Prevalence and risk factors for key infectious diseases amongst migrants to the UK: a systematic review

Abstract

Background

Migrants are at increased risk of infections including HIV, tuberculosis and viral hepatitis, with poorer outcomes. Early diagnosis and management can reduce morbidity, mortality and onward transmission. This systematic review summarises prevalence of HIV, latent and active tuberculosis and hepatitis B and C among UK migrants and evaluates associated risk factors.

Methods

PubMed/Medline, EMBASE, Web of Science and the Cochrane Library were systematically searched from 2004 to 11 June 2025. The review was conducted using PRISMA guidelines and registered with PROSPERO (registration CRD42024521191). Quality assessment was performed using the Joanna Briggs Institute Critical Appraisal Checklist for Prevalence Studies. High heterogeneity (I² = 95.2%, 99.2%, 87.2%, 96.9% and 91.6% for IGRA, active TB, HIV, HBV and HCV yields, respectively) indicated that meta-analysis was not appropriate. The impact of risk factors on prevalence was explored through meta-regression and descriptive analysis.

Results

Of 2033 identified records, 36 were included, reporting Interferon Gamma Release Assay (IGRA) (n = 13), active TB (n = 10), HIV (n = 12), HBV (n = 16) and HCV (n = 11) test yields. An additional two publications excluded from the main analysis for reporting duplicate study data were included in the risk factor analysis because they stratified prevalence by additional risk factors. Highest yield was for IGRA which, excluding one lower prevalence outlier (6.9% (n = 1617)), was 15.1%–22.1%. There was high heterogeneity in active TB prevalence: 62–1,484/100,000. HIV prevalence among larger studies (n > 200) was 0.18%–0.48%. HBV prevalence was 0.00%–8.93% (all studies) and 1.06%–4.75% for larger studies (n > 1000). HCV prevalence was lower: 0.00%–1.67%, with only two of 11 included estimates above 0.50%. There was considerable heterogeneity in risk factors analysed making comparisons difficult.

Conclusions

Despite heterogeneity, infection prevalence was generally high, particularly IGRA yield and HBV. This underscores the need to maintain effective monitoring, testing and treatment for key infections among migrant populations, especially given the rapidly evolving epidemiological and demographic landscape for this population.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-026-12953-z.

Introduction

Migrants to the UK are at increased risk of communicable diseases compared to the general population, and the chronic infections tuberculosis (TB) and blood-borne viruses (BBVs) HIV, hepatitis B, and hepatitis C have been identified as key health priorities for migrants by international and national organisations, advocating an integrated approach to testing [1–3]. In 2024, 82% of TB notifications were in people born outside the UK [4]. An international systematic review, including UK-based studies, has shown that HIV prevalence is higher in migrants than in native-born populations [5]. The United Kingdom Health Security Agency (UKHSA) reported in 2024 that 95% of new chronic hepatitis B diagnoses in the UK are in migrants [6], and HCV seroprevalence in women delivering live-born infants in North Thames, England is higher in non-UK-born than UK-born mothers [7].

Increased infection risk stems from multiple factors across the migration journey – before, during and after arrival – including greater exposure to infections, substandard living conditions, and limited access to healthcare [8]. Evidence shows that migrants are more likely to present later with these infections (e.g., overseas-born individuals are significantly more likely to present with HIV infection at CD4 counts below 350 cells/mm³ [9]). They may also experience more aggressive disease progression, as seen with TB [10], and may transmit infections to contacts if left undiagnosed [11, 12].

Early diagnosis and management of these infections are crucial for improving health outcomes by reducing morbidity, mortality and onward transmission [13]. This position is supported by several national guidelines, including those from the National Institute for Health and Care Excellence (NICE), which recommend screening migrants for active and latent TB [14, 15], HIV [16–18], hepatitis B [19, 20] and hepatitis C [19–21] to support national prevention targets [19, 22–24].

