Abstract
Background
People experiencing homelessness (PEH) are amongst the most marginalised populations in the society. Despite facing extreme health inequalities and poorer outcomes than the general population, they are under-represented in clinical trials. Little is hence known about the physical, mental health and social characteristics of PEH as clinical trial participants. Cross-sectoral complex interventions aiming to mitigate health, wellbeing and wider inequalities faced by PEH are urgently needed.
Methods
This paper reports on baseline characteristics of PEH from the PHOENIx Community Pharmacy multicentre Pilot randomised controlled trial (RCT) with recruitment from community pharmacies in Glasgow and Birmingham, UK. Participants were randomised 1:1 to receive PHOENIx (collaboration between National Health Service pharmacist independent prescribers and third sector workers) intervention in addition to usual care (UC) or UC only. Data were collected using face-to-face patient questionnaires which included validated tools (e.g. Equation 5D), direct clinical observations (e.g. blood pressure, grip strength) and healthcare utilisation records (e.g. hospital admissions).
Results
A total of 100 participants were recruited as planned from five community pharmacies. Participants (n = 99 baseline records for one were lost) were on average 42 years old, mostly white (87, 88%), and male (81, 82%). Most had recently slept rough (57, 58%), were registered with an Alcohol and Drug Recovery Service (79, 81%), and prescribed opioid substitute therapy (74, 75%). A minority reported having a recent street substance-related overdose (15, 15%). Most participants reported one or more chronic physical health condition (83, 85%). Fifteen (15%) participants rated severity of breathlessness as 4. Thirty-four (44%) participants were deemed pre-frail and 43 (56%) as frail. Participants had an average EQ-5D-5L health-related quality of life score of 42.2 (SD 24.7).
Seventy-six (77%) participants reported having a mental health condition or difficulty; depression (n = 56, 56%), anxiety (n = 47, 47%), self-harm/suicide attempt (n = 21, 21%). Only half were known to housing/casework services. Moreover, half walked to appointments and a minority of those diagnosed with a blood borne virus were in receipt of treatment. Over three quarters had been imprisoned in the past.
Conclusions
Baseline characteristics of PEH who participated in the PHOENIx Community Pharmacy pilot RCT demonstrates multiple, complex and unmet health, social and practical needs. The data emphasises the need to engage, support, refer and treat PEH through integrated and bespoke pathways to promote people’s engagement with care, prevent repeat homelessness and early deaths.
Trial registration
International Clinical Trials Registration ISRCTN88146807. Date of registration 18/07/2022.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12939-025-02627-8.
Keywords: Homelessness, Community pharmacy, Inclusion health
Background
People experiencing homelessness (PEH) have 12 times higher mortality than the general population [1]. The average age of death of PEH men and women in the UK is 45.4 years and 43.2 years respectively; the majority of early deaths resulting from drug poisoning, alcohol consumption and suicide [1–3]. Substance use and injuries are also leading causes of emergency department presentations in PEH [4]. Homelessness is increasing worldwide, including rough sleeping and other forms of homelessness [5].
Homelessness often overlaps with history of abuse, offending behaviours, substance dependency, domestic violence and mental health problems and these co-presenting inequalities are often collectively referred to as severe and multiple disadvantage [6]. Poverty, especially childhood poverty correlates strongly with homelessness [7]. Adverse childhood experiences, incorporating potentially traumatic events prior to 18 years of age, have a higher prevalence in PEH than the general population [8]. Unsupported and temporary shared living space accommodation presents challenges, particularly for PEH who are substance dependent due to the availability of street substances, a paucity of outreach by health services, and the re-traumatising effects of the accommodation itself [9]. Poor discharge planning from hospital to accommodation for PEH, and a lack of service co-ordination negatively impact on the outcomes for PEH [10]. Assertive outreach is one way to successfully engage with PEH and is recommended by the National Institute for Health and Care Excellence (NICE) [11].
Community pharmacies and people experiencing homelessness
Evidence on the role of community pharmacies supporting PEH is mixed. PEH receiving opioid substitution therapy (OST) have reported facing stigma in community pharmacies, perpetuating social exclusion and worsening existing health inequalities [12]. Community pharmacies however provide a range of healthcare and harm reduction services such as needle exchange, and relevant to health needs of PEH [13]. Studies highlight lack of education, training and clinical guidelines in counselling and management of PEH in pharmacies [14]. Despite many PEH utilising community pharmacies, there is a paucity of research on the role of community pharmacies in supporting PEH to focus on their holistic health, social and practical needs. Community pharmacies are ideally located, often in principal sites, with no appointment required, creating scope to proactively identify and collaborate with other teams to support the management of PEH.
PHOENIx model of care
PHOENIx is an outreach model of care consisting of NHS employee clinical pharmacist independent prescribers collaborating with third sector homelessness link workers to engage with PEH, assess health and wider needs, and intervene. PHOENIx acts as an important conduit back into mainstream care for PEH, while providing direct clinical care also. The PHOENIx intervention originated in Glasgow and previously underwent feasibility testing [15] and pilot in other community/outreach settings [16].
Randomised controlled trials in people experiencing homelessness
There is a lack of ‘gold standard’ RCT evidence on what works to meet the needs of PEH (Supplementary Material 1). There is a need for high quality studies evaluating collaborative health and social care outreach interventions to address wider needs. Previous RCTs involving healthcare professional interventions in PEH in the UK include a multicentre trial of General Practitioner-led hospital in-reach, and a pilot RCT of the PHOENIx intervention for PEH with a history of overdose which was delivered in Glasgow [16]. Remaining gaps include whether community pharmacies can play a role in the identification of PEH, and whether PHOENIx can be delivered successfully in a multicentre study.
Given the important gap around inclusion of PEH in RCTs and hence reflecting inequalities in opportunities for research participation, very little is known about health, social and practical needs of PEH who participate in trials. The purpose of this paper is to report baseline characteristics of PEH identified in community pharmacies who participated in the multicentre PHOENIx Community Pharmacy Pilot RCT and to evaluate how these characteristics compares with the existing literature.
Methods
A protocol paper with full methods for the pilot RCT have been described previously including identification of potential participants, the informed consent process, clinical governance, outcome measures, intervention fidelity, and details on the process evaluation and economic evaluation [17]. The primary focus of the Pilot RCT was to determine the feasibility of proceeding to a definitive study.
Design and setting
Set in Glasgow (Scotland) and Birmingham (England) UK, this study was a multisite, prospective, parallel group, pilot RCT with embedded economic and qualitative process evaluation. Participants were recruited from five participating community pharmacies (2 in Glasgow and 3 in Birmingham), based in city centre locations. Table 1.
Table 1.
Recruitment distribution from community pharmacies
| Recruiting community pharmacy | Participants (n = 100) |
|---|---|
| Glasgow community pharmacies | |
| Pharmacy 1 | 35 |
| Pharmacy 2 | 15 |
| Birmingham community pharmacies | |
| Pharmacy 1 | 5 |
| Pharmacy 2 | 27 |
| Pharmacy 3 | 18 |
Participants
Adults aged 18 years and over, experiencing homelessness and utilising one of the five participating community pharmacies were eligible to take part in the study. The European Typology on Homelessness and Housing Exclusion (ETHOS) definition was adopted for the study including: rooflessness, houselessness or insecure or inadequate housing [18]. This includes people who sleep rough (street dwellers), living in temporary and night shelters, sofa surfing, and people experiencing any other forms of temporary or repeat homelessness. Eligibility was established through prescription records (e.g. annotation of ‘no fixed abode’ within address field in the prescription) or a reference to specialist homelessness general practices as care providers). Participants’ acknowledgement of homelessness was the most important source of information for eligibility assessment.
Utilisation of the community pharmacy by potential participants could relate to range of reasons including OST dispensing, wound management, needle exchange services or dispensing of other prescribed medication. Exclusion criteria for the study were: living in residential or community-based rehabilitation which has direct access to in-house medical and/or nursing care or intoxicated or (in the opinion of the researcher) posing a safety risk to staff and lacking capacity to consent.
