Medicine burden experiences of people living with HIV and association with stigma

ABSTRACT The medicine burden of people living with HIV (PLWH) is unknown. Between 2018 and 2020, participants completed a survey comprising outcome measures for medicine burden (LMQ-3) and stigma experiences (SSCI-8). Participants were HIV+ adults (≥18 years), using antiretrovirals (ARV) with or without non-ARV medicines, recruited via two outpatient clinics in southeast England and online via HIV charities across the UK. Spearman’s correlations between medicine burden levels and stigma scores were calculated. Participants were mostly males (72%, 101/141) of mean (SD) age 48.6 (±12.31) years. Total number of medicines ranged from 1-20. High medicine burden was self-reported by 21.3% (30) and was associated with polypharmacy (≥ 5 medicines) (101.52 Vs 85.08, p = 0.006); multiple doses versus once daily regimes (109.31 Vs 85.65, p = 0.001); unemployment (98.23 Vs 84.46, p = 0.004); and ethnicity (97 Vs 86.85, p = 0.041 for non-White versus White participants). A correlation between medicine burden and stigma was observed (r = 0.576, p < 0.001). The LMQ-3 demonstrated adequate construct validity and reliability (domain loadings ranging 0.617-0.933 and Cronbach’s α of 0.714-0.932). Assessment of medicine burden and psychosocial stigma in PLWH could enable identification of those needing additional support in future research and practice.


Introduction
The UK surpassed the United Nations' 90:90:90 goals, with 98% of people living with HIV (PLWH) on longterm antiretrovirals (ARVs) to manage this chronic condition (Kirby, 2018).For the 105200 PLWH in the UK, engagement with care is critical for optimal health outcomes and prevention of HIV transmission (Genberg et al., 2016).Due to advances in treatments and successful early initiation of ARVs, by 2030, 75% of PLWH will be aged 50 years or older, and up to 28% will have at least three comorbidities (Smit et al., 2015).HIV increases individuals' risk of co-morbidities (e.g., cardiovascular-, mental health-, bone-and renal-conditions) and is associated with polypharmacy (using ≥ 5 medicines).
Managing ARVs and non-ARV medicines, which now account for more than 50% of all medicines used by PLWH, can be challenging.Up to 65% of PLWH in the UK use at least one non-ARV medicine while 17% experience non-ARV polypharmacy, thereby juggling the demands of concomitant medicines use (Okoli et al., 2020).Polypharmacy is associated with increased incidence of adverse drug reactions, druginteractions, and medicine burden (Back & Marzolini, 2020;Carter, 2018;Edelman et al., 2020;Guaraldi et al., 2017).Prescribing cascades, aiming to manage side effects of ARVs, may further complicate medicine use and contribute to problematic polypharmacy (Marzolini & Livio, 2019).
Medicine burden is "a patient's subjective experience in response to physical, psychosocial and financial impacts of medicines to maintain their daily lives, health and well-being [and involves] adapting to challenges of living with a medicine" (Mohammed et al., 2016).Medicine burden, including but not limited to side effects and interruptions to daily routines, impacts adherence to treatments and contributes to disengagement with care and treatment discontinuation (Claborn et al., 2017).
Various systematic reviews and meta-analyses have shown that lower adherence to ARVs is associated with more complex regimens, leading to higher risk of treatment failure and risk of antiretroviral drug resistance (Altice et al., 2019;Bhatta et al., 2017;Shah et al., 2019;Shubber et al., 2016;Taiwo et al., 2023).Few studies have attempted to determine extent to which HIV-related treatment burden affects PLWH generally (Schreiner et al., 2019;Tran et al., 2019).We are not aware of studies that have explored HIV-related medicine burden in the UK.
Despite HIV services and ARVs being offered free within the National Health Service (NHS), some PLWH in England may face challenges with making co-payments for their non-ARV prescriptions.Organising and attending different healthcare appointments, and difficulties accessing some or all their medicines can be challenging (Tran et al., 2019).Moreover, stigma and secrecy related to HIV (and medicine use) may further exacerbate actual or perceived medicine burden (Katz et al., 2013;Pound et al., 2005;Shubber et al., 2016).Higher levels of HIV-related stigma and not sharing information about HIV status are also associated with nonadherence to ARVs (Katz et al., 2013).Research is needed to understand medicine burden in PLWH and to ascertain the extent to which it correlates to stigma.
Person-centred care, particularly monitoring health outcomes from the patient's perspective, is increasingly accepted, despite predominant biomedical approaches to HIV care (Bristowe et al., 2019).Patient-reported outcome measures (PROMs) play a role in supporting person-centred care and they are continuing to emerge as relevant tools for identifying challenging aspects of HIV care (Bristowe et al., 2018).We measured the extent of medicine burden as perceived by PLWH using a valid and reliable instrument.The Living with Medicines Questionnaire (LMQ-3) is a 41-item, 8domain, generic, PROM for medicine burden which measures a broad range of medicine-related issues which affect everyday lives of people using medicines long-term (Katusiime et al., 2018;Krska et al., 2018).It has been used internationally to assess medicine burden in the general population (Tordoff et al., 2019) and in older Chinese over 50 years living with HIV (Zheng et al., 2022) but not in other PLWH.This study also investigated factors correlated with medicine burden and stigma.