In many low TB incidence countries (World Health Organization definition of < 10 per 100,000 population [25]) such as the UK, as well as countries in the European Union, North America, Australia and New Zealand, most active TB cases in migrants result from reactivation of latent TB infection [26]. The stressors associated with migration may act as a trigger for this reactivation, underscoring the importance of screening and treating both latent and active infections [27]. Furthermore, TB prevention and control are particularly important now, given that TB notifications are increasing in the UK [28].

As we attempt to reach targets for control of these infections, we need to know prevalence of infection within specific risk groups such as migrants for effective health planning, particularly for approaches to screening. In this systematic review, we aimed to evaluate recent (within the past 21 years) evidence regarding prevalence of key infections (HIV, TB and viral hepatitis) amongst migrants to the UK, to aid planning appropriate, effective interventions to screen migrant populations. We also aimed to explore prevalence stratified by established and potential risk factors for infection, which could inform targeted testing strategies adopting a risk-based approach.

Methods

This systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [29] to uphold the integrity, transparency and reproducibility of the research process. The study was registered with PROSPERO (CRD42024521191). The primary outcomes of interest were point prevalence and/or test yield for active TB, HIV, HBV, HCV and IGRA test positivity (used as a proxy for latent TB infection or exposure to TB infection). Secondary outcomes were primary outcomes stratified by risk factors, where reported. We did not predefine these risk factors but extracted data for all factors reported by included studies.

Search strategy

We searched PubMed/Medline, EMBASE, Web of Science and the Cochrane Library from 2004 to 11 June 2025 for studies reporting prevalence of any one or more of the included infections (HIV, TB, HBV and HCV) among any international migrant type migrating to the UK. The search conducted up to March 2024 included publications from the preceding 20 years, as data from earlier periods were considered less relevant to current infection prevalence. A 20 year-window was considered sufficient to yield an adequate number of studies for each infection and to facilitate the examination of temporal trends. This original search was subsequently updated in March and June 2025, resulting in a total 21 years of data being included. There were no language or publication status restrictions. We used the following Medical Subject Heading (MeSH) and keyword terms: (HIV OR tuberculosis OR TB OR hepatitis B OR HBV OR hepatitis C OR HCV OR “viral hepatitis”) AND (migra* OR immigra* OR refugee* OR asylum) AND prevalence AND (England OR Scotland OR Wales OR Ireland OR Britain OR United Kingdom OR UK).

Eligibility criteria

Studies were included if they reported prevalence of HIV, HBV, HCV or active TB infection, or test positivity rates (test yields) for these infections, including IGRA testing for latent TB infection/past exposure (or from which prevalence could be derived), for any international migrant type settling or settled in the UK. We used the International Organization for Migration (IOM) definition of migrant [30]. Estimates were also included where the study population, or a subgroup of the study population, was described as non-UK-born and where there was clear implication that these study participants were residing in the UK long-term. Studies reporting results from pre-entry migrant screening programmes were included. Exclusion criteria included estimates based on self-reported diagnoses, prevalence reported for travellers/international exchange students rather than migrants, case reports, contact tracing studies, outbreak investigations, editorials and reviews. Modelling studies, where the denominator was an estimate of the migrant population size rather than representing all migrants who were tested for/clinically examined to diagnose infection, were excluded. Studies which only reported prevalence of specific TB conditions (e.g., pulmonary, extrapulmonary, spinal TB) were also excluded. Studies were excluded if the data they reported were superseded by more recent studies, unless the earlier study reported more relevant information. For the secondary outcomes (prevalence/test yields stratified by risk factors), estimates of relative risk were insufficient for inclusion – number tested and number diagnosed/testing positive had to be stated. There was no restriction based on age, sex or ethnicity.

Study selection

Records identified through database searching were merged and duplicates removed using Rayyan Software [31]. Two reviewers (CMH, LS, LZ, PB), including the lead investigator (RFB), independently screened the records by title/abstract and full text according to the inclusion/exclusion criteria. Any disagreements were resolved by the lead investigator.

