“A fog that impacts everything”: a qualitative study of health-related quality of life in people living with HIV who have cognitive impairment

Background Cognitive impairment (CI) in people living with HIV (PLWH) is an important health concern in the context of an ageing HIV population. Impacting 14–28% of PLWH, CI is associated with lower health-related quality of life (HRQoL), however, evaluation of the illness-specific factors comprising HRQoL in PLWH with CI have not been assessed. Objective We sought to contribute evidence toward an understanding of HRQoL and identify domains of HRQoL in PLWH with CI. Methods Qualitative interviews with 25 PLWH with objective CI related to HIV disease were conducted with participants attending HIV clinics in the UK. Clinically significant CI was defined based on The European AIDS Clinical Society guidelines, requiring: (i) subjective reporting of cognitive symptoms; (ii) symptoms to be related to HIV (e.g. potentially confounding non-HIV related conditions have been excluded or are being optimally managed) and; (iii) formal neuropsychological assessment confirming CI. Median age was 56 years (range 35–80); 18 participants were men (72%); 11 (44%) were white British and 8 (32%) were Black African; 14 (56%) were men that have sex with men and 10 (40%) were heterosexual; median number of years living with HIV was 17 (range 1–34); and all participants were on combination antiretroviral therapy. Analyses employed techniques from grounded theory, underpinned by an inductive, collaborative team-based approach. Results Findings revealed seven interrelated domains comprising HRQoL experiences were identified: Physical function, Cognition, Social connectedness, Physical and mental health, Stigma, Self-concept, and Control and acceptance, and each was defined by specific descriptive components. Conclusion This study provides valuable insights on the factors that drive HRQoL in PLWH with CI and contribute to a body of evidence which provides targets for the development of targeted interventions to maintain or improve quality of life.


Introduction
Cognitive impairment (CI) due to HIV disease affects 14-28% of people living with HIV (PLWH) [1], and impacts on medication adherence, employment, and quality of life [2]. Recent cohort studies have estimated 73% of Europe's estimated 2.2 million [3] PLWH will be over 50 years of age by 2030 [4]. With the attenuation of cognitive reserve seen in PLWH, as well as an increased incidence of neurodegenerative disorders with age, there is likely to be an increased burden of cognitive difficulties in PLWH in coming years.
The advent and availability of effective antiretroviral medication in the last 20 years has transformed HIV from a life-limiting illness to a chronic condition which can be successfully managed on a long-term basis. In the UK, men who have sex with men (MSM) and Black African people are disproportionally affected [5] and research generally reports health-related quality of life (HRQoL) in PLWH as being worse than that seen in non-infected populations [6][7][8][9]. PLWH have to cope with a range of HIV-related symptoms and experience more co-morbidities (including cognitive impairment), than age-matched negative populations [10,11]. In addition to the biological consequences, PLWH frequently face wide-ranging psychosocial influences such as stigma, poverty, substance abuse, poor mental wellbeing, and cultural beliefs which can impact HRQoL [12]. HRQoL is a multi-dimensional construct comprised of aspects including physical, psychological, and social functioning. Generic models (i.e. those applying to all illnesses) may also include domains of general health perceptions, along with dimensions of pain, energy, independence, environment and spirituality [13][14][15][16][17][18]. Generic models may miss components of HRQoL that are important in particular disorders, so disease-specific conceptualisations and measures have been developed to identify and measure domains of HRQoL which have particular salience in specific illness experiences. For PLWH, alongside generic aspects, domains of stigma and discrimination, barriers to care, engagement coping, cognition, and sexual functioning have been identified as important [19][20][21][22]; in individuals with mild/moderate dementia domains of self-concept, behaviour and mood, and self-awareness have been identified as central to HRQoL [23][24][25][26]. HRQoL provides a broad and patient-reported appraisal of the impact of a condition on that individual and the overall effects of treatment and care, and so are of increasing importance in the holistic evaluation of interventions and services. Therefore, understanding more about HRQoL in PLWH with CI might assist in the provision of improved person-centred care and inform the development of targeted interventions to maintain or improve quality of life.
The clinical characteristics of CI in the context of HIV (often referred to as HIV-associated Neurocognitive Disorder (HAND), in research contexts [27]) are well documented with biomedical and neuropsychological studies reporting deficits in memory, learning, attentional and executive processes which have impact on health and function [28]. Less is known about the impact of these impairments with a recent scoping review of research finding a paucity of studies examining HRQoL and its associated dimensions in PLWH with CI [29]. Quantitative studies report lower overall HRQoL, and attenuated physical, psychological, and social HRQoL in PLWH with CI, compared to HIV [30][31][32] and non-HIV matched controls without CI [32]. Specifically, these studies found PLWH with CI reported poorer overall QoL across all domains of the MOS-HIV questionnaire [30,31] and the SF-36 [32] which include domains related to physical functioning, physical health, mental health, and social functioning. Three qualitative studies, have sought to understand the lived experiences of PLWH with CI, and found impacts on concentration, multi-tasking, and fulfilling daily activities) [2,33,34], negative emotional reactions (embarrassment, stress, worry, depressive symptoms) [2,33], and dimensions of discrimination and stigma [2] that affect them in work and social settings. These studies, along with the calls to include good quality of life as a fourth '90' in the UNAIDS Testing and Treatment targets, support the value of further work to assess HRQoL directly and the illness-specific domains comprising this important outcome in PLWH with CI. Here we describe the results of a study conducted systematically to understand HRQoL experiences and identify domains of HRQoL in PLWH with CI.