Design and sample
Between 2018 and 2020, a cross-sectional survey was conducted to gather views of eligible HIV + adults, aged 18 years or older, using ARVs for at least 6 months.Participants who were not using ARVs long-term or on pre-exposure prophylaxis were excluded.We also excluded those unable to understand English to a level necessary for providing consent to study procedures; logistical challenges prevented the use of translation services.

Data collection and setting
Participants were recruited through two outpatient HIV clinics in southeast England.Screening to identify eligible patients was done by the care team.Eligible participants were informed about the study, provided an information pack, and gave their written consent.The anonymous survey could be completed in the clinic or at home.A printed paper version or electronic survey was available with the latter accessed via a scannable QR code, or online link.A prepaid envelope was provided for paper surveys.Survey links were also promoted via social media and websites of eligible patient organisations providing HIV-related services/support across the UK.Informed consent was recorded for all survey participants via Qualtrics™ platform.

Measures
The LMQ-3 assessed overall medicine burden, which was defined as a sum of all participants' collated ratings on the questionnaire, in eight domains: practical difficulties (7 items), general concerns (7 items), interferences to day-to-day life (6 items), perceived effectiveness (6 items), patient-doctor relationships/communication (5 items), side effects (4 items), cost-related burden (3 items), and autonomy/control over medicine use (3 items).Items were rated on a 5-point scale (strongly agree to strongly disagree), and reverse scoring used as appropriate, with higher scores depicting higher medicine burden.Overall medicine burden was then categorised using previously published cut-off values for levels of burden (low, moderate, high as 41-87, 88-110, and > 110 respectively).The LMQ-3 also includes a section for obtaining free-text qualitative data and participant characteristics (Katusiime et al., 2018).
The 8-item Stigma Scale for Chronic Illnesses (SSCI-8) is a generic measure of stigma related to living with a long-term condition.It was used, with permission, to assess forms of stigma: internalised or self-stigma correlated with awareness of negative stereotypes around an individual's illness (2 items) and enacted/actual experiences of stigma and discriminatory behaviour (5 items) or internalised and enacted stigma (1 item).Items were rated on a 5-point scale (never, rarely, sometimes, often, always) and total scores calculated (range 8-40) with higher scores indicating greater stigma (Molina et al., 2013;Rao et al., 2009).We hypothesised a priori that stigma would be positively correlated with medicine burden.

Data analysis
Validity and reliability of the LMQ-3 were assessed by conducting a confirmatory factor analysis (CFA) and measuring Cronbach's alpha for the eight domains respectively.To investigate the correlations between medicine burden and stigma, Spearman's correlations were calculated using scores obtained by the previously described outcome measures with p-values < 0.05 considered statistically significant.

Ethics
Ethics approval was obtained from the Northeast Newcastle and North Tyneside Research Ethics Committee for recruitment via outpatient NHS clinics (REF 18/ NE/0321) and Medway School of Pharmacy Research Ethics Committee (SREC) for the UK-wide on-line survey (REF001018).