Data extraction

Data were extracted using a predefined form to capture study information including author, year of publication, study characteristics, study participant demographics, test type, numbers tested and test outcomes, including coinfection estimates. While we only included prevalence of active HCV infection in the analysis, we also recorded estimates of HCV antibody test positivity, indicating previous history of HCV infection. Where reported, we included test outcomes stratified by patient characteristics/risk factors. The data extraction form was refined during the extraction of the first few articles to ensure the forms were comprehensive. Two reviewers (RFB with CMH, LS, LZ or PB) independently extracted data, with discrepancies resolved through consensus. We contacted eight authors for further information; seven replied and one provided further data.

Quality assessment

All studies were assessed independently by two reviewers (RFB with CMH, LS, LZ, or PB) using the Joanna Briggs Institute Critical Appraisal Checklist for Prevalence Studies [32, 33] to identify potential biases and determine the quality of evidence the studies provide, and were categorised as being of high, moderate or low quality. Discrepancies were resolved through consensus.

Data synthesis and analysis

We summarised prevalence data from included studies using forest plots (produced in R version 4.4.2 using the meta package), stratified by infection type. All prevalence estimates and 95% confidence intervals (95%CI) were recalculated using data from each study on numbers tested and numbers testing positive, to ensure consistency (95%CI were calculated using the Wilson score method). We evaluated heterogeneity in prevalence estimates using the I² statistic and, given the high level of heterogeneity observed for all infections (I² >75% is assumed to represent considerable heterogeneity) and the variability in study populations, pooling of estimates through formal meta-analysis was not considered to be appropriate. Instead, qualitative synthesis was conducted to describe and summarise findings and identify patterns across studies.

For HIV, estimates were categorised as extremely high (> 0.5%), high (0.2–0.5%) and low (< 0.2%) prevalence, based on established thresholds used to define UK local authorities [34, 35]. HBV estimates were categorised as high (≥ 5%), intermediate (2-4.9%) and low (< 2%) prevalence to align with current World Health Organization national endemicity categories [36]. In the UK, HCV prevalence is not typically defined using formal numerical categories. We therefore defined a pragmatic, informal categorisation of prevalence, for descriptive purposes only, using the 0.12% estimated prevalence for the general population in England in 2023 [37], classing up to 150% of general population prevalence as low (up to 0.18%), up to three times the general population prevalence as intermediate (> 0.18% up to 0.36%) and above that as high (> 0.36%) prevalence. A similar approach was applied for IGRA yield, because IGRA testing has been restricted to targeted populations rather than broad, unselected population surveys. A hospital-based TB outbreak study reported an IGRA yield of 8.7% among unexposed age- and other demographics-matched control patients [38], which may overestimate yield in the general population, as patients may experience more exposures. In contrast, a study of farm workers in England with potential Mycobacterium bovis exposure reported an overall IGRA positivity rate of only 1.1%, despite occupational exposure (but likely few other risk factors for exposure) [39]. We therefore assumed a UK general population IGRA positivity rate of 5% to inform categories, using the same methodology as for HCV: classing up to 150% of general population IGRA positivity as low (up to 7.5%), up to three times as intermediate (> 7.5% to 15%) and above that as high (> 15%) test positivity. In the absence of formal prevalence categories for active TB, we used migrant country-of-origin TB incidence screening thresholds as a pragmatic, informal indicator. High prevalence was defined as ≥ 40 per 100,000 (the threshold used for UK pre-entry active TB screening [40]), and extremely high prevalence as ≥ 150 per 100,000 (the threshold used for UK latent TB infection testing [40]).