Study design
We conducted an exploratory qualitative study with participants recruited from specialist HIV clinics in the South-East of England from January to July 2020. Analysis of transcripts and decisions regarding the final domains were led by an experienced qualitative researcher (SD) and the team included a doctoral student (KA), an HIV medical consultant/academic (JV), and a Dementia medical consultant/Professor of Dementia (SB).

Participants
Participants were patients attending specialist HIV services in London (King's College Hospital) or Brighton (University Hospitals Sussex). All had a diagnosis of HIV, were on combination antiretroviral therapy, and had clinically significant CI. Clinically significant CI was defined based on the European AIDS Clinical Society (EACS) [35] definition of CI related to HIV. This requires: the patient reporting subjective symptoms of CI; the HIV physician considering the symptoms to be related (partly or completely) to HIV disease (e.g., potentially confounding non-HIV related conditions have been excluded or are optimally managed); and formal objective neuropsychological assessment, based on multiple cognitive domain testing, confirming CI. Patients who were judged by their HIV consultant to be unable to give informed consent were not considered eligible.

Procedure
Clinicians in each HIV service advised potential participants about the study and referred interested participants to researchers. All participants gave informed consent prior to participating. Interviews were conducted face-to-face (n = 10) or over the telephone (n = 15) depending on patient preference. Interviews lasted 45-97 min and were audiorecorded, transcribed verbatim and checked for accuracy. Interview guides based on a review of the HRQoL literature and experiences of PLWH with CI [29] were used to structure the interview. Questions were open-ended, and probing. Key issues explored included: experiences of living with HIV and CI; how the conditions impact on one another; personal understanding of quality of life; and effect of conditions on HRQoL and on associated domains (impact of difficulties on function, physical health, psychological health, social experiences, and relationships). See Table 1 for the interview key questions and probes. Severity of CI was determined based on most recent MoCA (Montreal Cognitive Assessment [36]) score, which is widely used and well validated screening instrument used in HIV clinics.

Data analysis
A team-based analysis of interview transcripts was conducted to identify components and domains relating to the impact of CI on the HRQoL of PLWH. Data were analyzed using techniques from grounded theory, informed by Hardy and Bryman's (2004) adapted framework which posits guidelines for the iterative relationship between data collection and analysis [37,38]. This approach enables researchers to inductively "discover" categories, themes, and patterns that emerge from the data [37], and is shown in Fig. 1.
In phase 1 of the analysis, five interviews were undertaken and two researchers (KA and SD) read and reread 3 transcripts and gave meaningful segments of text descriptive codes. The researchers met together to identify areas of similarity and differences and agreed upon an initial coding framework. KA then coded the remaining transcripts using this coding framework. In Phase 2, KA carried out a further 14 interviews, which were coded on NVivo 12 (QSR International) to support data collation. The coding was reviewed regularly by KA and SD, who also undertook constant comparison techniques. This involved the comparison of new data and themes with previous coding and transcripts to further develop codes and begin to identify relationships between codes. The process was supported by memo writing, namely identifying areas of uncertainty and possible theoretical explanations, and at this stage four themes were perceived to need further exploration. This involved theoretical sampling, i.e. further data was collected with participants with specific features, in this case, we sought women, those with children, and older PLWH (over 70 years). In phase 3, further interviews with 8 participants took place, themes were further reviewed using constant comparison techniques with the wider research team involved, and seven final conceptual domains were agreed upon. These final domains were identified inductively from the analysis, rather than from existing QoL models, each including sub-themes which were grouped as being the most appropriate fit within each specific domain.

