What do primary care staff know and do about blood borne virus testing and care for migrant patients? A national survey

Background UK migrants born in intermediate to high prevalence areas for blood borne viruses (BBV) including hepatitis B, hepatitis C and HIV are at increased risk of these infections. National guidance from Public Health England (PHE) and National Institute for Health and Care Excellence (NICE) recommends primary care test this population to increase diagnoses and treatment. We aimed to investigate primary care professionals’ knowledge of entitlements, and perceptions of barriers, for migrants accessing healthcare, and their policies, and reported practices and influences on provision of BBV testing in migrants. Methods A pre-piloted questionnaire was distributed between October 2017 and January 2018 to primary care professionals attending the Royal College of General Practitioners and Best Practice in Primary Care conferences, via a link in PHE Vaccine Updates and through professional networks. Survey results were analysed to give descriptive statistics, and responses by respondent characteristics: profession, region, practice size, and frequency of seeing migrant patients. Responses were considered on a per question basis with response rates for each question presented with the results. Results Four hundred fourteen questionnaires were returned with responses varying by question, representing an estimated 5.7% of English GP practices overall. Only 14% of respondents’ practices systematically identified migrant patients for testing. Universal opt-out testing was offered to newly registering migrant patients by 18% of respondents for hepatitis B, 17% for hepatitis C and 21% for HIV. Knowledge of healthcare entitlements varied; fewer clinical staff knew that general practice consultations were free to all migrants (76%) than for urgent care (88%). Performance payment structure (76%) had the greatest reported influence on testing, followed by PHE and Clinical Commissioning Group recommendations (73% each). Language and culture were perceived to be the biggest barriers to accessing care. Conclusions BBV testing for migrant patients in primary care is usually ad hoc, which is likely to lead to testing opportunities being missed. Knowledge of migrants’ entitlements to healthcare varies and could affect access to care. Interventions to improve professional awareness and identification of migrant patients requiring BBV testing are needed to reduce the undiagnosed and untreated burden of BBVs in this vulnerable population.


(Continued from previous page)
Conclusions: BBV testing for migrant patients in primary care is usually ad hoc, which is likely to lead to testing opportunities being missed. Knowledge of migrants' entitlements to healthcare varies and could affect access to care. Interventions to improve professional awareness and identification of migrant patients requiring BBV testing are needed to reduce the undiagnosed and untreated burden of BBVs in this vulnerable population.
Keywords: Hepatitis B, Hepatitis C, HIV, Blood borne virus, Testing, Screening, General practice, Migrants, Healthcare access, UK Background Migrants in the UK from intermediate or high prevalence areas for HIV, hepatitis B (HBV) and hepatitis C (HCV) are at increased risk of these infections compared to the UK-born population, and experience a disproportionate burden of disease [1][2][3]. In 2017 almost half (49%) of newly diagnosed cases of HIV among heterosexuals were among black Africans or heterosexuals born in a high prevalence country [2] and an estimated 95% of new chronic HBV infections in the UK are among individuals who probably acquired their infection overseas in endemic countries, often perinatally or in childhood [4]. For all three blood borne viruses (BBV) the undiagnosed fraction remains high, with a high proportion of infections diagnosed late, potentially contributing to avoidable morbidity and mortality and increased risk of onward transmission [5][6][7][8].
In 2016, the UK government signed up to the World Health Organisation (WHO) goal of eliminating viral hepatitis as a public health threat by 2030 [9]. HBV and HCV testing in primary care for migrants from countries with intermediate or high prevalence (> 2%) has been recommended by National Institute for Health and Care Excellence (NICE) guidance since 2012 [10], and since 2018 European Centre for Disease Prevention and Control (ECDC) recommends screening newly arrived migrants for BBV [11]. Although surveillance data indicates that an increasing proportion of BBV diagnoses take place in primary care, evidence suggests a low proportion of eligible migrants receive the recommended BBV testing in this setting [12,13].
Language, stigma and poor knowledge and understanding of disease create barriers for migrants to accessing testing and treatment for viral hepatitis [14,15] and migrants often experience discrimination and bureaucratic obstacles when trying to access healthcare [16,17]. In August 2017 UK healthcare entitlements for migrants changed and the list of chargeable services was expanded [18,19]. While many services, including primary care, are still free to all migrants, for others a patient should now be deemed ordinarily resident in the UK to access services for free. Sexual health, family planning, GP and nurse consultations in primary care, emergency department (ED) and walk-in centres, and treatment of mental and physical conditions caused by torture, female genital mutilation (FGM), domestic or sexual violence are free to all migrants, while communicable disease diagnosis and treatment is free to all for a list of specified diseases, including HIV, viral hepatitis, TB and Middle East respiratory syndrome [18][19][20]. For surgery or outpatient services in secondary care, a patient should be deemed ordinarily resident in the UK to access the service free of charge. This additional complexity may lead to uncertainty among both migrants and professionals about entitlements, and could create additional barriers to accessing services.
We, therefore, conducted a national survey to understand primary care staff's knowledge, attitudes, policies and practice regarding BBVs in migrant patients, influences on these, and their perceptions of the barriers for migrants in accessing healthcare.