We then conducted a risk factor analysis, which included conducting univariable and multivariable meta-regression analyses for each infection, using continuous covariates (median date of study data collection) and covariates that could be categorised into a limited number of groups: study quality (high, moderate, low) and timing of testing (pre-entry, at entry or post arrival to the UK). Number of covariates included in multivariable meta-regression was restricted to these, to reflect the limited number of study estimates for each infection and the incomplete data reported by studies for many potential covariates. Characteristics of migrant populations (migrant type e.g., asylum seekers, students, sexual health clinic attendees etc.) reported by studies were very diverse, and so they were summarised in text form on forest plots, for descriptive analysis. We attempted visually to synthesise estimates stratified by age, sex and country of birth/origin of migrants using heat maps, aligning mismatched risk factor categories (e.g., different age group definitions between studies) as far as possible and using the infection prevalence categories described above. Spatial visualisations were produced in R using the tmap package to create choropleth maps showing the distribution of studies across UK counties and unitary authorities using Office for National Statistics 2023 data [41]. In addition, we compared infection prevalence estimates for UK migrants stratified by country of origin with prevalence in country of origin, using data for HIV [42], HBV and HCV [43] (there were insufficient comparable data to perform a comparison for active TB and IGRA yield). Results were summarised as scatter plots.

Results

The database search yielded 2033 articles. 1572 articles were unique, after removal of duplicates, and of these, 85 were selected for full-text screening. Review of the full-text articles led to the inclusion of 36 articles in the analysis (see PRISMA flow diagram, Fig. 1). Included studies reported prevalence of latent TB infection/IGRA positivity (n = 13), and active TB (n = 10), HIV (n = 12), HBV (n = 16) and HCV (n = 11) infection. Two studies reported prevalence from pre-entry screening. Crawshaw et al. reported testing data from International Organisation for Migration clinics enrolled in the UK pre-entry migration health assessments, for refugees resettled in the UK [44]. Menezes et al. reported data extracted from the UK new entrant screening programme (which does not include asylum seekers, who undergo domestic health checks and are not part of pre-entry screening) to compare infection prevalence with other European countries [45]. The remaining studies were conducted at entry to the UK (defined as immediately after arrival e.g., testing at airports [46] or shortly afterwards e.g., asylum seekers at initial accommodation centres [47]); or post-arrival (after the at entry stage e.g., testing upon registration with primary care [48]).

Fig. 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram of included studies. *Two studies reporting prevalence of active TB are not included in the forest plot analysis because of overlapping data: overall prevalence estimates are reported by Menezes et al. [45], but Aldridge et al. [56] and Zenner et al. [57] both provide additional prevalence data for the same study population stratified by additional risk factors and have been included in the risk factor analysis

More studies were conducted in London than elsewhere in the UK (n = 23 in London or Greater London, including studies with catchment defined as England or the UK) (see map of distribution of study settings, Supplementary Material Fig. S1). The next most frequent study location was Birmingham (10 studies). Two studies were classified as UK-based because they reported results of UK pre-entry screening [44, 45]; one was a national screening programme for Ukrainian refugees in Wales [49] and five were based on testing migrants across England [50–54]. The remaining studies had smaller study regions but often included multiple study sites; for example, Flanagan et al. reported from areas of the UK with a high density of migrants (Bradford, Yorkshire, and northeast and southeast London) [55]. However, these studies did not usually report prevalence estimates stratified by study region, and the limited number of included studies makes comparisons between studies by UK study region difficult.

Two of the ten studies reporting prevalence of active TB [56, 57] were not included in the forest plot analysis because Menezes et al. [45] reported prevalence from the same study population, but these studies were included in the analysis of prevalence stratified by risk factors because they provided additional data. Therefore the 34 articles reporting on different study samples shown across the forest summary plots (Figs. 2 and 3) included a total of 2,313,965 migrants tested. A detailed summary of all included studies is presented in Table S1, Supplementary Material. Table S2 lists studies that were excluded because they reported prevalence of HCV antibody test positivity only (indicating prevalence of history of HCV infection rather than current, active infection).

Fig. 2.