Glasgow, Scotland
Glasgow city has the lowest life expectancy in Scotland [19]. The largest city in Scotland, Glasgow is synonymous with high levels of excess mortality (mortality that is not explained purely in terms of differentiating socioeconomic deprivation), hypothesised in terms of significant socioeconomic and political events. The so-called ‘West Central Scotland Effect’ is derived from Glasgow recording 30% excess deaths per head of population than Liverpool and Manchester; cities with comparable levels of socioeconomic deprivation [19]. A substantial number of individuals are estimated to habitually inject substance in Glasgow city centre, prompting calls for safer injecting facilities and an Enhanced Drug Treatment Service (EDTS); an initiative where pharmaceutical grade diamorphine is prescribed to patients with complex health and social care needs. Until recently, Glasgow was the only local authority in Scotland to house dispersed asylum seekers from overseas and hosts the largest number of asylum seekers per head of population than any other local authority in the UK [20].
Birmingham, England
Birmingham, the second most populous city in the UK and the largest local authority in Europe, has a high ethnically diverse population [21]. Located in the West Midlands region of England, Birmingham, like Glasgow is an area of high socioeconomic deprivation, with 40% of the city living in 10% of the most deprived areas of England [22]. There is a life expectancy gap of 8.9 years for men and 6.6 years for women in the most deprived areas of Birmingham compared to the least deprived areas of the city [23]. High incidence of mental health, substance use, and infectious diseases was observed in PEH in Birmingham in a previous study [24].
The National Health Service (NHS) in the UK is a publicly funded healthcare system and is free at the point of care to all residents regardless of their ability to pay. NHS is the main provider of health services. PEH are eligible to use all services available in the community as well as secondary and specialist care. This includes utilisation of mainstream general practices. There are also specialist homelessness general practices for PEH in most urban areas including the two study cities Birmingham and Glasgow. Further details on usual care provisions are described in our protocol paper [17].
Recruitment, sample size and statistical power
Opportunistic recruitment of PEH from community pharmacies was adopted in our study. Community pharmacy staff would briefly describe the study, prior to introducing them to the independent researcher responsible for obtaining informed written consent. The sample size calculation was based around two of the main feasibility objectives: rate of recruitment and retention; the power calculation has been previously reported [17]. Data summaries (e.g. frequencies and percentage, mean (SD), median [IQR] as appropriate), were generated using SAS v9.4 following transfer from a secure Trial Database System, located at the University of Birmingham.
Data collection
Baseline demographic, physical and mental health information data were collected through face-to-face meetings between researchers and participants using case report forms (CRFs) (Table 2). External data were sourced from community pharmacies, Third Sector Organisations (TSOs), substance treatment services, mental health services, Emergency Departments, data linkage from NHS, and Scottish Prison Service. Baseline alcohol, tobacco, overdose, and substance use data were collected through participant facing CRFs and primary care health records (Table 3). Prescribed medications data were taken from community pharmacy and hospital records (Table 4).
Table 2.
Baseline demographic, physical and mental health characteristics (N% or mean (SD)/median [IQR]
| Characteristic | Participants (n = 99) | Missing data |
|---|---|---|
| Age (years) | 42.3 (8.9) | |
| Age range (years) | 24–60 | |
| Ethnicity | ||
| White | 87 (88%) | |
| Asian/Asian British | 3 (3%) | |
| Black, African, Black British or Caribbean | 3 (3%) | |
| Mixed or multiple ethnic groups | 2 (2%) | |
| Any other ethnic group | 2 (2%) | |
| Prefer not to say | 1 (1%) | |
| Sex | ||
| Male | 81 (82%) | |
| Female | 18 (18%) | |
| Age first experienced homelessness | 21.5 [16–33] | 5 |
| Type of homelessness | ||
| Hotel (supported) | 45 (45%) | |
| Hotel (unsupported) | 17 (17%) | |
| No fixed abode (NFA) | 12 (12%) | |
| Other | 10 (10%) | |
| Rough sleeping | 9 (9%) | |
| Temporary furnished flat (TFF) | 6 (6%) | |
| In past three months, slept rough on streets | 57 (58%) | |
| How many nights slept rough on streets | 8 (3,40) | 2 |
| Registered with GP | 87 (88%) | |
| Type of GP registered with (n = 87) | ||
| Mainstream | 53 (61%) | |
| Homeless health | 33 (38%) | |
| Unknown | 1 (1%) | |
| Ever been in full-time employment (yes) | 70 (71%) | |
| Type of employment1 | 1 | |
| Manual work | 29 (41%) | |
| Driving | 5 (7%) | |
| Engineering/technology | 4 (6%) | |
| Factory | 4 (6%) | |
| Catering | 3 (4%) | |
| Armed forces | 2 (3%) | |
| Bar work | 2 (3%) | |
| Supervisor | 2 (3%) | |
| Administration | 1 (1%) | |
| Other | 1 (1%) | |
| Highest educational attainment | 1 | |
| Primary or below | 15 (15%) | |
| Lower secondary (e.g. GCSE or National 5) | 38 (38%) | |
| Upper secondary (e.g. A levels, Highers) | 4 (4%) | |
| Post-secondary (non-tertiary e.g. vocational) | 37 (37%) | |
| Bachelor degree | 2 (2%) | |
| Does participant have children (yes) | 57 (59%) | 2 |
| Number of children | 2 (1,4) | |
| Who cares for children1 | ||
| Mother | 22 (39%) | |
| Grown-up | 17 (30%) | |
| Family | 8 (14%) | |
| In care | 7 (12%) | |
| Ex-partner | 2 (4%) | |
| Participant | 1 (2%) | |
| Ever had counselling (yes) | 38 (40%) | 3 |
| Current debt/loans (yes) | 24 (24%) | 1 |
| Ever been in residential or foster care (yes) | 39 (39%) | |
| Known to alcohol and drugs recovery service (ADRS) | 79 (81%) | 1 |
| Known to mental health services | 30 (30%) | 1 |
| Any other team participants known to | 23 (23%) | 1 |
| Housing/caseworker | 43 (43%) | 1 |
| Currently receiving state support | 95 (96%) | |
| Type of state support receiving (n = 95) | ||
| Universal credit (UC) | 66 (69%) | |
| Personal independence payment (PIP) | 30 (32%) | |
| Housing benefit | 28 (29%) | |
| Employment support allowance | 22 (23%) | |
| Job seekers allowance | 1 (1%) | |
| Carers allowance | 1 (1%) | |
| Incapacity benefit | 1 (1%) | |
| Currently applying for state support | 21 (21%) | |
| How does participant normally travel to appointments | ||
| Walk | 49 (50%) | |
| Bus or public transport | 31 (31%) | |
| Taxi | 9 (9%) | |
| Any other | 4 (4%) | |
| Wheelchair | 2 (2%) | |
| Ambulance | 1 (1%) | |
| Not applicable | 3 (3%) | |
| Participant has transport paid, bus pass | 30 (30%) | 3 |
| Does anyone go with participant | 33 (35%) | 6 |
| Number of participants with a chronic physical health condition | 83 (85%) | 1 |
| Number of chronic physical health conditions | 1 | |
| 0–1 | 22 (22%) | |
| 2–4 | 29 (29%) | |
| 5–8 | 25 (25%) | |
| 9–16 | 7 (7%) | |
| Diagnoses: physical health | ||
| Respiratory | 33 (33%) | |
| Wounds | 25 (25%) | |
| Blood borne viruses | 22 (22%) | |
| Chronic painful condition | 19 (19%) | |
| Skin condition | 19 (19%) | |
| Seizures | 18 (18%) | |
| Infection | 18 (18%) | |
| Fracture | 18 (18%) | |
| Dental condition | 16 (16%) | |
| Rheumatic | 15 (15%) | |
| Head injury | 14 (14%) | |
| Upper gastrointestinal | 13 (13%) | |
| Neurological | 12 (12%) | |
| Musculoskeletal | 12 (12%) | |
| Head/brain injury | 11 (11%) | |
| Epilepsy | 7 (7%) | |
| Anaemia | 7 (7%) | |
| Alcohol-related seizures | 7 (7%) | |
| Eye condition | 6 (6%) | |
| Endocrine condition | 5(5%) | |
| Hearing condition | 4 (4%) | |
| Genitourinary/Pelvic | 2 (2%) | |
| Cardiovascular disease | 2 (2%) | |
| Alcohol-related brain injury | 1 (1%) | |
| Coronary heart disease | 1 (1%) | |
| Other | 22 (22%) | |
| Allergies | 17 (18%) | |
| Currently pregnant, or ever been pregnant | 11 (11%) | |
| Participants with a mental health diagnosis | 76 (77%) | |
| Diagnosis: mental health | ||
| Depression | 56 (56%) | |
| Anxiety | 47 (47%) | |
| Self-harm/suicide attempt | 21 (21%) | |
| Post-traumatic stress disorder | 18 (18%) | |
| Personality disorder | 15 (15%) | |
| Schizophrenia/psychosis | 13 (13%) | |
| Adverse childhood experience | 11 (11%) | |
| Drug-induced psychosis | 6 (6%) | |
| Mania/hypomania | 5 (5%) | |
| Complex trauma | 6 (6%) | |
| Other mental health condition | 9 (9%) | |
| Ever been in jail or prison | 75 (76%) | 2 |
| Number of times in jail or prison | 8 [4–20] | 2 |
| Assaulted in past 6 months | 29 (30%) | 2 |
| Feel unsafe | 27 (27%) | 1 |
1More than one answer may be given
Table 3.