Study design
A cross-sectional survey conducted among primary care staff in England.

Study population
The target population for this study were primary care staff working in general practices in England. All primary care staff were eligible for inclusion, but it was expected that respondents would primarily be general practitioners (GPs) and practice nurses.

Sampling procedures
A convenience sampling approach was used, with survey participants recruited through several routes during October 2017. Firstly, the link to the online survey was circulated via primary care networks of local Public Health England (PHE) Health Protection Teams across England. Secondly, the link to the online survey was circulated in the PHE publication Vaccine Updates, a newsletter primarily subscribed to by general practice professionals with distribution~46,000. These two routes were used opportunistically and in parallel in order to access primary care professionals, as we did not have access to another primary care network with equivalent coverage across England.
In addition, stalls were held at the Royal College of General Practitioners (RCGP) annual conference in Liverpool and the Best Practice in Primary Care conference in Birmingham, both in October 2017. Attendees were approached and asked to participate in the survey via tablet devices, or offered shorter paper versions of the survey to take away and return once completed.
Due to the timeframe and budget of the project, participants were not followed up to respond to the survey, and the online survey was closed in January 2018.

Data collection
The survey was created as an online questionnaire using SelectSurvey.net™ software (ClassApps LLC, Kansas City, Missouri) and responses were exported into Excel for analysis. Shorter paper versions of the survey were also distributed among participants attending the two primary care conferences. Responses from paper surveys were manually input into Excel by the lead author.

Questionnaire design
The questionnaire was developed at PHE in collaboration with a project group of clinicians, academic GPs and qualitative researchers and piloted before launch with a group of 4 academic GPs.
The questionnaire consisted of 37 questions covering the following topics. The full questionnaire took approximately 17 min to complete (see additional file 1 for full online questionnaire and additional file 2 for shorter paper questionnaire). Most questions were multiple choice, apart from those relating to barriers, practice size and practice population which were free text.
To investigate knowledge of healthcare entitlements respondents were asked, as of August 2017, which services were free irrespective of migration status, using legislation and implementation guidelines to define correct responses [18,19].
Respondents were asked about location of practice and professional role. No additional respondent characteristics were sought as the focus of the survey was on practice, not individual characteristics.

Definitions
A migrant is defined as someone who changes their country of usual residence, in whatever manner, for whatever reason and irrespective of their legal status. The UN defines a permanent migrant as someone who moves to a country for a period of 12 or more months, often for work, education or joining family, effectively making it their new country of usual residence [21]. A refugee is defined as someone who has fled their home country for reasons of feared persecution, conflict, generalised violence or other circumstances that have seriously disturbed public order and are seeking protection in another country [21]. In the UK, for a person to be officially recognised as a refugee they first have to make a claim for asylum and have this accepted by the government. An asylum seeker is a person who seeks safety from persecution or serious harm in a country other than their own and has formally applied for asylum (i.e. refugee status) from that country, but is awaiting a decision on their application [22].
A definitions sheet was provided with the survey to help respondents understand terms relating to migration which also included refused asylum seeker and undocumented migrant (not specifically mentioned in the questionnaire) (additional file 3).
Definitions of other terms used are as follows: Clinical staff: respondents who identified their role as either GP (GP partner, Salaried GP, Locum GP, or 'other' and specified GP), or nurse (Practice Nurse or 'other' and specified nurse). In questions relating to practice policy for BBV testing, new migrants refer to newly registering patients who were migrants from outside the UK (regardless of when they had migrated), and existing migrants refer to patients already registered with the practice who had ever migrated from outside the UK. When asking about practice characteristics, the proportion of new migrants on a practice list was defined as migrants registered with the practice who had lived in the UK for 5 years or less, regardless of when they registered.
The roles of GPs and nurses in the diagnosis and management of BBVs vary by practice. Practice nurses undertake a range of nursing assessments and provide appropriate care / treatment in conjunction with GPs according to practice policy and protocols and are likely to be responsible for vaccination and testing, whereas GPs would usually be responsible for clinical diagnoses and referrals to secondary care.
Practice size was defined as small, medium or large using tertiles of practice populations in England reported by NHS Digital for December 2017 [23].