Study participants
Twenty-five PLWH with CI completed interviews. This included 18 men and 7 women, with an age range of 38-80 years. See Table 2 for descriptive data on participant characteristics.

Key domains
Using continued constant comparison techniques among the study team, first yielded six overarching domains which was then expanded to seven following whole study team consultation. The seven interrelated conceptual domains each had a number of descriptive components ( Table 3). Each of the seven domains is described below from the perspective of PLWH with CI.

Physical function
For PLWH with CI difficulties performing at least some instrumental activities of daily living (ADL) (e.g., cooking, shopping, food preparation, housekeeping, laundry, managing finances, managing medications, using technology) were reported by all to varying degrees, and appeared dependent on CI severity (component: activities of daily living). For everyone we spoke to this negatively impacted self-concept/ esteem and sense of independence ( Table 3, example 1). PLWH with CI described attenuated activity levels since the onset of cognitive difficulties and described how they missed doing things previously enjoyed (e.g., reading, holidaying, going to concerts, dancing, seeing friends) ( Table 3, example 2; component: recreational activities). For two participants, substitution for another activity helped to maintain positive self-appraisal. For many, PLWH with CI engaging in outside activities was described as effortful and stressful, with individuals expressing a strong preference for staying at home ( Table 3, example 2). For eight participants, positive perceptions came from maintaining a sense of normality, particularly if they were able to maintain employment via workplace adaption or changing roles (component: employment and vocational activities). Not working was a considerable concern for many PLWH with CI, causing significant financial burden and increasing feelings of uselessness. Individuals reported concerns speaking to their employers about their cognitive issues, fearing they would be labelled as incompetent or that they would need to reveal the cause of their CI (i.e., HIV) resulting in anxiety and stress. For those able to speak to employers, workplace adaptation and continued employment was a strong moderator of good QoL appraisal ( Table 3, example 3).

Cognition
PLWH with CI described issues with short-term recall which impacted a range of activities. Depending on CI severity, individuals described difficulty remembering past events, names or faces, and dates and times (including those of appointments or meetings). Difficulties multitasking impacted ADLs and many described their CI as a 'mental fog' which led to their feeling overwhelmed in busy or loud environments, and resulted in information processing difficulties ( Table 3, example 4; component: memory and executive functioning). All PLWH with CI reported problems attending to and concentrating on activities, along with feelings of restlessness, which impacted their ability to engage in activities lowering self-efficacy and HRQoL (Table 3, example 5; component:

Social connectedness
In the majority of cases, a consequence of cognitive difficulties was social withdrawal. Many PLWH with CI described feeling overwhelmed, distracted, and stressed when in group situations or busy environments, preferring only to socialise one-on-one in quiet places or not at all. Social withdrawal was described as a protective strategy by four participants; one which was necessary, despite its negative impact on HRQoL. Participants described difficulties maintaining social networks (Table 3, example 7). However, where social networks could be maintained, these provided emotional support and consolidation of identity (component: social engagement and withdrawal). While everyone we spoke to described attenuation of social activities following onset of CI, for 18 participants this appeared to compound an already limited social life with few avenues for support: many participants (particularly homosexual males) described limited social contact, with no partner, children, few friends, and limited contact with families ( Table 3, example 8). In contrast, for those participants who had strong support networks, higher levels of social engagement and lower levels of social withdrawal were reported. Having a supportive partner, children, and friends with which they could discuss health concerns openly appeared to be a considerable moderator of HRQoL, which, along with conferring positive affect and reassurance, provided functional assistance and hope (

