This qualitative study employed a descriptive phenomenological approach, within a constructivist paradigm. Phenomenology is concerned with obtaining the lived-experience—the essence, or true meaning of a phenomenon (16). Descriptive phenomenology supported the physiotherapist researcher to describe the experience in an unbiased way (17). Bracketing and separating preconceived beliefs out of consciousness were used to capture the essence of the experience in its purest form (18). In the constructivist paradigm, description of the experience is accessed through the construction of knowledge obtained from the view of the person perceiving the experience (19). Ethics approval was obtained from the participating facility (HREC/18/WH/47137) and university (2019-17R) ethics committees. This study is reported using the Standards for Reporting Qualitative Research (20). Throughout, healthcare professional refers to professionals involved in stroke survivor health care, and rehabilitation service refers to facilities providing rehabilitative care.

Stroke survivors who experienced fatigue were purposively sampled inclusive of a range of ages, stroke severities and cultural backgrounds from a cohort attending physiotherapy at an outpatient rehabilitation facility that has two sites in Melbourne, Australia. Residents in the geographical area are from culturally and linguistically diverse backgrounds—~25% of stroke survivors admitted do not speak English as their primary language (21).

Stroke survivors were screened and invited to participate by their treating physiotherapist if they met the following inclusion criteria:

Stroke survivors with other health conditions causing fatigue, such as cancer or chronic respiratory conditions were excluded. The Fatigue Assessment Scale was completed by treating physiotherapists when screening their existing caseload and new admissions for the study. Upon meeting inclusion criteria, the stroke survivor was immediately referred to the study. Stroke survivors were asked if they identified as having a carer and were provided the option of their carer attending the interviews. There were no other carer inclusion or exclusion criteria.

Face-to-face semi-structured interviews were conducted by the primary researcher (EB) and audio recorded. An open-ended, pre-piloted interview schedule was used to elicit perceptions of stroke survivors, together with their carers where present (Figure 1). Pilot testing of the interview schedule, conducted with a stroke survivor and carer not included in the study, resulted in minor changes. No changes were made to the interview schedule between participants. Participants could share what they wanted in the interviews, with probes seeking further depth and clarification (23) and minimizing the researcher imparting assumptions (24). Carers provided further insight into stroke survivors' experiences as well as their own.

Researcher thoughts and participants' non-verbal communication were documented. Interviews of 30–60 min were conducted in a private room at the facility or participant's homes, between March-December 2019. Professional interpreters were utilized at the preference of the participant. Potential participants with stroke-related communication impairment were seen by a Speech Pathologist as part of routine care, to determine suitability for interview. Interviews were conducted by a clinician with substantial experience communicating with people with aphasia. Verbal communication in the interview was adapted to be aphasia friendly (25). This included modifying questions as required, being patient and encouraging, noting non-verbal communication, allowing participants to access the questions and write responses before the interview, and having carers present to support communication. Interview questions were provided to participants one week prior to their interview (26).

Audio recordings were transcribed verbatim by the primary researcher (EB) to increase data familiarity. Anonymised transcripts were exported to the encrypted web-based qualitative data-analysis application Dedoose (27). Comments recorded in the interview schedule were added to the transcripts. Data were analyzed after blocks of four to six interviews according to Colaizzi's seven-step, non-linear analytic method (28) (Figure 2), aligned with descriptive phenomenology (29). To enhance credibility of results (30), the essential structure was presented to participants and carers at a second interview either face-to-face or over the phone. Participants confirmed if their experience was accurately depicted and provided elaboration of data, but no new themes emerged. Recruitment continued until study size was sufficient to elucidate the study's aims (31), applying the concept of information power whereby study aim, specificity of information participants hold, theoretical background, quality of dialogue, and data analysis strategy are continually considered in guiding sample size (32). Prior research about the topic, including a theoretical model for post-stroke fatigue (33) informed development of the study aim and research questions. Quality of dialogue was strengthened due to the primary researcher's extensive experience communicating with stroke survivors, including those with aphasia (32). Sample size was expected between six and twenty, as is commonly necessary for adequate information power in phenomenological studies (17). No new knowledge emerged relevant to the study's aims in the final three interviews.