Data analysis
Results were collated and de-duplicated using email addresses. Practice names were checked to identify if there were multiple respondents from the same practice, but these were not de-duplicated as responses differed and were from different professionals.
Due to the methods used to circulate the survey via an e-bulletin to which practices subscribe, it was not possible to calculate a formal response rate. Instead, the number of responses as a proportion of GP practices in England (based on NHS data from 2017) was calculated to provide an indication of the proportion of England GP practices represented in responses [23].
Results were analysed in Excel and R for overall descriptive statistics (percentages and 95% confidence intervals (CI)) and then stratified by respondent and practice characteristics. Responses for clinical knowledge and practice questions were restricted to those who gave their role as GP or nurse and stratified by profession. Responses for practice policy questions were analysed for all respondents, and stratified by region, frequency of seeing migrant patients, and practice size.
To assess the statistical significance of differences in responses to questions, and for differences by stratified analysis, variables were dichotomised where required before performing Chi squared tests. Where this was done, the dichotomous outcome is highlighted in bold in results tables. Fishers exact tests were used where the cross-tabulation did not meet the criteria for Chi squared tests. Differences were considered statistically significant where p < 0.05.
Free text responses on barriers were grouped by thematic analysis into 10 themes: language/culture, patient information/knowledge, staff information/knowledge, psychological, accommodation, prejudice/discrimination, patient financial, organisational resource, geographical, and service/organisational issues.
Each question had a different response rate, and descriptive statistics were based on the number of respondents for each question, which are presented with the results.

Respondent and practice characteristics
The survey was circulated through the PHE Vaccine Updates circulation list, although it is not known how many will have actively read this, and others may have been forwarded the survey link by contacts. After deduplication 414 responses were obtained; 16 paper and 398 online, an estimated 5.7% response rate, based on the number of GP practices in England. The majority of responses came from circulation of the survey link as response at conferences was poor; no attendees completed the online survey on the day, and only 16 paper surveys were returned. Where practice name was reported (145 respondents), there were 3 practices that each had 2 respondents, all other practices were unique. Most respondents were practice nurses (49%, 202/414) and GPs (40%, 165/414) with the highest responses from the Midlands (21%, 86/414), the South East (16%, 65/ 414) and London (15%, 63/414) ( Table 1). Responses were equally distributed among those who reported they saw migrant patients frequently (34%, 108/317), sometimes (33%, 103/317) and rarely (33%, 106/317) (77% response rate).

Professional knowledge and practice
Issues considered for newly registering patients who are migrants When asked which issues were always addressed for newly registering patients who are migrants (response rate 78% of GPs and nurses), vaccination history was the most common always addressed issue (58.2%, 167/287, 95% CI 52.4-63.7), and family planning, TB screening, sexual health and BBV screening more often considered on a situation-specific basis ( HCV and HIV respectively and this did not differ by the respondent's profession (p = 0.763, p = 0.953 and p = 0.867 respectively). TB screening was addressed for all newly registering migrant patients by 28.6% (82/287, 23.7-34.1) of respondents, and this did not differ by profession (p = 0.072). There were significant differences in practice by profession for vaccination history, family planning, and sexual health advice and screening (all p < 0.01), with nurses more likely than GPs to address these issues for all newly registering migrant patients.