Participant characteristics
A total of 217 participants accessed the survey.There were variations in response rates by the source of data and recruitment methods used: relatively more questionnaires (66.8%, 145) were returned from participants using social media/online platforms of HIV-related patient organisations/charities or support groups compared to sample data collected from two outpatient clinics in southeast England (33.2%, 72).Non-responders were mostly those who viewed the electronic survey-no demographic data were provided by all nonresponders.The first and last LMQ-3 items were completed by 175 and 154 participants, respectively.Two thirds (141, 65%) completed all LMQ-3 items.Seventy-two percent were male, of mean age in years 48.6 (SD ± 12.31); age range 18-79 (Table 1).

Experiences of medicine burden
High-and moderate-levels of medicine burden were experienced by 21.3% (30/141) and 29.1% (41/141) respectively.The median LMQ-3 total score was 88 suggesting moderate medicine burden for this sample population (range 57-170).Table 3 shows results from LMQ-3 statements.Notably, over a third (36.7%, 62/169) were concerned that they may forget to use their medicines, while 11.4% (18/159) felt it was neither easy to keep their medicines routine nor were they comfortable with the timing of their regular doses (5.7%, 10/174).Within the domain  "interference to daily life, a third (32.9%, 51/155) agreed/ strongly agreed that their medicines interfered with their sexual life while 23.6% (37/157) had interferences to their social relationships.Quotes taken from participants' written comments illustrate aspects of medicine burden in the eight LMQ-3 domain areas (Table 4).

Factors correlated with medicine burden
Polypharmacy was significantly correlated with medicine burden (Figure 1).Mean LMQ-3 burden scores were higher among participants using 5 or more medicines, those using medicines more frequently than once daily (Table 5).Data showed statistically significant differences in medicine burden scores with respect to employment status.Domain analyses showed cost burden scores were higher in the unemployed group (mean ± SD 6.44 ± 3.31) compared to those employed (5.04 ± 2.38) (p = 0.018) and comparable results were obtained with Mann-Whitney tests (mean rank, 80.27 Vs 62.46 respectively, p = 0.012).Non-White participants had higher medicine burden scores (97.00 ± 24.12) than White participants (86.85 ± 20.33, p = 0.041); domain analyses revealed statistically significant differences in the following areas: practical difficulties (16 Vs 13, p = 0.006), cost-burden related to non-ARV prescriptions (7.0 Vs 5.2, p = 0.005), general concerns (21.3 Vs 18.3, p = 0.017), and lack of autonomy (10.5 Vs 11.9, p = 0.021).

Correlations between medicine burden and stigma
A third (32.8%) of the participants often or always felt embarrassed about living with HIV (and taking medicines) while 22.8% often or always felt blamed for their HIV diagnosis (Table 6).A 49-year-old male illustrated the burden of stigma: "some medicines are arguably difficult to take in front of people as this may Table 4. Qualitative analysis of free-text comments in the eight medicine burden domains.

Medicine burden domains
Experiences of medicine burden reported by PLWH

Practical difficulties
Nearly a quarter (22.4%) of participants felt that they put a lot of planning and thought into using their medicines, while over 10% felt it was difficult getting prescriptions (12.3%) and medicines from the pharmacist (13.5%).Other practical challenges related to demands of strict adherence and associated food/drink a 59-year-old female on 4 medicines revealed that "the need for food with my ART means I must always have an evening meal.This is sometimes annoying when, say, going to a concert".Nonetheless, some participants had developed routines to manage any practical difficulties associated with their medicines, as illustrated by a 61-year-old male on 3 medicines "I don't have any problems taking medicines.I have developed a system that works well for me".However, changes to individual routines appeared to disrupt medicine use behaviours of some as illustrated by a 49-year-old male on 3 medicines: "I am in a good routine with my medication, so very rarely have any issues with them.Only in unusual circumstances, like being on holiday, may I have to think more carefully about how I am going to take my medication".

Interferences to day-to-day life
A third of participants (32.9%) reported medicine-related interferences to their sexual life, and to their social relationships (23.6%).One in five (19.1%) participants felt that medicines interfered with their social or leisure activities.For 17.2% of participants, using medicines caused problems with daily tasks (including work) as illustrated by a 54-year-old male: "I do have some difficulty in juggling my meds with working nights and it does cause some issues in taking them or missing sometimes".