All researchers contributed to study design. Colaizzi's method was executed by a novice in qualitative research (EB) who underwent tertiary provided training, interview observation in another study, and received practical mentorship from the research team. Data analysis was independently confirmed by another researcher with extensive experience in qualitative research (SK). The decision trail at each step of the research process was audited and emerging results and information power challenged by two other members of the research team (CS, KH) with experience in rehabilitation (CS in stroke specifically) and in qualitative research (29). Final themes and illustrative quotations were critically analyzed for verification (CS, KH). The researcher conducting data collection (EB) had 10 years' experience in stroke rehabilitation. Participants were not patients of the researcher but may have had indirect contact during their rehabilitation. Continual reflexive analysis of the researcher's impact on the research process was addressed through journaling and discussions with the research team (SK, CS), to assist with debriefing and bracketing.

For some, fatigue was noticed when the stroke occurred. However, fatigue was often uncovered once home from hospital when returning to usual daily activities, particularly for younger stroke survivors. Fatigue could be physical or mental, or a combination of both. Fatigue could be unpredictable leaving stroke survivors unsure of how much they could achieve a given task, “You don't know when it's going to hit” (John^*). Carers and stroke survivors reported lacking insight into their experience initially, or confused fatigue with other stroke impairments like low mood or emotional lability. This links strongly with the universal sense of feeling unprepared for fatigue.

Post-stroke fatigue was different to pre-stroke fatigue, but was influenced strongly by pre-morbid personal factors such as self-perceived work ethic and level of activity pre-stroke. Participants acknowledged trying to ignore fatigue, thinking it would ease as recovery proceeded. When it persisted, admitting fatigue was seen as admitting defeat, connected to pre-morbid personal traits such as identifying as a strong and independent person who should not be troubled by a circumstance like fatigue. Stroke survivors reported experiencing feelings of anxiety and fearing the unknown of fatigue, worrying about anticipated events. This manifested in concern about injury if they pushed too hard and worry about exacerbating fatigue in the short and long-term. If improvement was noted, fear of fatigue setback was described.

There was angst about where one could rest when participating outside the home, contributing to restriction in meaningful activities and life roles. Fatigue limited social participation with friends and family, leisure, holidays, accessing the community, vocational work, and familial roles. Frustration was common, heavily influenced by a sense of not meeting the expectations one has of oneself.

Lowered motivation, whilst acknowledged as present in the experience, did not limit initiation of meaningful activities. Stroke survivors overwhelmingly spoke of always trying hard, determined to participate even if that meant spending time resting in days proceeding. As such, there was a notion of fighting with fatigue; participants used language such as “battle”, “trauma” and fatigue having “taken its toll”.

Participants described attempting to accommodate and reduce fatigue through self-management. This commenced with a level of acceptance as a starting point for coping and finding some normality and enjoyment in life. Faith and spiritual beliefs helped some cope, when other strategies hadn't worked, more commonly described by stroke survivors born overseas.

Self-management strategies included adapting activities and plans to suit fatigue levels each day, pacing and grading activities. Exercise, reducing simultaneous mental stimulants, reducing cognitive and emotional stressors, and minimizing alcohol intake were also helpful. Sleeping helped for some, often utilized mid-morning after usual morning activities. For others sleep made no difference and they still awoke tired.

Universally there was benefit of having predictable, daily routines with rests, such as finding a quiet area to sit or lie down. This involved gathering thoughts, sleeping or just quietening the mind whether physically or mentally fatigued.

With successful self-management, participants measured progress by noting a reduction in fatigue severity, or requiring shorter and less frequent rest. Stroke survivors recounted having no choice but to manage their fatigue because of the negative manifestations. They had to proactively pay attention, be aware of limitations, listen to their body, look after themselves, be flexible and continuously adjust. Carers also highlighted that they felt responsible for supporting self-management of post-stroke fatigue.

Participants noted rehabilitation was challenging when experiencing fatigue. Rehabilitation was described as exhausting at times, resulting in the remainder of the day spent sleeping or resting.

Stroke survivors and carers developed strategies to manage this and reported benefits of rehabilitation on fatigue over time, with carers quietly noting the improvement. They tried to plan therapy schedules to optimize fatigue management, requiring rehabilitation services to be flexible to their varied needs. Carers often viewed schedule planning as their duty.