Knowledge of migrants' entitlements to health services
When asked about migrants' entitlements to health services (100% response rate), 88.3% (323/366, 84.5-91.2) of clinical staff correctly identified emergency departments and walk-in centres as free to all and 76.0% (279/367, 71.4-80.1) correctly identified GP and nurse consultations in primary care as free to all, with no difference by profession (p = 0.565 and 0.262 respectively, Table 2). Surgery or outpatient services in secondary care were correctly identified as not free to all by 52

Identification of migrant patients for BBV testing
When asked if existing migrant patients were identified for testing opportunistically during consultation, 78.9% (191/242, 73.4-83.6) of respondents stated they were (61% response rate), while only 13.6% (33/242, 9.9-18.5) stated that systematic identification methods, either periodic flagging of the GP system by automated software, or periodic manual audit of the GP system, were used (61% response rate) (  (18/241, 4.8-11.5) had health support teams and 4.6% (11/241, 2.6-8.0) had incentive schemes for GPs. Outreach services (p = 0.024) and projects to register migrants (p = 0.019) were both more common in small practices, and projects to register new migrants, specially designated clinics, longer appointments at registration (all p < 0.01) and outreach services (p = 0.029) were all more common in practices where respondents saw migrant patients frequently. Provision of longer appointments at registration varied significantly by region and was highest in Yorkshire and Humber (46.9%, 15/32, 30.9-63.6), followed by the North East (41.7%, 5/12, 19.3-68.0) and was lowest in the North West (6.3%, 2/32, 1.7-20.1) (p < 0.01) (supplementary Table 2).

Perceived barriers for migrants to access healthcare across the care pathway
The content and proportion of responses in each theme varied according to which part of the care pathway was considered (Fig. 1, additional file 4).
Language and culture was also the most commonly cited barrier to accessing BBV services in primary care ( For patients to access BBV services in secondary care, again language and culture was the most common theme (41.5%, 98/236, 35.4-47.9, 60% response rate), followed by patient information and knowledge, with responses often relating to patients not understanding the significance of the diseases and importance of accessing treatment (28.4%, 67/236, 23.0-34.5). Geographical issues were cited by 24.5% (58/236, 19.5-30.4), including transport costs and lack of locally available services, and 21.2% (50/236, 16.5-26.8) cited service issues, including communication methods around arranging appointments, waiting times and lack of an organised clinic.
Perceived barriers for asylum seekers accessing healthcare (74% response rate) were largely similar to those for all migrants, with some notable differences (Fig. 2). Language and psychological barriers were more frequently reported for asylum seekers than they were for all migrants. Psychological barriers again included fear, as well as highlighting mental health concerns for asylum seekers. To a lesser degree, accommodation, and prejudice and discrimination were also identified more often for asylum seekers than for all migrants.