Forest plots of prevalence estimates for a) IGRA yield (%) and b) prevalence of active TB (yield per 100,000). Estimates are listed from oldest to most recent based on the median year of data collection. * Overseas-born individuals 16–65y, UK arrival < 5 years, from a high TB incidence country (≥ 150/100,000) or from sub-Saharan Africa or a refugee/asylum seeker (IGRA testing for patients ≤ 35 years only) [48]. Only patients testing IGRA positive were examined for signs of active TB. ** Asylum seekers in the UK undergo domestic health checks and are not part of pre-entry screening. Menezes et al. reported that the majority of screening episodes were among persons with student (45%) or settlement visas (24%), with lower proportions among those on work visas (8%), family reunification (4%) and working holiday maker visas (2%) [45]. E – East; ED – Emergency Department; ESOL – English for Speaker of Other Languages course; exc – excluding; GP – general practice and primary care; HPU – Health Protection Unit; IQR – interquartile range; NS – not stated; SSA – sub Saharan Africa; UASC – unaccompanied asylum seeker children; y – years

Fig. 3.

Forest plots of % prevalence estimates for a) HIV, b) HBV and c) HCV. Estimates are listed from oldest to most recent based on the median year of data collection. * Overseas-born individuals 16–65y, UK arrival < 5 years, from a high TB incidence country (≥ 150/100,000) or from sub-Saharan Africa or a refugee/asylum seeker (IGRA testing for patients ≤ 35 years only) [48]. DBS – neonatal dried blood spot testing; ED – Emergency Department; FSW – female sex worker; GP – general practice and primary care; GUM – genitourinary medicine or sexual health clinic; IAC – Initial Accommodation Centre for UK asylum seekers; MSW – male sex worker; NS – not stated; UASC – unaccompanied asylum seeker children; y – years

Study quality

Table S3, Supplementary Material, reports the quality assessment of included studies using the Joanna Briggs Quality Assessment Tool [32]. A large majority of included studies were of high quality. In total, 25 studies were appraised as high, seven as moderate and four as low quality (10 of 13 IGRA yield estimates were from high quality studies and similarly, 5 of 8, 6 of 12, 10 of 16, and 8 of 11 estimates for active TB, HIV, HBV and HCV prevalence were from high quality studies, respectively). Three low quality studies were conference abstracts, which necessarily are limited in the level of detail that can be reported.

Overall prevalence estimates

Figure 2 summarises IGRA yield and infection prevalence estimates for active TB, while Fig. 3 summarises infection prevalence estimates for HIV, HBV and HCV. As anticipated, given the heterogeneity in included study populations, the I² statistic for each infection indicated substantial heterogeneity (95.2% for IGRA yield and 99.1%, 87.2%, 96.9%, and 91.6% for prevalence of active TB, HIV, HBV and HCV, respectively). We therefore did not conduct a meta-analysis. Study populations varied from a broad inclusion of migrants (e.g., migrants registering with primary care [48, 58]) to specific populations (e.g., army recruits among Nepalese new arrivals [51]; results from Unaccompanied Asylum Seeking Children (UASC) health checks [59]; Ukrainian refugees [49]; Eastern European Genitourinary Medicine (GUM) clinic attendees [60]), making comparisons between the limited number of studies difficult.

The highest prevalence values were observed for IGRA yield where, excluding one outlier with lower prevalence (6.86%, 95%CI 5.73–8.20%, n = 1617, Ukrainian refugees [49]), central estimates of yield were relatively homogeneous, varying between 14.58% (95%CI 8.89-23.00%) and 22.09% (95%CI 20.87–23.36%) (n = 12) with no suggestion of a change in yield over time.