Baseline alcohol, tobacco, overdose and problem drug use (N% or mean (SD)/median [IQR]
| Characteristic | Participants (n = 99) | Missing data |
|---|---|---|
| Daily alcohol consumption | 18 (18%) | 2 |
| Ever had a detox/hospital stay/rehab for alcohol | 19 (20%) | 3 |
| Age of first drink | 14 [12–16] | |
| Alcohol units per week | 34.3 (89.3) | 13 |
| Previous DTs or withdrawals | 25 (26%) | 4 |
| Ever had alcohol-related counselling | 35 (38%) | 7 |
| Current tobacco smoker | 89 (91%) | 1 |
| Age started smoking | 14 [11–15] | 31 |
| Number of cigarettes per day | 8 [0–15] | 4 |
| Vaping | 32 (33%) | 3 |
| Current nicotine replacement therapy | 9 (10%) | 5 |
| Drug-related overdose in past three months | 15 (15%) | 2 |
| Number of overdoses in past three months | 0.2 (0.7) | 2 |
| Why participant overdosed (n = 15) | ||
| Took too much illicit drug | 4 (27%) | |
| Unknown | 3 (20%) | |
| Accidental | 2 (13%) | |
| Street valium | 1 (7%) | |
| Strength of drug taken | 1 (7%) | |
| Greed | 1 (7%) | |
| Guilt | 1 (7%) | |
| Money related | 1 (7%) | |
| Family related | 1 (7%) | |
| Possesses Naloxone | 50 (56%) | 10 |
| Does participant have any injection sites (yes) | 37 (38%) | 2 |
| Heroin: currently using | 37 (37%) | |
| Age first used heroin (years) | 23.1 (6.0) | |
| Heroin: frequency of use | 1 | |
| Once or more per day | 22 (59%) | |
| Every few days/weekly | 6 (16%) | |
| Every two weeks/monthly | 7 (19%) | |
| Rarely | 1 (3%) | |
| Dose of heroin used | 2 | |
| ≤ 0.4 g (≤£10) | 2 (5%) | |
| > 0.4 g but ≤ 2 g (£11–50) | 7 (19%) | |
| > 2 g but ≤ 4 g (£51–100) | 19 (51%) | |
| Refused to answer | 1 (1%) | |
| Heroin: route of administration | ||
| Injection | 25 (68%) | |
| Smoke | 12 (32%) | |
| Cocaine: currently using | 50 (51%) | |
| Age first used cocaine (years) | 24.2 (8.4) | 2 |
| Cocaine: frequency of use | 1 | |
| Once or more per daily/most days | 0 | |
| Every few days/weekly | 22 (44%) | |
| Every two weeks/monthly | 9 (18%) | |
| Rarely | 3 (6%) | |
| Refused to answer | 1 (2%) | |
| Dose of cocaine used | ||
| ≤ 1 bag (o.4 g, 2 lines, ≤£10) | 8 (16%) | |
| > 1 bag – 2 bags (310 − 20) | 14 (28%) | |
| > 2 bags – 1 g (£21–25) | 4 (8%) | |
| > 1 g (2.5 bags, >£25) | 23 (46%) | |
| Refused to answer | 1 (2%) | |
| Cocaine: route of administration | ||
| Injection | 29 (58%) | |
| Smoke | 18 (36%) | |
| Intranasal | 3 (6%) | |
| Street Valium: currently using | 14 (14%) | 1 |
| Age first used street Valium | 26.0 (12.0) | 2 |
| Street valium: frequency of use | ||
| Once or more daily/most days | 3 (12%) | |
| Every few days/weekly | 7 (50%) | |
| Every two weeks/monthly | 1 (7%) | |
| Rarely | 3 (12%) | |
| Dose of street valium used | ||
| 1–10 tablets | 7 (50%) | |
| 11–25 tablets | 2 (14%) | |
| 26–50 tablets | 3 (21%) | |
| 51–100 tablets | 2 (14%) | |
| Spice: currently using | 8 (8%) | 2 |
| Age first used spice | 37.4 (10.3) | 1 |
| Spice: frequency of use | 1 | |
| Once or more daily/most days | 6 (75%) | |
| Every few days/weekly | 1 (13%) | |
| Every two weeks/monthly | 1 (13%) | |
| Rarely | 0 | |
| Street gabapentinoids: currently using | 10 (10%) | 1 |
| Age first used street gabapentinoids | 37.6 (10.2) | |
| Street gabapentinoids: frequency of use | ||
| Once or more daily/most days | 5 (50%) | |
| Every few days/weekly | 2 (20%) | |
| Every two weeks/monthly | 1 (10%) | |
| Rarely | 1 (10%) | |
| Refused to answer | 1 (10%) | |
| Cannabis: currently using | 31 (31%) | |
| Age first used cannabis | 17.0 (9.2) | 3 |
| Cannabis: frequency of use | ||
| Once or more daily/most days | 16 (52%) | |
| Every few days/weekly | 8 (26%) | |
| Every two weeks/monthly | 1 (3%) | |
| Rarely | 5 (16%) | |
| Refused to answer | 1 (3%) | |
Table 4.
Prescribed medicines (N% or mean (SD)/median [IQR]
| Characteristics | Participants (n = 99) | Missing data |
|---|---|---|
| Opioid substitution treatment (OST) | 74 (75%) | |
| Methadone | 65 (88%) | |
| Methadone: daily dose (mg) | 76.6 (28.1) | |
| Methadone range of dose (mg) | 30–150 | |
| Buprenorphine oral/orodispersible/sublingual/with naloxone (1 miss) | 5 (7%) | |
| Buprenorphine: daily dose (mg) | 14.8 (1.8) | |
| Buprenorphine range of dose (mg) | 12–16 | |
| Buprenorphine injection (every four weeks) | 4 (5%) | |
| Buprenorphine injection monthly dose (mg) | 120 (16) | |
| Buprenorphine injection range of dose (mg) | 96–128 | |
| Diazepam treatment | 17 (17%) | 1 |
| Diazepam: daily dose (mg) | 26.7 (7.4) | |
| Diazepam range of daily dose (mg) | 10–40 | |
| Number of medicines for problem drug use | 1 [1–2] | 25 |
| Medicine for physical health | 43 (44%) | 1 |
| Analgesia | 14 (33%) | |
| Respiratory | 12 (28%) | |
| Nutrition and anaemia | 9 (21%) | |
| Antiepileptic | 7 (16%) | |
| Upper gastrointestinal | 7 (16%) | |
| Antihypertensive | 5 (12%) | |
| Topical for skin condition | 4 (9%) | |
| Antiplatelet | 4 (9%) | |
| Diabetes | 2 (5%) | |
| Antiretroviral | 2 (5%) | |
| Laxative | 2 (5%) | |
| Antibacterial/antifungal | 1 (2%) | |
| Statin | 1 (2%) | |
| Hormone replacement therapy | 1 (2%) | |
| Drug for movement disorder | 1 (2%) | |
| Number of medicines for health condition | 2 [1–3] | 3 |
| Medicine for mental health | 37 (38%) | 1 |
| Antipsychotic | 12 (32%) | |
| Antidepressant | 29 (78%) | |
| Anxiolytic | 3 (8%) | |
| Number of medicines for mental health | 1 [1–2] | 1 |
We used various validated instruments to measure health-related quality of life (EQ-5D-5 L) [25], dyspnoea (MRC) [26] and COPD assessment test [27]. We also carried out a number of near patient tests, including weight and height to calculate body mass index, blood pressure, heart rate, oxygen saturation and strength via a hand-held dynamometer. Fried’s adapted frailty phenotype scores were calculated using data on participant’s weight loss in the past three months, experience of exhaustion in the last two weeks, physical activities undertaken in the past 4 weeks, walk pace in the last three months, grip strength (as recorded with a hand-held dynamometer), BMI and gender [28]. Frailty status was reported as follows: (a) non-frail (0 frailty indicators); (b) pre-frail (1–2 frailty indicators); and (c) frail (≥ 3 frailty indicators).