Effectiveness /Efficacy
As expected, most participants (95.8%) felt their medicines prevent their condition getting worse, and that their medicines were working (94.1%).Most expressed beliefs about necessity of their ART medicines and gratitude towards enabling them to achieve a healthy, long life.Participants' own risk-benefit assessments appeared to influence their views on effectiveness as illustrated by a 65-year-old female on 3 medicines: "There is no choice as I have HIV.No one wants to take medicines for life, but if I did not, I would be long dead so I'm lucky effective treatment for HIV exists".

Side effects
A quarter (25.5%) of participants agreed that side effects interfered with daily activities and that side effects were bothersome (24.2%).18% participants felt that the side effects were worse than the condition for which they used medicines and some reported concerns about counteractive prescribing cascades.The latter was illustrated by a 66-year-old male on 8 medicines: "Side effects have caused me problems that continue today, even though I have changed from the meds that gave me problems.Old HIV drugs left me with lipodystrophy and peripheral neuropathy.I take meds for the pain, but these cause erectile dysfunction which affects my sex life.I have to take statins because the HIV meds raise cholesterol and lipids, and the statins give me muscle pain".Communication/relationships with healthcare professionals Most participants reported good relationships and communication with healthcare professionals, as demonstrated by a 46-year-old male on 3 medicines: "The team at clinic [ x] in my opinion are outstanding".
Up to 75% felt that their doctor takes concerns about side effects seriously, and a similar proportion (77.7%) felt the doctors listened to their opinions about their medicines (Table 3).10% wanted more information about their medicines from their doctors

Concerns about medicine use
Nearly two-thirds of participants were concerned about possible long-term effects of ART (62.9%), while 23.4 -43% were concerned about polypharmacy and potential drug interactions respectively.These concerns were exemplified by a 54-year-old male on 2 medicines "My only concern is the unknown long-term effects" and by a 52-year-old female on 2 medicines "It's hard when you have HIV and have diabetes as well when both drugs fight against each other".Some were concerned about adherence to ART were revealed by a 57year-old female "As I live my life sometimes the days are so normal, I forget to take my medication even with my alarm and dosette box"., A 32-year-old male on 2 medicines acknowledged the burden of taking longterm ART "It's every day, forever.The burden is so high!"About a third (29.2%) were concerned about genericbrand switching of ART, including a 47-year-old male on two medicines: "My medication brand changed recently which caused me some concern.I hadn't been informed about it so when the medication arrived, I didn't recognise the packaging or the ingredients in the tablets.It did mean I wasn't 100% sure I should take them … ."

Cost burden
For most participants, the cost of medicines was not a burden to them.as illustrated by a 32-year-old female: "The cost of my medication [ART] doesn't affect me as I live in the UK, and it's completely covered by the NHS.I don't have to worry about finding the money for them, ever".However, a quarter (25.7%) worried about paying for their non-ART prescriptions.Some felt that the prescription co-payment was more than they could afford (7.2%): "I find affording supplementary medicines a problem sometimes, I am on the wrong kind of ESA [Employment and Support Allowance]" 55-year-old male on 6 medicines.

Patient autonomy over medicine use
Most participants reported limited autonomy to alter their regimes to suit personal lifestyles.Just over twothirds (68.1%) felt that they little or no choice in deciding whether or not to use their medicines as illustrated by a 54-year-old male on 5 medicines: "I either take my medicines or die, no contest!"Almost a third (32.3%) felt able to vary the times they use their medicines, while 8.4% felt they could change the dose of their medicines.
Qualitative free text comments ART-Antiretroviral therapy, also used interchangeably with ARVs (Antiretrovirals).
lead to being stigmatised".Some participants were concerned about the size of tablets, like this 58-year-old female "I don't like the size of some of them as it makes it difficult to be discreet when taking them in public".Medicine burden scores were strongly positively correlated with total stigma scores (0.576, p < 0.001 respectively).The correlation between total LMQ-3 scores and internalised/perceived stigma (0.543) was stronger than total LMQ-3 and enacted/actual stigma (0.441).Of the eight LMQ-3 domains, "general concerns' had the strongest positive correlation with total stigma scores (rho = 0.565, p < 0.001) (Table 7).