There was high heterogeneity for active TB studies, both in terms of prevalence estimates and study sample sizes (which ranged from 119 [49] to more than two million [45]). Extremely high prevalence was reported from studies screening UASC (Pinto et al. abstract, diagnostic method not stated: 1,681/100,000, 95%CI 462-5,921, n = 119 [61]) and those restricting screening to individuals from high TB incidence countries (recent migrants: 1484/100,000, 95%CI 1083–2031 [48]; 455/100,000, 95%CI 125-1,642 [62]; and new entrants: 339/100,000, 95%CI 314–365 [46]). The extremely high prevalence reported by Baggaley et al. reflects a study in which active TB was diagnosed among IGRA-positive migrants, supported by detailed clinical, biological, and radiological assessment—an approach likely to have enhanced diagnostic sensitivity [48]. In contrast, Menezes et al. reported a much lower (though still high according to our pragmatic, informal indicator) prevalence of 69/100,000 (95%CI 65–73) for the 2,006,671 individuals undertaking pre-entry screening in their country of origin by designated clinics [45], although again, these individuals undergoing pre-entry screening were from high TB incidence countries (> 40 per 100,000 population [45, 56]). Earlier studies, including Menezes et al., involved use of less sensitive diagnostic methods such as symptom-based and/or chest X-ray. The two more recent (2018 onwards) published studies which reported relatively low prevalence estimates, despite using more detailed assessment for diagnosis [44, 49], included migrants who were likely to be at lower risk of infection because their countries of origin have lower TB prevalence (primarily Eastern Mediterranean countries [44]; Ukraine [49]) than the countries of origin of migrants in the other recent studies (sub Saharan Africa, high TB incidence countries in Asia) [63].

Sample sizes tended to be smaller for studies reporting HIV prevalence. HIV prevalence estimates were generally high compared to the UK general population (nine of 12 estimates falling in the high or very high prevalence categories). Populations tested were often restricted to high-risk groups such as commercial sex workers and/or GUM attendees. Even studies reporting on similar subgroups of migrants found substantial differences in HIV prevalence. For example, UASC health check studies reported contrasting HIV prevalence levels (1.68%, 95%CI 0.46–5.92% [61] and 0.19%, 95%CI 0.05–0.68% [64], although the relatively small sample sizes mean that 95%CIs overlap). Prevalence among GUM attendees varied from 0.18% (95%CI 0.06–0.52% [60]) and 0.26% (95%CI 0.10–0.68%, female sex workers [52]) to 6.48% (95%CI 3.17–12.78%, GUM/Emergency Department attendee asylum seekers (64)) and 6.92% (95%CI 3.68–12.63%, male sex workers [53]). Less heterogeneity is observed by restricting to larger studies only (sample sizes more than 200) with HIV prevalence varying from 0.18% (95%CI 0.06–0.52%) [60] to 0.48% (95%CI 0.30–0.77%) [48].

HBV prevalence estimates again showed considerable variation, with study estimates between 0.00% (95%CI 0.00-9.64%) [65] and 8.93% (95%CI 5.85–13.39%) [66]. Estimates for larger studies (more than 1000 study participants) varied between 1.06% (95%CI 0.90–1.27%) [55] and 4.75% (95%CI 4.34–5.19%) [67]. Prevalence of HCV was lower, with studies reporting prevalence between 0.00% (four studies: [61, 65, 66, 68]) and 1.67% (95%CI 1.33–2.09%) [69]. Only two of the 11 included HCV estimates were above 0.50% [69, 70].

Table S4 summarises coinfection prevalence reported by studies estimating prevalence of multiple infections. Eight included studies provided estimates of prevalence of coinfections, or sufficient data from which to derive them [44, 48, 55, 64, 65, 68, 69, 71]. Excluding three studies with small sample sizes (n < 200 [64, 65, 71]), HBV-HCV coinfection prevalence varied between 0.00% [44, 48, 68] and 0.09% (95%CI 0.04–0.23%) [69] (n = 4). HIV-HBV coinfection prevalence was reported by two studies (0.02% [44] and 0.06% [48]). Two studies reported 0.00% coinfection prevalence of HCV (with active TB [44] and with IGRA, HIV and HBV [48]). IGRA-HIV and IGRA-HBV co-prevalence values were 0.14% and 0.80% respectively, reported by one study [48]. One study reported 0.00% coinfection of active TB with HIV, HBV and HCV (but based on only 9 diagnosed active TB cases [44]).