Discussion
Primary care staff knowledge, attitudes, policy and practice regarding migrants' healthcare entitlements and their BBV testing and care was very variable in this sample of English practices. One in four were unaware that GP and nurse consultations are free for all. Universal opt-out BBV testing for migrants was not common practice, and only one in three clinical staff would routinely consider BBV risk assessments for newly registering migrant patients. Where testing was requested, most respondents requested the recommended tests, and over 90% indicated they would refer diagnosed cases to secondary care. Respondents identified access barriers relating to a range of factors including language; patient knowledge, stigma and fear; time pressures of appointments, practices not having a BBV testing policy and lack of knowledge and awareness among staff.
Knowledge of entitlements to health services varied. Of particular concern is that a quarter of clinical staff were unaware that GP and nurse consultations are free to all. Vulnerable patients, including migrants could be wrongly refused GP registration, despite national guidance that lack of documentation should not be a barrier to registering with a GP [24]. Three UK studies estimate that refusal ranges from 20 to 39% [25 -27]. As primary care is the first point of access, 'gatekeeper' to other services, and a trusted source of health information, GP staff's poor knowledge of entitlements is likely to create a further access barrier in addition to the system, cultural and language barriers migrants face, especially when it comes to communicable disease control [28,29]. There were no substantial differences in knowledge of healthcare entitlements of migrants between GPs and nurses; GPs were more likely to know that operations and outpatient appointments are not free, and nurses were more likely to know that care for FGM, torture etc. is free, which likely reflects their roles in general practice.
BBV testing was more often conducted on an ad-hoc than opt-out basis, and BBV risk assessments for newly registering migrant patients were routinely considered by only a third of respondents, and more often considered on a situation specific basis. Most respondents stated existing migrant patients were identified for testing opportunistically during consultation, and only 14% systematically flagged migrant patients for BBV testing. These findings suggest that migrant patients eligible for BBV testing are not often routinely identified as such, which is likely to hinder adherence to BHIVA, ECDC and NICE guidelines which recommend BBV testing for migrant patients from higher HBV and HCV prevalence (> = 2%) areas and HIV testing for all patients in high (> 2 per 1000) prevalence areas in the UK [10,11,30], and is consistent with other studies which report low adherence to testing guidelines for BBVs outside of the sexual and antenatal health setting [12,31]. Opt-out rapid testing in general practice has been shown to increase diagnosis rates for HIV, and universal optout testing initiatives for migrants which offer opportunities to normalise testing are subject to ongoing evaluations and have been well received by staff and patients [32][33][34].
Improved practice staff knowledge is needed to help reduce barriers and variations in service provision. Resources are available, such as the PHE Migrant Health Guide which outlines NHS entitlements and gives guidance on assessing new patients from overseas [35], RCGP e-learning courses [36], but also PHE and charities co-branded promotional videos, posters and leaflets on BBV in multiple languages for patient waiting areas can be downloaded or ordered from gov.uk [37]. Further work is needed to understand current awareness and to increase uptake of these resources.
Use of software programmes to identify and automatically flag patients for testing based on risk factors or risk proxies recorded on practice registers has been shown to be effective for groups including migrants in primary care and other settings [38][39][40]. This requires migrant status to be recorded on GP systems for migrants to be flagged as at risk. Although 59% of respondents stated that their practice routinely recorded country of birth, data from the Clinical Practice Research Datalink suggests recording is a lot lower at 1.6% [41]. This discrepancy could in part be due to biases in our sample or respondents giving socially desirable answers, however it may be that practices are recording this data on other systems where the data is not extractable. This highlights a need to increase awareness of the importance of recording patient country of birth amongst practitioners and information service providers, and for practitioners to assess risks associated with country of birth as a part of the first registration consultation.
Flagging and opt-out testing for newly registering migrant patients was more frequently reported where respondents said they see migrant patients frequently, in smaller practices and highest in Yorkshire and Humber and London, but lowest in the South East. A similar pattern was observed for services such as specially designated clinics, longer appointments at registration and projects to register migrantsthese were more frequently reported where respondents saw migrants frequently. Outreach services and projects to register migrants were higher in small practices. These variations may reflect GP surgeries adopting policies and practices that reflect the needs and diversity of their local populations, and the ability of smaller practices to be more responsive and adaptive in implementing migrant focused services.
Performance payment structure may be the most influential way to incentivise BBV testing [42][43][44], and PHE and CCG recommendations were also more often stated to provide motivation to test than alternative sources. Only half of respondents indicated that NHSE and CMO recommendations would provide motivation for testing, which may reflect a relative lack of confidence in these bodies versus the perceived scientific credibility and independence of PHE and the more direct relationship and understanding of local PHE teams with CCGs. Interestingly, nurses were more likely than GPs to be motivated by local targets, national goals, CCG recommendations and NICE guidance.
Over 90% of GPs reported referring diagnosed cases of HBV and HCV to secondary care, and over half followed patients up if they did not attend secondary care appointments. Three quarters would offer close contacts of HBV infected cases testing, and three quarters would offer close contacts hepatitis B vaccination. These proportions seem high, as many patients diagnosed with HBV and HCV are known not to be accessing secondary carein a 2012 UK study only a third of HBV diagnosed patients were on treatment [45] and an analysis of England sentinel surveillance data showed that between 2004 and 2017 only 21% of viraemic HCV diagnosed patients had evidence of treatment [46], and baseline HBV testing of close contacts was only 34% in a 2018 UK study [47]. Our results may be due to sampling and responder bias, due to initially contacting networks who were interested in migrant health, and those who were interested being more likely to respond, as well as wishing to give socially acceptable or desirable responses.
Differences in knowledge of BBV diagnosis and management were noted between nurses and GPs, with nurses being less likely to know whether all diagnosed HBV/HCV patients are referred and whether referrals are followed up, and less likely to know the diagnostic tests to request for HCV and HIV but there was no significant difference for HBV. Generally, there is a need for improved understanding of diagnostic testing for BBV in practices. Regardless of service, respondents cited language as the main barrier to accessing healthcare, which is consistent with the literature [14,15,48]. For patients who require linguistic support, interpreting services are recommended in General Medical Council guidance and were provided by 88% of respondents' practices [49][50][51]. Nevertheless, there is evidence that interpreting services are underused, despite incurring no direct cost for practices [52]. This may be partly due to time pressures within appointments and the indirect costs of longer appointment times, as well as the need for forward planning and a lack of professional confidence in working with interpreters [53]. Brief education may help improve professional confidence in when and how to use interpreters [54]. Improved provision of translated materials is also likely to be beneficial in improving patient access to health information; efforts to ensure materials about BBV are translated are ongoing [55].
Patient knowledge and understanding of disease, and the location of secondary care services were cited as barriers to accessing secondary care. Community based treatment pathways, as have been demonstrated to be effective for HCV [56,57], could help to overcome geographic barriers, although one recent study found that (first and second generation) migrant patients were no more likely to complete treatment on a community pathway [40]. Assessing patient attitudes towards BBV, and increasing patient knowledge would also be valuable and could be considered by practices to improve their services.