Public and participant involvement and engagement (PPIE)
Our approach to PPIE have been reported separately [29]. In summary, we sought extensive engagement with people with lived and living experience of repeat and long-term homelessness. A lived experience panel in each study site met regularly to advise on the study as it progressed. Previous qualitative studies with PEH and various stakeholders, including service commissioners, healthcare professionals and public health organisations have informed perspectives in the design and implementation of this study.
Results
Recruitment
One hundred and eighty-four eligible participants were approached, with 84 declining participation, resulting in 100 participants recruited to the study as planned; 50 participants recruited to the Glasgow arm from 17th November 2022 to 9th December 2022 (23 days) and 50 participants to the Birmingham arm from 8th November 2022 to 16th March 2023 (129 days). Fifty participants in total were randomised into the active group (25 from Glasgow and 25 from Birmingham) and assigned to the PHOENIx intervention plus UC and 50 were randomised into UC only. One set of baseline research notes was lost. The records contained no identifiable personal data, but merely an identifier known only to the research team. Only 99 participants were hence included in the analysis.
Baseline demographic, physical and mental health characteristics (Table 2)
The mean age of participants was 42 (SD 8.9) years (males n = 81, 82%). The median age participants first experienced homelessness was 21.5 [IQR 16,33] years; 45, (45%) participants lived in supported accommodation and (n= 17, 17%) in unsupported hostels. Most participants were white and male; reflecting significant evidence gaps, not only in minoritised communities, but also studies involving women, lesbian, gay, bisexual, or transgender (LBGT+) people and people with disabilities [30]. Most participants were registered with a GP (87, 88%), with only 33, (38%) of that total registered with a specialist homeless health GP. Ninety-five (96%) participants were receiving financial state benefit support. These results are fairly typical of previous similar studies [16]. Half of all participants (n = 49, 50%) walked to appointments, while 31, (31%) took the bus or another form of public transport.
Seventy-nine (81%) participants were known to addiction teams (Alcohol and Drug Recovery Services), 30 (30%) had mental health team support, and 43 (43%) a housing/caseworker. Sixty-one (62%) participants had two or more chronic physical health conditions, with respiratory (n = 33, 33%) the most frequently reported, followed by wounds (a combination of recent injuries and chronic leg ulcers) (n = 25, 25%), blood borne viruses (n = 22, 22%), and chronic painful conditions (n = 21, 21%). Seventy-six (77%) participants reported having a mental health condition or difficulty; depression (n = 56, 56%), anxiety (n = 47, 47%), self-harm/suicide attempt (n = 21, 21%), post-traumatic stress disorder (PTSD) (n = 18, 18%), personality disorder (n = 15, 15%), and adverse childhood event (n = 11, 11%) most commonly reported.
Seventy-four (76%) participants stated ever been in prison/jail, with a median of 8 [IQR 4,20] occasions. Prison is associated with repeat and continuous homelessness and study reflects this relationship. Criminal justice forms one of the three overlapping severe and multiple disadvantages (SMD) and is also associated with high levels of loneliness, isolation, unemployment, poverty and mental ill-health [6], making care and treatment of health and social problems more difficult and complex.
Baseline alcohol, tobacco, overdose, and substance use (Table 3)
Eighteen (18%) participants reported daily alcohol consumption. Median age of first alcohol, use was 14 [IQR 12, 16] years old, the mean reported alcohol units per week was 34.3 (SD 89.3), and equivalent to around half a bottle of wine per day. Of those consuming alcohol, 25 (26%) had previous delirium tremens (DTs) or experienced alcohol withdrawals, reflecting high levels of past alcohol consumption. Eighty-nine (91%) reported smoking tobacco, and (n = 9, 10%) stated current nicotine replacement therapy.
Fifteen (15%) participants reported a recent substance-related overdose with the following reasons reported: took too much (n = 4, 27%), unknown (n = 3, 20%), accidental (n = 2, 13%), street benzodiazepines (“Valium”) (new psychoactive substance) (n = 1, 7%), strength of drug (n = 1, 7%), ‘greed’ (n = 1, 7%), ‘guilt’ (n = 1, 7%), money-related (n = 1, 7%) and family-related (n = 1, 7%). Thirty-seven (37%) participants reported currently using heroin with 25 (68%) currently injecting. Fifty (51%) participants stated currently using cocaine with 29 (58%) injecting. Fourteen (14%) participants stated current street benzodiazepines use. Eight (8%) participants reported taking the synthetic cannabinoid spice. Ten (10%) reported currently using street gabapentinoids. Thirty-one (31%) participants reported currently using cannabis.
Prescribed medicines (Table 4)
Seventy-four (75%) participants were prescribed OST. Of these participants, (n = 65, 88%) were prescribed methadone. Oral buprenorphine was prescribed to five (7%) participants, and injectable buprenorphine was prescribed to four (5%). Diazepam was prescribed to 17 (17%) participants. Three-quarters of participants in our study reported being prescribed OST with almost two-thirds of all participants in the study receiving a prescription for methadone. A much smaller number of participants were prescribed buprenorphine. Street benzodiazepines such as etizolam have increasingly been implicated in substance-related deaths, and diazepam can be prescribed to manage this.
Thirty-seven (38%) participants reported being prescribed medicines for mental health, with (n = 29, 78%) prescribed antidepressants, (n = 12, 32%) prescribed antipsychotic medications, and (n = 3, 8%) anxiolytic medications. Forty-three (44%) participants were prescribed medicines for physical health, with 14 (33%) participants prescribed analgesics, (n = 12, 28%) medicines for respiratory problems, and (n = 9, 21%) nutrition and anaemia medicines. This wide range of medicines reflect the numbers of multi-morbidities experienced by PEH [1], although less than half of participants in our study received a prescription for physical health problems. This suggests a disparity between physical health problems and medicine prescribing as described previously and addressed by PHOENIx [15, 16].
Baseline functional assessment, quality of life (Table 5)
Table 5.