Stroke survivors recognized benefits of home sessions and independent exercises for fatigue, but also valued coming to the center for building motivation through personal attention of the therapist and the equipment available. Contrary to all-day exhaustion, many stroke survivors described a second wind of energy after rehabilitation or exercise sessions. This might relate to the physiological effects of exercise, or to a sense of achievement, even if that meant pushing beyond their known limits.

Fatigue was also felt to negatively impact rehabilitation, resulting in missing part or whole sessions and home practice. Whilst this concerned participants, it was not harbored on, and stroke survivors focused on moving forward through their rehabilitation toward set goals. Goals for many stroke survivors were often related to participation in family-focused activities, particularly for participants from culturally diverse backgrounds, and family members were an additional source of motivation in these experiences.

Stroke survivors reported largely learning to deal with fatigue independent of healthcare professionals. They accessed previous knowledge about their bodies and experiences, and used trial and error to find something that helped, recognizing mental and physical triggers over repeated exposure. Stroke survivors learnt the benefit of preventing exacerbation of fatigue from negative experiences of “overdoing it”. They sometimes learnt from family, friends, other stroke survivors, and internet sources. Some expressed caution when accessing information from the internet, however participants from culturally diverse backgrounds identified the importance of resources in appropriate languages which often came from the internet.

Learning from health professionals was dependent on circumstance; receiving education about fatigue was an exception rather than a rule. If the healthcare professional was more experienced, stroke survivors and carers perceived fatigue information was obtained more easily and considered more trustworthy; advice was accepted, and demonstration of how to pace in occupational activities occurred.

Participants felt fatigue was often missed by health professionals. They either “hadn't focused on it”, had not brought it up or put fatigue down to something else such as medication, which the stroke survivor or carer disagreed with. Fatigue was noted by stroke survivors and carers to be missing in stroke education groups. When engaging in outpatient rehabilitation, participants placed less emphasis on learning about fatigue from medical staff they saw only periodically.

Carers appeared to facilitate the uncovering of fatigue and the journey to self-management, reported by both stroke survivors and carers. Carers identified they could enable learning through reinforcing education when received, researching independently, or noticing fatigue triggers. Rather than performing a task for the stroke survivor, carers assisted with task completion or monitored pacing, allowing stroke survivors to gradually increase capacity.

Carers adopted rehabilitation scheduling and routine planning responsibilities, played a protective role, and often took a large physical load. Carers often reported frustration at setting pacing boundaries for fatigue management that stroke survivors did not adhere to. Many stroke survivors recognized the role of carers in fatigue management and worried about those without carers. Several carers expressed anxiety about how they would continue this role once formal rehabilitation finished.

Some stroke survivors reported family members could be dismissive of fatigue and possess unrealistic expectations of them, contributing to further distress. This was predominantly reported by stoke survivors with minimal stroke deficits or by young stroke survivors within a busy family unit. Universally, the invisibility of fatigue was perceived as influencing the expectations of others, “I suppose it's like the silent killer.” (Kapil)

Carers also identified they can alter their existing approach to be more positive in supporting the stroke survivor with managing fatigue.

Results of this study are potentially not representative of stroke survivors who were not receiving intensive outpatient physiotherapy services, and therefore limits understanding of the experience for those without physical deficits. However, other studies with community-dwelling stroke survivors reported similar experiences of searching for answers due to a lack of information about fatigue from health professionals (37, 40). We have extended understanding by describing where and how stroke survivors and carers sought information. Interviewing stroke survivors and carers together may have restricted participant comfort to speak freely. We set out to understand stroke survivor experiences and the compelling insights about caring for a stroke survivor with fatigue highlight carer burden should be explored in future research. The Fatigue Assessment Scale is one of several existing fatigue measures. Limitations in accurate detection and measurement of available fatigue measures have been highlighted in the literature (49) so we may have unintentionally omitted other stroke survivors with fatigue. Sample size was small but was justified by our in-depth analysis of strong, clear interview dialogues, and a sample that had dense specificity of information relevant for the study aims, providing adequate information power (32). Qualitative research results are not designed to be reproducible as eliciting participant lived-experience in constructivist-based descriptive phenomenology is situation-specific (50), reflecting the voices of recruited participants. We also cannot make inferences about the typical experience of specific cultures as our participants were from a broad range of cultural groups.