “collective identity” [48]. The development of such relationships would counter other published results indicating that PWAs communicate with fewer friends and have smaller social networks [49].

The current results using participants from an aphasia therapy center suggest otherwise: social was ranked toward the bottom of the 11 factors, suggesting that participation in the aphasia therapy center may coincide with increased social networking. Other studies have also shown the beneficial effects of participation in an aphasia therapy-type center [50,51,52], with the rationale being that such targeted and focused intervention may act as a catalyst and spread to other areas that are affected post-stroke [50]. Clearly, more work and research needs to be completed before making generalizations, but intuitively, increased socialization should be a goal for PWAs.

4.2. Reliability of Proxy Responses

The current results suggest that when a proxy is allowed to respond to a QOL questionnaire or survey for a PWA, their responses may not be entirely consistent or reliable. While this is not the first research study to report such results, the finding does add an additional note of caution when obtaining QOL information from PWAs. Other studies with similar results include Oczkowski et al. [27], Carod-Artal et al. [37], and Hilari et al. [33]. Each of those studies also demonstrated the difficulty of comparison between self and proxy responses, with each reporting overall poor agreement, particularly for the more subjective domains such as pain, emotion, and social functioning [27]. Alternatively, more objective domains (e.g., physical functioning, self-care) showed better agreement. With regard to the 11 aspects included in the current study, the physical problems associated with stroke might be the most objective domain of the eleven aspects. Nevertheless, this domain shows the fourth worst agreement between self and proxy, whereas ability to self-care shows the best agreement. This may be due to the overlap of this domain with regard to the PWA’s life and the caregiver’s life. That is, the better the self-care of the PWA, the less support the caregivers are required to provide. This is why the caregivers may have the best overview about the actual needs of the PWAs in this particular domain.

Also, the ICC was used to investigate the agreement between the proxy report and the self-report and shows values of <0.4, and therefore a poor agreement, for ten of eleven factors. Again, only the factor ability to self-care shows a single ICC value of >0.4, which indicates a fair to good agreement. ICC values show that only two aspects, family life and, again, ability to self-care, had acceptable values. Results from the current study are similar to others in the literature, specifically Oczkowski et al. [27], Carod-Artal et al. [37], and Hilari et al. [33]. However, different from these results, the current results do not show the best agreement in the most objective domain, which should be the physical problems associated with stroke. The current study shows the best results in domains such as family life and ability of self-care, which may have the biggest overlap in everyday life between the two groups, PWAs and proxies.

Finally, the results of the study showed that none of the SLP responses were in agreement with the PWAs’ responses. This is particularly concerning, as SLPs are the primary healthcare providers treating communication for PWAs and should be theoretically more in tune with the needs and concerns of PWAs. Although limited, there is similar research showing that healthcare professionals as a whole may not be reliable to serve as proxies and may overestimate impairments and underestimate areas concerning health-related quality of life [53,54,55].

4.3. Limitations

The first limitation that should be noted and that prevents generalization to all PWAs is the fact that all of the participants were recruited from a dedicated aphasia therapy center. Therefore, the PWAs as well as their caregivers already had a modicum of support. This may have effects on the value of agreement, which may positive or negative. On the one hand, support by a professional institution may enhance mutual understanding and therefore the value of agreement. On the other hand, the support by others (such as a professional institution) may bolster the independence of the PWAs in relation to the caregivers and may lead to a decreased overlap in everyday life, with effects on agreement. Regardless, it would be prudent for future studies to account for the degree of social support during recruitment or provide data analysis to elucidate the role of this important factor when discussing QOL in stroke patients.

Furthermore, the group of caregivers includes several types of caregivers, such as brother/sister, fiancé/fiancée, and son/daughter. The degree of connection as well as the average time that the PWA and caregiver spend together may also have effects on the proxy answers and the agreement between the two groups.

In addition to degree of social support, the total sample size should also be noted. Specifically, the current study included only 31 PWAs, which greatly decreases the generalizability of the results.

Finally, results of the Cronbach’s Alpha showed that none of the parameters included in the QOL scale were found to be internally reliable. While consistent with the overall results of the study suggesting the poor reliability of proxy responses in general, the results should yet again be interpreted with caution, as there may not be complete agreement across participant groups as to the true definition of each parameter. Factors such as current emotional state, culture, employment, and sociodemographic factors, among others, may play a role and should be taken into consideration in future studies [56].

4.4. Suggestions for Further Research

Further research about the quality of life of PWAs should also involve people who do not belong to any kind of professional institution. As already noticed, recruitment might be more difficult, but the collected data might highlight differences between the groups (belonging or not belonging to a professional institution) and might enlarge the possibilities in interpreting the collected data. Such a project could not only provide references as to the quality of proxy reports but could also enable a comparison of the ranking of most and least impacted quality of life domains regarding aphasia. This would decrease the knowledge about PWAs who are not part of professional support systems and about the positive or even negative influence of support systems.

Additionally, future research should consider the total number of factors included in a QOL scale targeted at PWAs. The scale utilized in the current study included 11 factors, which were derived from existing QOL scales. Considering that a high percentage of people who have had a stroke also have attention or other cognitive deficits [57], the increased number of factors may have had unintended consequences on the outcomes. That is, any cognitive/attention deficits present in the participants may have reduced their ability to complete the task. While the examiners did not report any such instances per se, participant responses may have been unintentionally impacted. Future research should attempt to pare down the total number of factors to expedite the process and reduce the potential of decreased attention. Finally, a survey about the kind of caregiver would be interesting. This would enable information on which kind of caregiver (brother/sister, fiancé/fiancée, or son/daughter) might provide the best results in terms of agreement. This could help to improve the research field comprising QOL studies of PWAs on the whole.

5. Conclusions

The current study provided valuable insight on the QOL of PWAs and their caregivers. The results show low values of agreement of the two included groups (PWAs and caregivers), which was also the main result of other similar studies in this research field. The identified limitations of the current study give helpful advice and information for further research, which can lead to an overall improved quality of QOL studies of PWAs.