Baseline functional, quality of life (N% or mean (SD)/median [IQR]
| Characteristics | Participants (n = 99) | Missing data |
|---|---|---|
| MRC Dyspnoea Scale | 96 (97%) | 3 |
| Not troubled by breathless except on strenuous exercise | 31 (32%) | |
| Short of breath when hurrying on a level or when walking up a slight hill | 15 (16%) | |
| Walks slower than most people on the level, stops after a mile or so, or stops after 15 min walking at own pace | 18 (19%) | |
| Stops for breath after walking 100 yards, or after a few minutes on level ground | 17 (18%) | |
| Too breathless to leave the house, or breathless when dressing/undressing | 15 (16%) | |
| COPD Assessment Test | 13.0 (6.5,22.5) | 3 |
| COPD Assessment Test (range) | (0–40.0) | |
| Quality of life (EQ-5D-5 L) | ||
| Mobility | 1.5 [1,3.25] | |
| (1 = no problem; 2 = slight; 3 = moderate; 4 = severe; 5 = unable to mobilise) | ||
| Self-care | 1 [1,3] | |
| (1 = no problem; 2 = slight; 3 = moderate; 4 = severe; 5 = unable to self-care) | ||
| Usual activities | 2 (1,3) | |
| (1 = no problem; 2 = slight; 3 = moderate; 4 = severe; 5 = unable to do usual activities) | ||
| Pain/discomfort | 3 [1,4] | |
| (1 = no problem; 2 = slight; 3 = moderate; 4 = severe; 5 = extreme pain/discomfort) | ||
| Anxiety/depression | 3 [3,5] | |
| (1 = no problem; 2 = slight; 3 = moderate; 4 = severe; 5 = extreme) | ||
| Visual analogue scale (EQ-VAS) (5 miss) | 42.2 (24.7) | |
| (0 = worst health imaginable; 100 = best health imaginable) | ||
| EQ-5D index score (van Hout method used) | 0.431 (0.397) | 3 |
| Participants with a negative (worse than death) score | 16 (17%) | 3 |
| (− 0.5 = lowest score on all five domains; 1 = highest score on all five domains) | ||
| Fried’s frailty phenotype | 22 | |
| Non-frail | 0 (-) | |
| Pre-frail | 34 (44%) | |
| Frail | 43 (56%) | |
| BMI (mg/kg2) | 23.3 (4.6) | 16 |
| BMI (minimum – maximum) | 14.2–41.8 | |
| Overweight (> 25 kg/m2) | 25 (28%) | 11 |
| Underweight (< 18.5 kg/m2) | 12 (14%) | 13 |
| Blood pressure: systolic (mmHg) | 122 (17.5) | 11 |
| Blood pressure: diastolic (mmHg) | 78.4 (13) | 11 |
| Heart rate (BPM) | 76.1 (12.5) | 5 |
| Oxygen saturation (%) | 96.5 (2.2) | 4 |
| Grip strength (kg) | 30.8 (10) | 9 |
| Peak expiratory flow rate (PEFR) | 385.8 (149.6) | 7 |
| How many meals ate in a typical day | 1 | |
| Breakfast, lunch and dinner | 10 (10%) | |
| One meal only per day | 31 (31%) | |
| Two meals per day | 32 (32%) | |
| No daily meals | 23 (23%) | |
| Refused to answer | 1 (1%) | |
| Blood samples (urea and electrolytes) | ||
| Sodium (within normal range) | 2 | |
| Yes | 31 (32%) | |
| No | 0 | |
| Not collected | 66 (68%) | |
| Potassium (within normal range) | 2 | |
| Yes | 30 (31%) | |
| No | 1 (1%) | |
| Not collected | 66 (68%) | |
| Chloride (within normal range) | 2 | |
| Yes | 27 (28%) | |
| No | 2 (2%) | |
| Not collected | 68 (70%) | |
| Creatinine (within normal range) | 2 | |
| Yes | 29 (30%) | |
| No | 1 (1%) | |
| Not collected | 67 (69%) | |
| Estimated glomerular filtration rate (eGFR) (within normal range) | 3 | |
| Yes | 31 (32%) | |
| No | 1 (1%) | |
| Not collected | 64 (67%) | |
| Blood samples (liver function tests) | ||
| Alanine transaminase (ALT) (within normal range) | 1 | |
| Yes | 27 (28%) | |
| No | 4 (4%) | |
| Not collected | 67 (68%) | |
| Aspartate transferase (AST) (within normal range) | 1 | |
| Yes | 22 (22%) | |
| No | 8 (8%) | |
| Not collected | 68 (69%) | |
| Alkaline phosphatase (ALP) (within normal range) | 1 | |
| Yes | 26 (27%) | |
| No | 4 (4%) | |
| Not collected | 68 (69%) | |
| Albumin (within normal range) | 1 | |
| Yes | 23 (23%) | |
| No | 7 (7%) | |
| Not collected | 68 (69%) | |
| Bloods samples (bone profile) | ||
| Calcium (within normal range) | 1 | |
| Yes | 11 (11%) | |
| No | 2 (2%) | |
| Not collected | 85 (86%) | |
| Calcium adjusted (within normal range) | 2 | |
| Yes | 13 (13%) | |
| No | 0 | |
| Not collected | 84 (87%) | |
| Phosphate (within normal range) | 1 | |
| Yes | 12 (12%) | |
| No | 0 | |
| Not collected | 86 (88%) | |
| C-reactive protein (within normal range) | 1 | |
| Yes | 8 (8%) | |
| No | 6 (6%) | |
| Not collected | 84 (86%) | |
| Vitamin B12 (within normal range) | 1 | |
| Yes | 5 (5%) | |
| No | 3 (3%) | |
| Not collected | 90 (92%) | |
| Full blood count (within normal range) | 1 | |
| Yes | 6 (6%) | |
| No | 25 (26%) | |
| Not collected | 67 (68%) | |
| Magnesium (within normal range) | 1 | |
| Yes | 5 (5%) | |
| No | 1 (1%) | |
| Not collected | 92 (94%) | |
| Folate (within normal range) | 1 | |
| Yes | 3 (3%) | |
| No | 6 (6%) | |
| Not collected | 89 (91%) | |
| Cholesterol (within normal range) | 1 | |
| Yes | 5 (5%) | |
| No | 1 (1%) | |
| Not collected | 92 (94%) | |
| Blood samples (blood borne viruses) | ||
| Hepatitis C virus | 1 | |
| Positive | 7 (7%) | |
| Negative | 21 (21%) | |
| Not collected | 69 (70%) | |
| Human immunodeficiency virus | 1 | |
| Positive | 2 (2%) | |
| Negative | 23 (23%) | |
| Not collected | 73 (74%) | |
| Hepatitis B virus | 1 | |
| Positive | 0 | |
| Negative | 21 (21%) | |
| Not collected | 77 (78%) | |
Severity of breathlessness was rated on a scale of 0 (less severe) to 4 (more severe). Fifteen (15%) participants rated severity of breathlessness as 4, and 17 (17%) rated it as 3. Participants’ COPD Assessment Test (CAT) scores ranged from 0 the minimum score to 40 the maximum one, with a median score of 13.0 [IQR 6.5, 22.5]. Poor respiratory health in PEH in determined often by high levels of smoking, substance use in the population and living in communal shelters for PEH, making the spread of respiratory infections, such as tuberculosis easier. In our study one-third of patients reported that they were either too breathless to leave the house/breathless when dressing/undressing, or stopped for breath after walking 100 yards, or after a few minutes on level ground.
Thirty-four (44%) participants were deemed pre-frail and 43 (56%) as frail. Increased frailty is associated with less resilience to health encounters [31] and presents earlier and at higher rates in PEH than in the general population. Only five participants presented as non-frail, demonstrating extremely high levels of frailty for the majority of participants, and consistent with elderly populations.
Participants had an average EQ-5D-5L health-related quality of life score of 42.2 (SD 24.7). EQ-5D-5L index score was 0.43 (SD 0.40), using the original cross-walk method [32], and was similar 0.43 (SD 0.38) when checked against the method of the EEPRU project team (as described in the NICE Reference Case 2022) [33], and compared to a general population index of 0.86 (SD 0.23). Health-related quality of life is a key outcome in research with PEH [34]. Sixteen participants had a value that represents a health state “worse than death”. The health-related quality of life Visual Analogue Scale (VAS) represents a visual self-report tool ranging from the best health imaginable (scoring 100) to the worst health imaginable (score of 0) [32]. The VAS of the EQ-5D-5L at baseline in our previous study was 42.2, compared to an average of 34.4 in PEH with a recent substance-related overdose [16].
Mean body mass index (BMI; mg/kg2) was 23.3 (SD 4.6). According to BMI, (n = 25, 28%) participants were considered overweight and (n = 12, 14%) underweight. Mean health measures were: systolic blood pressure 122 mmHg (SD 17.5); diastolic blood pressure 78.4 mmHg (SD 13); heart rate 76.1 BPM (SD 12.5); oxygen saturation 96.5% (SD 2.2); grip strength 30.8 kg (SD 10).