Discussion
To the best of our knowledge, this is the first study investigating the extent to which PLWH experience medicine burden and correlations with stigma.High medicine burden was experienced by 1 in 5 participants (21%), slightly less than that found in the UK's general adult population (23.5%) (Krska et al., 2018).All eight domains of medicine burden were experienced by some individuals to varying magnitudes.For instance, "medicine-related interference" affected up to a third of all participants' social relationships and sexual life, a bigger impact than that seen in the general population.Side effects affected around 25% and generated most of the free-text comments compared to around 19% in the general population; General concerns about possible long-term effects of medicines were more common in PLWH than in the general population (62.9% versus 54.4%); PLWH may experience significant medicine burden from either ARVs or non-ARV medicine, although likely to report less burden due to placing more emphasis on medicine need and gratitude for lifesaving ARVs.Stronger beliefs around necessity, effectiveness, and longevity may explain the lower prevalence of medicine burden among PLWH (Fall et al., 2014;Horne et al., 1999).More PLWH are living longer due to advances in antiretroviral medicines, and this may influence risk-benefit assessments in this population (Pound et al., 2005).Further qualitative research is needed to fully investigate perceptions of burden in this population.
Factors significantly correlated with increasing medicine burden included polypharmacy (5 or more medicines), multiple daily dosing regimens, and needing a carer to managing medicines use.Like the general population, higher medicine burden increased with the total number of medicines and more frequent daily use (Katusiime et al., 2018;Krska et al., 2018).Polypharmacy among PLWH is increasing (Edelman et al., 2020;Halloran et al. 2019).Nearly half (47%, 2195/ 4630) of PLWH in the UK use two or more non-ARVs plus their usual ARV regime; the prevalence of polypharmacy is even higher in among PLWH over 50 years of age (61%) (Okoli et al., 2020).Polypharmacy   contributes to drug-drug interactions, adverse drug reactions, nonadherence and poor health outcomes (Carter, 2018;Edelman et al., 2020;Guaraldi et al., 2017;Okoli et al., 2020).More frequent use of ARVs and non-ARV medicines (> more than once daily) was associated with higher medicine burden.Practical difficulties, particularly concerns about adherence and timing of doses were also more common in PLWH than in the general population (36.7% compared to 25.2%).Lack of autonomy to vary the timing of doses appeared to have a greater influence on experiences of medicine burden for PLWH (55.7%) compared to the general population (45.8%) (Krska et al., 2018).Factor analyses of our sample data showed a stronger correlation between "lack of autonomy" and medicine burden (0.674, p < 0.001) compared to the general population coefficient (0.1, p = 0.224) (Katusiime et al., 2018), suggesting that autonomy is a crucial factor to consider in medicine-related consultations for PLWH.Not surprisingly, fewer PLWH reported being able to vary the dose and/or timing of their medicines due to a greater emphasis on adherence to achieve undetectable viral loads (Altice et al., 2019).Pound et al. (2005) hinted on the potential burden of strict adherence to HIV medicines, which can be very demanding and disruptive to life (including mealtimes, sleep, social life) leading to lack of control and regimen fatigue.Patient autonomy with regards to timing of ARVs, where clinically appropriate, may reduce burden for PLWH and achieve person-centred care.Interventions to reduce regimen complexity are key to preventing non-adherence (Zhou et al., 2014).Long acting antiretrovirals such as once-a-month cabotegravir/rilpivirine injections are promising (Voelker, 2021), yet to become adopted routinely within practice.
Increasing stigma was correlated with increasing medicine burden.PLWH may experience stigma associated with the condition itself, using medicines, or older age.
Higher stigma may exacerbate fatigue and contribute to cognitive decline (Rao et al., 2012) impacting medicines use and adherence behaviour.Our study showed higher medicine burden levels in non-White participants.Rao et al. (2016) who used a similar measure of stigma found higher internalised stigma among African Americans living with HIV.Medicine burden interventions need to be inclusive for all PLWH, in addition to providing more support to ethnic minorities.PLWH needing a carer to support medicine use perceived a higher medicine burden than those managing independently, like the general population (Krska et al., 2018).Decreasing treatment burden is correlated with increasing social support (e.g., from close friends and family) (r = −0.2,p = 0.03) (Schreiner et al., 2019).However, perceptions or experiences of stigma may deter PLWH from seeking social support and those requiring formal carers may feel burdened by issues around unintended disclosure of HIV status.

