Exploring Poststroke Mood Changes in Community-Dwelling Stroke Survivors: A Qualitative Study

Recruitment 

We identified participants from a larger, retrospective, cross-sectional study of 91 stroke survivors at 1, 3, and 5 years; the design and methods of this study have been previously reported.29 In brief, participants were identified from hospital records and separations from the JHH, the largest tertiary referral hospital servicing Hunter residents within New South Wales, Australia. Inclusion criteria included: admission to JHH, alive at discharge to the community, and history and examination consistent with stroke as determined by a stroke physician. We aimed to avoid excluding participants with cognitive or communication deficits. Thus, participants were excluded only if they had severe cognitive impairment (MMSE score ≤16)30 or were unable to communicate sufficiently to participate in an interview. Further exclusion criteria included inability to give informed consent and current residence in a nursing home: that is, only community-dwelling stroke survivors (including those in hostels) were contacted. Exclusion was determined at the time of the face-to-face interview.

Sampling 

To obtain more details on the experience of stroke survivors, we invited all participants from this larger cohort to participate in this qualitative study using a mail-out that included some brief questions about demographics and living situations.

In total, 37 participants consented to participate in semi-structured interviews and, based on information gained from the mail-out, 12 stroke survivors were selected, 4 participants from each of the 3 cohorts (1, 3, and 5 years). Purposive, maximum variation sampling was used to identify men and women of different ages and with different living situations so as to reach a sample as heterogeneous as possible.31 This allowed for the selection of participants who could best provide insight into specific and personal experiences regarding the issues being examined, rather than obtaining a representative sample, as would be sought in quantitative research. All participants chose to be interviewed in their own home.

Data Collection Methods 

Data Analysis 

We transcribed all interviews verbatim, with identifying data removed. Data analysis was undertaken using a modified grounded theory. Grounded theory methodology allows for the development of theory from the experiences of the study participants; that is, the theory is grounded in data. A modified grounded theory approach incorporates data collection, coding, and analysis using a process of constant comparison, without the theory development component.36 This entailed an ongoing cycle of data collection and simultaneous analysis, whereby each transcript was read and reread by 2 members of the research team (J.H.W., P.M.) prior to consensus coding.

Each interview was transcribed, conceptualized, and coded. The first step of the coding process was open coding, which involved a literal line-by-line reading and interpretation of the transcripts to identify a wide range of categories from the data. These categories were given a label or code consisting of 4 letters in order to facilitate retrieval between the transcripts (for example, feelings of frustration were labeled as FRUS). Constant comparison involved comparing the emerging codes between each transcript. The next level of analysis was axial coding, which involved identifying relationships between codes and the grouping of codes into coherent categories. For example, all codes that appeared to influence mood were grouped and influences of these relationships were sought. The final level of analysis involved the selection of major categories, which outlined the relationships and interactions between the codes related to the experience of stroke, emotions, and community participation.

To ensure accuracy, 2 researchers undertook the consensus coding (J.H.W., P.M.). After independent coding of the transcripts by each researcher, there was a process of comparative coding including discussion of the codes and how they were developed. Differences in researcher perspective were resolved by negotiation and consensus and fed back into the analysis to cross-check codes and themes and develop an overall interpretation of the data.

Rigor was further enhanced by incorporating a search for negative cases (cases that appear to disconfirm or refute any conclusions arising from data37), use of participants' words, and member checking (testing the data with the members of the group from which they were collected38) whereby transcripts were mailed to participants for them to review and to add or remove any information from their transcript. To assist participants in this process (in case of potential cognitive and communication difficulties), the participants were contacted by phone within a few weeks of receiving the mail-out to clarify their feedback. No changes were made to the transcripts or their interpretation by the participants, however. There was the provision for further interviews to be done with additional participants if required; however, thematic saturation (a situation in which no further themes were emerging) was achieved during the initial 12 interviews and, thus, we scheduled no subsequent interviews.

Ethics 

Approval for this project was obtained from the Hunter New England Human Ethics Research Committee.

Feelings of Frustration 

The experience of frustration and low mood were often concomitant and strongly linked to functional limitations. All participants expressed negative feelings about “not being able to do what I like” (participant 11, female, age 79) and being “fed up” (participant 4, female, age 78).

“Actually it is more frustrating than anything else because you cannot do what you did before. You think you can do it until you try to do it and there is no way in the world you can.” (participant 7, male, age 58)

Experiences of frustration were compounded by the need to take extra time or the inability to complete a task: “That is the frustrating part. You can't make your limbs move the way that you would like to” (participant 6, female, age 81).