Limitations
This survey has a low response rate and a relatively small sample size, and may be affected by sampling and responder bias, as the survey was disseminated by contacts and networks with an interest in migrant and asylum seeker health and through a publication which professionals with an interest in vaccines subscribe to, so practice professionals with interests in migrant health, BBVs and vaccine preventable diseases may have been more likely to respond. This is likely to bias the responses towards more knowledge of healthcare entitlements and better practices for BBV screening and management of migrants. Response to recruitment at professional conferences was poor, possibly due to competing priorities for professionals with limited time in a conference setting.
No demographic data on respondents was collected, so it was not possible to assess how responses vary by individual respondent characteristics, which may have associations with personal bias, stigma and discrimination against individuals who have migrated from specific countries or settings.
Our findings may therefore not be representative of all general practice staff. It was not possible to obtain data on non-respondents which would help to identify biases in responses; further work to survey practitioners within a fixed sampling frame (for example all GP practices in one CCG) may help to address this issue. In addition, not all questions were completed by all participants and fewer respondents completed later questions in the survey. Again, those more motivated may be more likely to complete the whole survey, leading to potentially greater bias in questions that had fewer respondents.
The small sample size limited our power to detect statistically significant differences between subgroups, particularly in detecting regional variations. Increasing the responses to the survey would help to address this, and in hindsight, following up the circulation of the survey link in Vaccine Updates could have helped to increase sample size and potentially reduce response bias.
Another limitation was that questions about barriers for migrant patients were asked to general practice professionals rather than migrants themselves; responses will reflect only barriers that these professionals were aware of and may exclude those which migrants experience, but professionals are less aware of. However, our findings echoed those cited by migrants in a qualitative study on barriers to accessing healthcare for viral hepatitis [14], which suggests that our respondents were relatively well informed about these issues.

Conclusions
Systematic or universal opt-out testing for migrant patients are uncommon and testing is more often done on an ad hoc basis, despite BASHH, PHE and NICE guidelines. Achieving the WHO goal of eliminating viral hepatitis as a public health threat by 2030 requires coordinated efforts to increase case-finding of patients with HBV and HCV; our results suggest that current testing practices for high-risk migrants in primary care are likely to be inadequate to meet the elimination goal among this population. Improved systematic risk-based flagging of migrant patients for BBV screening and checks or prompts for referral and attendance to specialist services is needed to reduce the burden of undiagnosed infection and improve uptake of treatment, noting this is curative for hepatitis C. Local and national commissioning specifications for primary care testing and management of BBV infected patients may be required to achieve this goal. General practice professionals' knowledge of migrants' entitlements to healthcare was variable and could affect migrants' access to care. Perceived barriers to accessing healthcare consistently included language and lack of patient and staff information and awareness. These are not insurmountable but require sustained commitment, professional awareness as well as resource. The ultimate goal is to reduce morbidity, mortality and inequalities associated with BBV infections in migrant populations, mindful of the fact that this population is already vulnerable to disparities in healthcare access and health outcomes.