Implications for practice and policy
To the best of our knowledge, this is the first study to assess medicine burden among PLWH with a validated generic measure.The LMQ-3 could help to identify those needing additional support given the increasing multimorbidity and polypharmacy in this population.Medicine reviews should be prioritised for PLWH experiencing polypharmacy including both ARVs and non-ARV medicines.We know that the number of medicines alone is not a sufficient indicator of medicine burden (Krska et al., 2018), therefore other factors, particularly the timing of doses and how this affects everyday life should be considered by clinicians, as well as experiences of stigma.Additional support for ethnic minorities needs to be considered when prioritising patients for intervention.As HIV treatment is lifelong, clinicians and policymakers should consider integration of medicine burden measures (such as the LMQ-3) in future HIV care models to achieve person-centred care.Future healthcare funding policies in England should include HIV on the list of long-term conditions exempted from the NHS prescription charge like exemptions provided to people with certain life-long conditions (e.g., diabetes, hypothyroidism, epilepsy) especially for those on low household income.Other studies involving larger samples of PLWH are needed to study how medicine burden impacts adherence and clinical outcomes (e.g., viral load).

Strengths and limitations
Online recruitment methods, via social media and websites of HIV-related patient organisations and within selected HIV clinics, were used to widen inclusion of participants across the UK, contributing to a relatively small but adequate sample size.We do not anticipate that the sample population was unusual, especially that increasingly more PLWH seek support from online platforms.This may have increased during the data collection period which partially encompassed the COVID-19 pandemic.A survey led by one of the UK's leading women's HIV charities (n = 75) found that nearly half (45%) of participants had problems accessing HIV care during the lockdown period, with a quarter (25%) highlighted their main concern as "lack of access to their HIV clinic".(National AIDS Trust, 2020).It is possible that these recruitment methods may have biased the sample to younger persons or those with access to the internet.
Validated instruments (the LMQ-3 and SSCI-8) were used, allowing measurement of medicine burden and correlations with stigma.This is the first application of a validated PROM of medicine burden in PLWH within the UK, and confirmatory factor analyses demonstrated adequate construct validity and reliability.All eight domains of the LMQ-3 loaded strongly on the overarching construct of medicine burden, suggesting reliability of the instrument.Nevertheless, missing data may be due to potential survey fatigue from the questionnaire length.In addition, no demographic data were collected for non-responders to the electronic or paper-based survey.Regardless, a shorter, tailored, version of the LMQ-3 may benefit future studies.Although demographic characteristics of our sample comprised mostly White males, this was representative of the HIV population in the UK.The sample population, however, may not fully represent the views or experiences of all PLWH.

Conclusion
One in five PLWH experience high medicine burden.Our data showed that higher medicine burden is correlated with increasing stigma, polypharmacy (5 or more medicines), multiple daily dosing (more than twice daily), as well as needing support with medicines use (e.g., from a carer).Of the eight aspects of medicine burden, lack of autonomy to vary the timing of medicines appeared to be the highest contributor to burden for PLWH.The LMQ-3 is a valid and reliable outcome measure in this population, but further work is needed to harness a larger sample.This study confirms that perceptions and experiences of medicine burden are related to stigma, both of which can affect an individual's adherence to treatments for HIV and co-morbidities.
Highlighting this interrelationship has the potential to shape the development of future interventions aimed at reducing the burden to improve overall health and quality of life for PLWH.

Figure 1 .
Figure 1.Medicine burden category of participants by number of medicines used.

Table 2 .
Medicine characteristics of participants.
Note: Due to missing data in the last section of the survey, not all participants completed questions about medicine regimes used.(*NHS-National Health Service -payment applicable to England residents only)

Table 1 .
Demographic characteristics of participants.

Table 3 .
Participants' responses to the statements in the LMQ-3 survey.

Table 5 .
Factors associated with medicine burden among people living with HIV.

Table 6 .
Responses to the eight stigma items in the stigma questionnaire.
SSCI -Stigma Scale for Chronic Illnesses Questionnaire.

Table 7 .
Correlations between medicine burden and stigma.