Participant blood test results were investigated from existing healthcare records with a significant minority of liver function tests out with normal range. A minority of participants were positive for hepatitis C virus (HCV) (n = 7, 7%), and human immunodeficiency virus (HIV) (n = 2, 2%). We do not have data on participants whose HCV cleared spontaneously or after treatment.
Table 6 describes participant encounters with services in the year prior to baseline data collection. Thirty-nine (85%) of those participants who collected medicines, did so daily from pharmacies. Twenty-five (26%) participants had a recent ED attendance. Nine (9%) participants received an inpatient hospital stay with intoxication given as the main reason for this. Eighteen participants (18%) attended an outpatient appointment, of which the most common reason related to infectious diseases (n = 12, 12%).
Table 6.
Encounters (N% or mean (SD)/median [IQR]
| Characteristics | Participants (n = 99) | Missing data |
|---|---|---|
| Criminal justice in last 3 months | ||
| Arrested booked or charged for breaking the law | 22 (23%) | 4 |
| Been convicted or pleaded guilty to any charges | 12 (13%) | 4 |
| Been under any form of criminal justice supervision, including on probation, in jail, or in prison | 22 (23%) | 3 |
| Currently on probation | 18 (18%) | 4 |
| How long spent in prison in last 3 months | 29.6 (33.2) | 86 |
| Collected medicines in past three months | ||
| Collected 80% of daily dose medicines from pharmacy | 39 (85%) | 53 |
| Participants with Emergency Department (ED) visit | 25 (26%) | 3 |
| Top 5 reasons for ED visit | ||
| Overdose | 5 (5%) | |
| Assault | 3 (3%) | |
| Self-referral | 3 (3%) | |
| Intoxicated | 3 (3%) | |
| Mental health | 2 (2%) | |
| Participants with hospital inpatient stays | 9 (9%) | 4 |
| Top 5 reasons for hospital inpatient stay | ||
| Intoxicated | 4 (4%) | |
| Sepsis | 2 (2%) | |
| Gave birth | 1 (1%) | |
| Infectious disease | 1 (1%) | |
| Respiratory | 1 (1%) | |
| Participants with outpatient visits | 18 (18%) | 3 |
| Departments participant visited | ||
| Infectious disease | 12 (12%) | |
| Gastroenterology | 3 (3%) | |
| Chiropody | 2 (2%) | |
| Ophthalmology | 2 (2%) | |
| Orthopaedics | 2 (2%) | |
| Housing contacts/visits | 9 (9%) | 3 |
| Third sector contacts/visits | 44 (52%) | 14 |
| Rehab contacts/visits | 7 (7%) | 3 |
Discussion
Previous trials with PEH have focused on housing interventions, such as Housing First (HF). These studies were mostly based in North America with very few originating from the UK (Supplementary material 1.). To our knowledge no previous studies utilised community pharmacies as the primary setting for recruitment of PEH into RCTs. Community pharmacies are frequently utilised by many PEH for collection of prescribed medicines, treatment of substance use disorders, needle exchange and other public health services. Yet there have been little efforts to utilise these contacts to offer holistic support. Our study aimed to address this gap.
We found that participants were mostly men, white, had an average age of 42 years. Median age first experienced homelessness was 22, and over half had slept rough on the streets in the three months prior to recruitment to the study. Most participants resided in large hostels or hostels, often unsupported, and environments that are often re-traumatising [9]. Most participants had a chronic health condition, with most having two or more conditions. Most had a mental health diagnosis, were known to substance treatment services, in receipt of state benefits, and had been in prison during their lifetime.
In this study, half of the study participants walked to appointments and less than a third took the bus or another form of public transport. Lack of access to bus pass/free public transport is an important barrier to attending care appointments, skills and employment training venues, thereby exacerbating disparities. In our previous qualitative studies, participants have described lack of access to free bus pass prevented timely collection of their prescribed medicines as they were physically too unwell to walk to the pharmacy themselves [35–37]. These findings emphasise the need to address wider determinants of health in intervention design and delivery.
We are aware that demographic characteristics of participants in our study is not representative of Glasgow and Birmingham populations. For example, 57.9% of populations in Birmingham and 85.7% of Glasgow populations respectively identify their ethnicity as ‘White’ compared to 80.0% and 94.0% respectively in our study sample [38, 39]. It is likely due to under-utilisation of community pharmacy by minority ethnic and women experiencing homelessness. This highlights the need to encourage more minority ethnic and women experiencing homelessness including people with insecure immigration status to utilise pharmacy as an accessible service in the community.
The majority of published research does not reflect the ethnic diversity of UK PEH population [40] and tend to under-represent the perspectives of women experiencing homelessness [41, 42]. Future definitive trials where possible should attempt to stratify sampling by individual characteristics (in individually randomised RCTs) or use minimisation based on characteristics of study setting (e.g. shelters exclusively housing women).
A number of outcomes from the baseline characteristics of participants do present as unique. For example, we discovered that even though the majority of participants reported having a chronic physical health condition, and a mental health diagnosis, half of all participants walked to appointments, with under a third having transport paid for them. These criteria are important to highlight, not just in terms of health-related quality of life, but also to address negative consequences surrounding missed appointments, or “missingness” such as poor health outcomes [43].
In this study, we found nine participants with a confirmed blood borne virus diagnosis. Of these the two participants with a HIV diagnosis were prescribed antiretroviral treatments. However, we could only identify one out of the seven participants on anti-HCV treatment. It is likely that our reliance mainly on primary care records to extract this data may have led to under-recording as anti-HCV treatments are prescribed through specialist blood borne virus clinics and in Glasgow. Anti-HCV treatments may be also be dispensed through hospital pharmacies. This may also suggest missed repeat appointments. For example, participants may present to infectious diseases treatment once for diagnosis, and may not return for treatment. Missingness is common in healthcare in inclusion health and makes a compelling case for a referral pathway from PHOENIx to infectious diseases treatment providers, with support for transport costs, workers accompanying patients to appointments, and community prescribing by PHOENIx pharmacists of antiretroviral medications. Missed appointments are correlated with a lack of inexpensive, reliable transport [44], providing a further rationale as to why so many people in our study have untreated health problems.
It was unanticipated to learn that despite all participants experiencing homelessness, just under a half were known to housing/caseworker services. These findings reinforce the need for all outreach teams to create a rapid referral pathway to housing services.
Evaluation of data from the three and six-month participant follow up will be presented in future publications which will enable us to describe any signal of improvements across various baseline measures of health and social outcomes presented in this publication. Process evaluation will involve qualitative interviews with participants from both intervention and UC, intervention staff and wider stakeholders. Results from process evaluation are separately reported identifying how intervention work for different participants and the enablers and barriers relevant to wider adoption in community pharmacies [45].
Conclusion
The baseline characteristics of PEH reported in this study demonstrates multiple, complex and unmet health, social and practical needs of PEH who participated in the PHOENIx Community Pharmacy RCT. Data on participant health, and substance and alcohol use reinforces the finding that PEH have severe and intersecting inequalities characterised by substance use, and a high prevalence of acute and long term physical and mental health problems and multimorbidity. These baseline characteristics build on research and make a compelling case for an evidence-based response to the wider health, social and practical issues facing PEH. There is a need to engage, support, refer and treat PEH through integrated and bespoke pathways to promote people’s engagement with care, prevent repeat homelessness and early deaths.
Supplementary Information
Acknowledgements
Angela Mackie, Nancy’s Chemist, Glasgow UK; Asgher Mohammed, Abbey Pharmacy, Glasgow UK; Hanif Rahman, Medisina, Birmingham UK; Mohammed Eshak, Ladywood Late Night Pharmacy, Birmingham UK; Ama Inechi, Attwood Green Pharmacy, Birmingham UK.