Stigma
PLWH with CI described considerable variation in community or cultural attitudes towards HIV. Black heterosexual participants all described the stigma and discrimination they would face if others in their community became aware of their HIV status, which affected their ability to discuss cognitive concerns, and so consequently limited their receipt of social support (

Self-concept
PLWH reflected on the how their CI had impacted negatively on their self-esteem, confidence, and independence, with many, referring to themselves using derogatory language, such as "the village idiot", "useless", "worthless" and "stupid" (component: self-esteem and confidence). Individuals who required additional support expressed ambivalence: appreciating the help provided by friends or family while expressing sadness and in some cases guilt at their perceived sense of burden (component: sense of independence). CI appeared to compound already low self-esteem experienced by many as a consequence of living with HIV and, in many cases, having concurrent mental health problems ( Table 3, example 13). This impact on self-esteem, coupled with changes in functional abilities, independence, and social difficulties, led to a sense of identity loss for a number of PLWH with CI we spoke to ( Table 3, example 14). Participants described being 'quieter' and 'reserved' particularly in social situations and spoke with some incredulity and loss when considering their former selves (component: identity). However, many were able to draw on their experiences of living with HIV, which led them to describe themselves as resilient individuals, able to cope with adversity, and who felt the difficulties they had experienced previously helped equip them to deal with their cognitive problems. They spoke of the importance of maintaining a positive attitude, accepting one's difficulties and 'getting on with it' ( Table 3, example 15; component: attitude and personality). This attitude helped combat the fears expressed by many regarding their cognitive health deteriorating, financial insecurities, and becoming a burden to friends and family (Table 3, example 16; component: future self).

Control and acceptance
PLWH with CI who understood the basic pathology of HIV and could describe their cognitive issues tended to report better HRQoL (Table 3, example 17; component: understanding of health conditions). These individuals appeared to compartmentalise their health, seeing and treating health problems (including HIV and CI) as distinct issues. Those reporting worse HRQoL saw their health globally, which increased perceived burden of problems and made it harder to address specific difficulties. Two Black women we spoke to described their illnesses as the result of a 'curse'-this spiritual attribution, hugely increased feeling of distress and powerlessness over health outcomes (

Physical and psychological health and wellbeing
PLWH with CI tended to describe their overall health negatively reporting significant burdens of mental and physical co-morbidities (related and un-related to HIV) (component: mental health, physical health) along with CI which impacted on HRQoL (component: global health appraisal). Particularly striking was the high prevalence of mental health issues, which were reported to have been started or exacerbated by the onset of cognitive problems (Table 3, example 20; component: mental health). Interestingly, five participants framed their difficulties in the context of having survived the AIDS crisis and compared themselves to peers they had lost during this time, describing themselves as 'one of the lucky ones' ( Table 3, example 21). Participants tended to refer to their health-related difficulties in terms of their ability to cope, and whether they were able to feel happiness or, failing that, contentment ( Table 3, example 22). PLWH with CI reported struggling with the relatively uncertain trajectory of impairments, however, for those reporting better HRQoL, strategies which increased perceived control over health outcomes (e.g., lifestyle changes) appeared important. These individuals reported better perceptions of coping and expressed optimistic attitudes towards their cognitive difficulties and overall HRQoL.