Anger was commonly associated with frustration and participants reported feeling “impatient” (participant 4, female, age 78) and “cranky” (participant 11, female, age 79) or had an increased tendency to “flare-up” (participant 7, male, age 58):

“You get frustrated with yourself and angry…I get very angry.” (participant 3, female, age 66)

Feelings of Reduced Self-Efficacy 

The second factor related to mood change, feelings of reduced self-efficacy, was expressed in a variety of ways such as feelings of being “useless” (participant 1, male, age 92) and “unproductive” (participant 3, female, age 66). The lost ability to partake in their prestroke lifestyle continually reinforced for 1 participant her being “different and not the same as everyone” (participant 6, female, age 81). Reduced self-efficacy was further manifested when participants were dissatisfied with performance outcome and the quality or “thoroughness” (participant 6, female, age 81) previously achieved:

“Well in a way it makes you feel like you want to give up because you are not improving as you expected to. You got to accept the fact that you are not going to get better.” (participant 7, male, age 58)

“I am just a hindrance. I accept that.” (participant 9, female, age 74)

Such feelings of lost competency were seen to impact on their self-confidence and led to further withdrawal from participation in activities. Some participants had made a conscious decision to withdraw due to feelings of “embarrassment” (participant 7, male, age 58). This phenomenon was most evident in participants with speech impairments who experienced stigmatization from the broader community:

“I think I am talking funny so I sort of get out of the road.” (participant 10, male, age 70)

Feelings of Dependency 

The third factor related to mood change, feelings of dependency, was identified when participants were required to rely on other people for assistance, which contributed to their experience of low mood. This was exacerbated by the inability to undertake valued activities that were central to a participant's prestroke identity, such as maintaining the home, leisure, and work activities. The extent of dependency varied according to residual impairments and the most common restrictions related to ADLs:

“Have to get help to do up a shirt or pull a jumper down, put on a coat … small things that you are used to doing for yourself.” (participant 10, male, age 70)

Participants found the need to rely on others was “hard to accept” (participant 4, female, age 78) and this was observed as being a major contributor to distress:

“You got to rely on people unfortunately.” (participant 7, male, age 58)

Feelings of Loss 

The final factor related to mood change, feelings of loss, was characterized by the experience of residual physical, cognitive, or communication impairments, which all contributed to role loss:

“I used to go round [hostel] units here and [serve] communion…I was really involved. I just can't any more…it is a bit upsetting.” (participant 4, female, age 78)

There was a sense that life had “changed forever” (participant 10, male, age 70) and that the quality was “nowhere near as good” (participant 9, female, age 74)

Feelings of guilt, loss, and having “paid the price” (participant 7, male, age 58) were common in participants who felt they were to blame for their stroke. Participants identified a range of causative factors such as “drinking” (participant 2, male, age 60) or “stress” (participant 3, female, age 66):

“I was working 16 hour days, plus overtime and it just caught up with me.” (participant 7, male, age 58)

Changes in social patterns were commonly reported, characterized by reduced frequency of contact with friends and family, particularly friends. The experience of social isolation was an additional cause of distress in participants, due to feelings of loss: “I do miss the kids [grandchildren], I miss them very much” (participant 11, female, age 79).

Modulating Factors 

These feelings of frustration, reduced self-efficacy, dependency, and loss were modulated by 4 internal factors. These 4 factors were insight, optimism, faith, and resilience, which existed on a continuum of high to low. When these modulating factors were present, participants experienced more positive mood. When these modulating factors were absent, participants were more likely to experience or report feelings of frustration, reduced self-efficacy, dependency, and loss.

Insight referred to a participant's ability to comprehend their abilities and limitations and to effectively problem-solve barriers in order to accomplish a desired task. This led to enhanced feelings of independence and choice, which assisted in the maintenance of a positive self-efficacy:

“I sit down now and think, ‘Well blow the house work, I will crochet.' I enjoy that more, whereas before I would run around doing everything. I have slowed down a lot.” (participant 5, female, age 43)

Optimism was another key factor that modulated the experience of stroke. For many participants optimism was centered on “living for others” (interview 10) and seeing milestones achieved in the lives of children and grandchildren. These included marriage, the birth of grandchildren, or the graduation of grandchildren from school and university.

The modulating factor of faith was characterized as a significant sense of hope through a relationship with God:

“If it was not for God I would be a lot worse with my stroke I think.” (participant 9, female, age 74)

The modulating factor of resilience (and of having achieved mastery during previous life experiences and challenges) was an interesting finding in this study, facilitating greater adjustment to stroke. Participants felt that overcoming previous adversity in life helped them “not to take things for granted” (participant 6, female, age 81) and to “make today count” (participant 9, female, age 74).

Resilience was strongly linked with experiencing ongoing support of family, friends, and loved ones. This included the provision of practical support or emotional support:

“Having a family that cared, or friends that cared, that's the main thing, getting those things when you need them. I have never been on my own.” (participant 10, male, age 70)

Study Strengths 

This study generates unique data on mood disturbance in Australian stroke survivors and provides baseline data for ongoing measurement of outcomes after stroke in this region. The strength of this study lies in the use of qualitative methodology to obtain a description of mood disturbance after stroke using semi-structured interviews. There was also the opportunity to draw links between this study and quantitative data from a larger study. The use of an objective sampling frame in this study and the consecutive identification of study participants has been previously reported. Such results suggest that the information is likely to be representative of the experience of PSD in community-dwelling stroke survivors.

Study Limitations 

The study used multiple cross-sectional samples at 1, 3, and 5 years poststroke; this limits the ability to make interpretations regarding changes in poststroke functioning in people over time. The next step would be to confirm these findings in a prospective, longitudinal cohort. The accurate identification of mood is also influenced by the use of the emotion domain of the SIS in contrast to a formal, standardized assessment of mood. The study was intentionally aimed at community-dwelling stroke survivors in order to inform community service provision; the study excluded people with more severe strokes, such as people in nursing homes, and is subject to survivor bias, limiting the generalizability of the study. Qualitative interviews were limited to 1 interview per participant and therefore may not capture any changes over time.