Abbreviations
- CRF
Case Report Form
- ETHOS
European Typology on Homelessness and Housing Exclusion
- IQR
Inter Quartile Range
- NICE
National Institute for Health and Care Excellence
- OST
opioid substitution therapy
- PEH
People experiencing homelessness
- SD
Standard Deviation
- SMD
Severe and Multiple Disadvantage
- UC
Usual Care
- UK
United Kingdom
Authors’ contributions
Vibhu Paudyal and Richard Lowrie were the chief investigators on the study and jointly wrote the protocol and obtained funding. Andrew McPherson, Helena Heath, Jane Moir and Shabana Akhtar were study researchers who collected the data. Natalie Allen, Nigel Barnes, Hugh Hill, Steven Ross and George Provan contributed to trial protocol and facilitated study sites and participant recruitment for data collection. Andrea Williamson, Nigel Barnes, Parbir Jagpal, Lee Middleton, Jennifer Hislop, Versha Cheed and Frances S. Mair contributed to funding acquisition as co-applicants and supported protocol development, study execution and reviewed the manuscript. Sarah Tearne facilitated trial data collection and analysis. Lee Middleton and Versha Cheed were study statisticians. Jennifer Hislop is a health economist. Andrew McPherson led the writing of the manuscript which was first reviewed by Vibhu Paudyal and Richard Lowrie. All authors contributed to the manuscript through review and editing. All authors agree to the final version of the manuscript.
Funding
This study was funded by the National Institute of Health and Care Research (NIHR) Health Services and Delivery Research scheme under commissioned call stream ‘20/56 Community Pharmacies’. [Grant award ID: NIHR133060]. Trial Sponsor: University of Birmingham, Birmingham, B15 2TT, United Kingdom. Views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care. Neither the study funder nor the sponsor had any role in the study design; collection, management, analysis, and interpretation of data; writing of the report; and the decision to submit the report for publication.
Data availability
Only scientifically sound proposals from appropriately qualified Research Groups will be considered for data sharing. Requests should be sent to bctu@bham.ac.uk. The request will be reviewed by the BCTU Data Sharing Committee in discussion with the Chief Investigators. Completion of a data sharing agreement will be required prior to the access.
Declarations
Competing interests
The authors declare no competing interests.
Ethics approval and consent for publication
The study was approved by East Midlands – Leicester South Research Ethics Committee. REC reference 22/EM/0119. Signed, informed consent was obtained from all participants prior to enrolment into the study.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Vibhu Paudyal, Email: vibhu.paudyal@kcl.ac.uk.
Richard Lowrie, Email: richard.lowrie@ed.ac.uk.
References
- 1.Aldridge RW, Story A, Hwang SW, Nordentoft M, Luchesnski SA, Hartwell G. Morbidity and mortality in homeless individuals, prisoners, sex workers, and individuals with substance use disorders in high-income countries: a systematic review and meta-analysis. Lancet. 2018;391(10117):241–50. 10.1016/S0140-6736(17)31869-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Office for National Statistics. Deaths of homeless people in England and Wales: 2021 registrations. https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/bulletins/deathsofhomelesspeopleinenglandandwales/2021registrations#:~:text=Numbers%20of%20homeless%20deaths,of%20658%20to%20824%20deaths. Accessed 10 Aug 2025.
- 3.National Health Care for the Homeless Council. Mortality and homelessness. https://nhchc.org/wp-content/uploads/2024/11/mortality-and-homelessness.pdf. Accessed 10 Aug 2025.
- 4.Vohra N, Paudyal V, Price MJ. Homelessness and the use of emergency department as a source of healthcare: a systematic review. Int J Emerg Med. 2022;15:32. 10.1186/s12245-022-00435-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.OECD. Global Homeless Population – OECD Report 2020. https://www.oecd.org/content/dam/oecd/en/data/datasets/affordable-housing-database/hc3-1-homeless-population.pdf. Accessed 10 Aug 2025.
- 6.Bramley G, Fitzpatrick S, Edwards J, Ford D, Johnsen S, Sosenko F, Watkins D. (2015) Hard Edges: Mapping severe and multiple disadvantage. England, Lankelly Chase. Available: https://lankellychase.org.uk/wp-content/uploads/2015/07/Hard-Edges-Mapping-SMD-2015.pdf. Accessed 10 Aug 2025.
- 7.Bramley G, Fitzpatrick S. Homelessness in the UK: who is most at risk? Hous Stud. 2018;33(1):96–116. 10.1080/02673037.2017.1344957. [Google Scholar]
- 8.Liu M, Luong L, Lachaud J, Edalati H, Reeves A, Hwang SW. Adverse childhood experiences and related outcomes among adults experiencing homelessness: a systematic review and meta-analysis. Lancet. 2021;6(11):E836–47. 10.1016/S2468-2667(21)00189-4. [DOI] [PubMed] [Google Scholar]
- 9.Farmer N, McPherson A, Thomson J, Reilly F, Williamson A, Lowrie R. (2023) ‘There’s No Hope for Any Kind of Decent Life’: A Qualitative Study to Explore the Perspectives of People Experiencing Homelessness with a Recent Non-Fatal Overdose in Scotland. British Journal of Social Work. 2024;54(2):548–67. 10.1093/bjsw/bcad160. [Google Scholar]
- 10.McNeill S, O’Donovan D, Hart N. Access to healthcare for people experiencing homelessness in the UK and ireland: a scoping review. BMC Health Serv Res. 2022;22:910. 10.1186/s12913-022-08265-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.NICE. Integrated health and social care for people experiencing homelessness. [NG214]. https://www.nice.org.uk/guidance/ng214. Accessed 10 Aug 2025. [PubMed]
- 12.Radley A, Melville K, Easton P, Williamson B, Dillon JF. Standing outside the junkie Door’ – service users’ experience of using community pharmacies to access treatment for opioid dependency. J Public Health. 2017;39(4):846–55. 10.1093/pubmed/fdw138. [DOI] [PubMed] [Google Scholar]
- 13.Jessop AB, Luu J, Holbert R, Mazzella S. Delivering care to a city’s most vulnerable residents: A harm reduction organization and community pharmacy working together to reduce health disparities. J Health Care Poor Underserved. 2022;33(5):202–8. 10.1353/hpu.2022.0169. [DOI] [PubMed] [Google Scholar]
- 14.Paudyal V, Gibson Smith K, MacLure K, Forbes-McKay K, Radley A, Stewart D. Perceived roles and barriers in caring for the people who are homeless: a survey of UK community pharmacists. Int J Clin Pharm. 2019;15(41):215–27. 10.1007/s11096-019-00789-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lowrie R, Stock K, Lucey S, Knapp M, Williamson A, Montgomery M, Lombard C, Maguire D, Allan R, Blair R, Paudyal V, Mair FS. (2021) Pharmacist led homeless outreach engagement and non-medical independent prescribing (Rx) (PHOENIx) intervention for people experiencing homelessness: a non- randomised feasibility study. Int J Equity Health. 2021;20(1):19. 10.1186/s12939-020-01337-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Lowrie R, McPherson A, Mair FS, Stock K, Maguire D, Paudyal V, Duncan C, Blair R, Lombard C, Ross S, Hughes F, Moir J, Scott A, Reilly F, Sills L, Hislop J, Wishart S, Brannan D, Robertson JR, Ramage R, Boyle A, Greenlaw N, Williamson AE. Holistic health and social care outreach for people experiencing homelessness with recent non-fatal overdose in glasgow, scotland: the pharmacist and third sector homeless charity worker outreach engagement Non-medical independent prescriber Rx (PHOENIx) pilot randomised controlled trial. BMJ Public Health. 2024;2:e000219. 10.1136/bmjph-2023-000219. [Google Scholar]
- 17.Paudyal V, Lowrie R, Mair FS, et al. Protocol for a pilot randomised controlled trial to evaluate integrated support from pharmacist independent prescriber and third sector worker for people experiencing homelessness: the PHOENIx community pharmacy study. Pilot Feasibility Stud. 2023;9:29. 10.1186/s40814-023-01261-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.FEANTSA. ETHOS Typology on Homelessness and Housing Exclusion. https://www.feantsa.org/en/toolkit/2005/04/01/ethos-typology-on-homelessness-and-housing-exclusion. Accessed 10 Aug 2025.