Discussion
This study represents the first qualitative study directly exploring HRQoL in PLWH with CI. It provides unique data from a small sample of PLWH with CI on how living with CI impacts and influences HRQoL in PLWH, identifying domains and descriptive components which make up HRQoL in this population. Seven interrelated domains were identified: Physical function, Cognition, Social connectedness, Physical and Mental health and wellbeing, Stigma, Selfconcept and Control and Acceptance. Overall, the participants appeared to describe their HRQoL in negative terms. Many described their cognitive symptoms as impairing functional abilities, including recreational and social engagement. Individuals identified the detrimental effect that this had on their self-esteem and identity, and how living with CI in the context of HIV, was moderated by the impact of stigma on their lives and their perceived sense of control over health outcomes. Additionally, many described how living with CI effected their mental health, with participants describing high levels of depression and anxiety generally, and when considering the relatively uncertain trajectory of their CI. Four domains identified in this study are consistent with findings from the wider literature, including attenuated function [39][40][41][42][43], and impairments in cognition [2,28,44] impacting on HRQoL appraisal. Physical and mental health and wellbeing is supported as a domain as PLWH with CI report worse summative physical [2,[30][31][32] and mental [30][31][32] HRQoL scores. HIV-stigma is a well reported barrier to conversations with friends, family and healthcare professionals about health [2,45], and as in another study [2], participants provided examples of self or enacted stigma which prevented them from talking about their cognitive difficulties. Along with broader information campaigns on HIV-stigma and its impacts, interventions aimed at improving knowledge and facilitating communication of HIV-associated functional health difficulties among friends, family, employers, and healthcare professionals, might help reduce stigma and subsequently improve support [46].
This study contributes to three areas where literature is limited. First, we found a particular impact on social connectedness, with participants reporting high levels of social withdrawal which negatively affected HRQoL and limited receipt of social support. Within this, a novel finding is the role of religion or religious community on HRQoL. Individuals described the social support and safety felt within their religious communities yet reported how discussion of health concerns, specifically HIV or HIV-related illnesses, would not be appropriate or well received. This warrants further examination. We further identified the importance of self-concept on HRQoL in PLWH with CI, the subthemes of identity, attitudes and personality, and future self-concerns, are relatively unexplored. Self-concept has been identified as an important domain of HRQoL for individuals with dementia [23], but is not captured in HIV-specific conceptualisations. Finally, among our participants we found factors related to the control and acceptance of CI to be important moderators on HRQoL, which require further exploration. While other studies have reported on the use of compensatory or remediation strategies in PLWH with CI [2], we extend this to show how among our participants acknowledgement and understanding of conditions facilitates strategy implementation and increases perceived control over health-outcomes, which in turn improves HRQoL. Interventions, which, from the point of diagnosis forward, regularly review and remind patients of factors impacting cognitive health, along with collaborative practical assistance which engender confidence and empowerment, may be helpful in instilling a sense of control over health outcomes and improving overall QoL appraisal.
This study has important limitations, it was an exploratory study limited to urban south-east England. The sample was made up of predominately MSM and Black heterosexual women all of whom were accessing specialist HIV services, and as such is unlikely to be fully representative of HIV populations in the UK. Furthermore, there may be different experiences of HRQoL in PLWH with CI in mainstream services, not examined in the current study. Due to our limited sample we did not explore certain demographic and clinical differences (e.g. including duration of time on antiretroviral therapy, time since HIV diagnosis and other lifestyle factors) across our sample. Similarly, among our participants the social context within which one lives with HIV appeared to be a strong moderator of life quality in the context of CI and we noted cultural differences between MSM and Black heterosexual women which would benefit from further study as would other groups.
There are, however, three main strengths to the study. First, it is the first study to examine and delineate the domains impacting HRQoL in those with HIV and CI. The World Health Organization has identified quality of life as an essential consideration in the care of PLWH [47] and there have been recent calls to include it as a 'fourth 90' in the 90-90-90 testing and treatment targets proposed in 2016 [47,48]. With an ageing HIV population, we know the numbers with CI will increase, policy in this area will need to change, and HIV and memory services will need to expand and adapt to the specific needs of this group. Second, using a good quality qualitative, 'bottom up' approach to data collection and analysis, we have enhanced the validity of results and accessed the experiences of this group to generate new insights into HRQoL in PLWH with CI. Third, the conceptualisation of HRQoL we have identified indicates there are domains of HRQoL which are not captured in generic, HIV, or dementia-specific conceptualisations. This appreciation of condition-specific multidimensionality helps clinical understanding of illness impact, and therefore can help to inform the development of tailored interventions to improve outcomes for PLWH with CI.

Conclusions
This study offers new evidence to help understand the HRQoL of PLWH with CI. We provide evidence suggestive of the poor HRQoL experienced by many PLWH with CI and identify domains where future care, policy and research developments could be focused. Caution is, however, encouraged given that this was a small sample comprised of predominately MSM and Black heterosexual women in the South-East of England. Future research should examine the relationships between domains identified and how different sociodemographic and clinical factors influence HRQoL. Moreover, the experiences of other key groups, such as Black heterosexual men and transgender PLWH should be considered in qualitative work going forward.