- 19.Walsh D, McCartney G, Collins C, UClbut M, Batty GD. History, politics and vulnerability: explaining excess mortality in Scotland and Glasgow. Public Health. 2017;151:1–12. 10.1016/j.puhe.2017.05.016. [DOI] [PubMed] [Google Scholar]
- 20.Kierans D. Where do migrants live in the UK? The Migration Observatory at the University of Oxford. 2nd edition: COMPAS. https://migrationobservatory.ox.ac.uk/resources/briefings/where-do-migrants-live-in-the-uk/. Accessed 10 Aug 2025.
- 21.Catney G, Lloyd CD, Ellis M, Wright R, Finney N, Jivraj S, Manley D. Ethnic diversification and neighbourhood mixing: A rapid response analysis of the 2021 census of England and Wales. Geographical J. 2023;189:63–77. 10.1111/geoj.12507. [Google Scholar]
- 22.Local Government Association. Birmingham City Council: Year 1 update. https://www.local.gov.uk/birmingham-city-council-year-1-update Accessed 10 Aug 2025.
- 23.Public Health England. (2018) Birmingham: Local authority. Local Authority Health Profile 2018. https://www.nice.org.uk/Media/Default/sharedlearning/PHE%20data%20Birmingham.pdf. Accessed 10/August/2025.
- 24.Bowen M, Marwick S, Marshall T, Saunders K, Burwood S, Yahyouche A, Stewart D, Paudyal V. Multimorbidity and emergency department visits by a homeless population: a database study in specialist general practice. Br J Gen Pract. 2019;69(685):e515–25. 10.3399/bjgp19X704609. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Herdman M, Gudex C, Lloyd A, Janssen MF, Kind P, Parkin D, Bonsel G, Badia X. Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L). Qual Life Res. 2011;20:1727–36. 10.1007/s11136-011-9903-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Bestall JC, Paul EA, Garrod R, Garnham R, Jones PW, Wedzicha J. Usefulness of the medical research Council (MRC) dyspnoea scale as a measure of disability in patients with chronic obstructive pulmonary disease. Thorax. 1999;54(7):581–6. 10.1136/thx.54.7.581. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Jones PW, Harding G, Berry P, Wiklund I, Chen WH, Leidy NK. Development and first validation of the COPD assessment test. Eur Respir J. 2009;34(3):648–54. 10.1183/09031936.00102509. [DOI] [PubMed] [Google Scholar]
- 28.Fried LP, Tangen CM, Walston J, Newman AB, Hirsch C, Gottdiener J, Seeman T, Tracy R, Kop WJ, Burke G, McBurnie MA, Cardiovascular Health Study Collaborative Research Group. Frailty in older adults: evidence for a phenotype. J Gerontol Biol Sci Med Sci. 2001;56(3):M146–56. 10.1093/gerona/56.3.m146. [DOI] [PubMed] [Google Scholar]
- 29.McPherson A, Paudyal V, Lowrie R, Heath H, Moir J, Allen N, Barnes N, Hill H, Araf A, Lombard C, Ross S. Patient and public involvement in research evaluating integrated care for people experiencing homelessness: findings from the PHOENIx community pharmacy pilot Randomised-Controlled trial. Health Expect. 2024;27(5):e70070. 10.1111/hex.70070. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.McCarthy L, Parr S, Green S, Reeve K. Understanding models of support for people facing multiple disadvantage: A literature review Centre for Regional and Economic and Social research. Available: 2025 https://groundswell.org.uk/wp-content/uploads/2020/10/FLLSL-Lit-Review_FINAL-September-2020.pdf. Accessed 10/August/2025.
- 31.Dawes J, Walters K, Frost R, et al. Prevalence, risks and outcomes of frailty in people experiencing homelessness: a protocol for secondary analysis of health needs audit data. NIHR Open Res. 2024;4:9. 10.3310/nihropenres.13545.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Van Hout B, Janssen MF, Feng Y, et al. Interim scoring for the EQ-5D-5L: mapping the EQ-5D-5L to EQ-5D-3L value sets. Value Health. 2012;15:708–15. 10.1016/j.jval.2012.02.008. [DOI] [PubMed] [Google Scholar]
- 33.NICE. Health technology evaluations: the manual. NICE process and methods. https://www.nice.org.uk/process/pmg36. Accessed 10 Aug 2025.
- 34.Henwood BF, Matejkowski J, Stefancic A, Lukens JM. Quality of life after housing first for adults with serious mental illness who have experienced chronic homelessness. Psychiatry Res. 2014;220:549–55. 10.1016/j.psychres.2014.07.072. [DOI] [PubMed] [Google Scholar]
- 35.Gunner E, Chandan SK, Marwick S, Saunders K, Burwood S, Yahyouche A, Paudyal V. Provision and accessibility of primary healthcare services for people who are homeless: a qualitative study of patient perspectives in the UK. Br J Gen Pract. 2019;69(685):e526–36. 10.3399/bjgp19X704633. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Paudyal V, MacLure K, Forbes-McKay K, McKenzie M, MacLeod J, Smith A, Stewart D. If I die, I die, I don’t care about my health’: perspectives on self‐care of people experiencing homelessness. Health Soc Care Commun. 2020;28(1):160–72. 10.3399/bjgp19X704633. [DOI] [PubMed] [Google Scholar]
- 37.Jagpal P, Barnes N, Lowrie R, Banerjee A, Paudyal V. Clinical pharmacy intervention for persons experiencing homelessness: evaluation of patient perspectives in service design and development. Pharmacy. 2019;7(4):153. 10.3390/pharmacy7040153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Birmingham City Council. Population overview. Available: https://www.birmingham.gov.uk/info/50265/supporting_healthier_communities/2438/ethnic_communities/2: Accessed 10 Aug 2025.
- 39.Glasgow City Council. Demographics and needs profile. Available: https://www.glasgow.gov.uk/article/6088/Population-Projections-and-Population-Characteristics. Accessed 10/August/2025.
- 40.Finney N. (2022) Ethnic inequalities and homelessness in the UK. Centre for Homelessness Impact. Available: https://assets-global.website-files.com/59f07e67422cdf0001904c14/63e6476b4d417106be76eae5_CHI.Ethnic.Inequalities.homelessness.pdf. Accessed 10 Aug 2025.
- 41.Office of the National Statistics. Research on women experiencing hidden homelessness in the UK. Available at: https://www.ons.gov.uk/methodology/methodologicalpublications/generalmethodology/onsworkingpaperseries/researchonwomenexperiencinghiddenhomelessnessintheuk. Accessed Aug 12, 2025.
- 42.Lowrie R, Paudyal V, McPherson A, Heath H, Moir J, Allen N, Barnes N, Hill H, Araf A, Lombard C, Ross S. Pharmacy homeless outreach engagement Non-medical independent prescribing Rx (PHOENIx) community Pharmacy-Based pilot randomized controlled trial. J Urb Health. 2025;102:540–63. 10.1007/s11524-025-00981-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Williamson AE. Inclusion health and missingness in health care: dig where you stand. Br J Gen Pract. 2023;73(735):436–7. 10.3399/bjgp23X734985. Editorial. [DOI] [PMC free article] [PubMed]
- 44.Guo J, Bard F, Morrice JJ, Jaén D, Poursani CR. Offering transportation services to economically disadvantaged patients at a family health center: a case study. Health Syst. 2021;11(4):251–75. 10.1080/20476965.2021.1936658. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Scobie H, MacKinnon S, Wood K, Albanese A, Cunningham Y, Williamson A, Moir J, McPherson A, Lombard C, Ross S, Araf A. Patient participant, healthcare professional, and stakeholder perspectives on the pharmacy homeless outreach engagement Non-medical independent prescribing Rx (PHOENIx) community pharmacy pilot randomised controlled trial. PLoS ONE. 2025;27(6):e0324654. 10.1371/journal.pone.0324654. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Only scientifically sound proposals from appropriately qualified Research Groups will be considered for data sharing. Requests should be sent to bctu@bham.ac.uk. The request will be reviewed by the BCTU Data Sharing Committee in discussion with the Chief Investigators. Completion of a data sharing agreement will be required prior to the access.