Risk factor analysis: study characteristics

While Figs. 2 and 3a-b suggested no trend in infection prevalence/testing yield over time for IGRA, active TB, HIV and HBV, there was a general reduction in HCV prevalence reported over time by median year of data collection (Fig. 3c). This relationship was supported by the meta-regression analysis, which showed a significant decline in HCV prevalence over time for univariable and multivariable analyses (p < 0.0001 and 0 = 0.0009 respectively, Table 1). Prevalence estimates for active TB showed a borderline significant increase over time in multivariable (p = 0.03) but not univariable (p > 0.10) analysis. The meta-regression suggested no other significant changes in infection prevalence over time. Lower study quality was significantly associated with higher HIV prevalence (multivariable analysis, low quality vs. high: meta-regression coefficient 0.19, 95%CI 0.09–0.29, p = 0.0001; moderate vs. high: 0.12, 95%CI 0.02–0.21, p = 0.02). Of the 12 HIV estimates, three were of moderate and three were of low quality (Table S3, Supplementary Material). Moderate/low quality studies tended to have smaller sample sizes, but risk factors for HIV infection (e.g., sexual health clinic attendance) did not appear to be more common among the low/moderate quality studies. Study quality was not associated with prevalence for any other infection, except for active TB, where lower quality was borderline significantly associated with higher prevalence in multivariable analysis (low quality vs. high: meta-regression coefficient 0.09, 95%CI 0.01–0.17, p = 0.04; moderate vs. high: 0.08, 95%CI 0.02–0.14, p = 0.01). Given the small number of included studies, the influence of outlier studies of low/moderate quality [46, 61], and the second lowest prevalence estimate being from one extremely large, high quality study [45], are likely to be responsible for this relationship. There were no other significant associations except again for active TB, where studies testing pre-entry or at the time of UK entry reported lower prevalence than studies of post-entry testing (multivariable analysis, pre- vs. post-entry: meta-regression coefficient − 0.04, 95%CI -0.06 to -0.01, p = 0.01; at entry vs. post-entry: -0.07, 95%CI -0.11 to -0.03, p < 0.001). Pre-entry testing studies tended to use less sensitive diagnostic methods such as symptom-based and/or chest X-ray.

Table 1.

Univariable and multivariable meta-regression analysis of factors potentially associated with prevalence of infection/test positivity. 95%CI – 95% confidence interval

	Univariable		Multivariable
	Coefficient (95%CI)	p value	Coefficient (95%CI)	p value
IGRA yield
Midpoint of study (year)	-0.01 (-0.02, 0.00)	0.05	-0.01 (-0.02, 0.01)	> 0.10
Study quality: low (vs. high)	0.05 (-0.10, 0.21)	> 0.10	0.13 (-0.02, 0.27)	0.092
Study quality: moderate (vs. high)	-0.01 (-0.14, 0.12)	> 0.10	-0.02 (-0.14, 0.11)	> 0.10
Pre-entry testing (vs. post UK arrival)	-	-	-	-
At entry testing (vs. post UK arrival)	-0.06 (-0.14, 0.01)	0.09	-0.06 (-0.17, 0.05)	> 0.10
Active TB prevalence
Midpoint of study (year)	0.00 (0.00, 0.00)	> 0.10	0.01 (0.00, 0.01)	0.03
Study quality: low (vs. high)	0.01 (-0.05, 0.07)	> 0.10	0.09 (0.01, 0.17)	0.04
Study quality: moderate (vs. high)	0.01 (-0.06, 0.08)	> 0.10	0.08 (0.02, 0.14)	0.01
Pre-entry testing (vs. post UK arrival)	-0.05 (-0.08, -0.02)	< 0.002	-0.04 (-0.06, -0.01)	0.01
At entry testing (vs. post UK arrival)	-0.04 (-0.07, -0.00)	0.03	-0.07 (-0.11, -0.03)	< 0.001
HIV prevalence
Midpoint of study (year)	0.00 (-0.01. 0.01)	> 0.10	0.00 (-0.01, 0.00)	> 0.10
Study quality: low (vs. high)	0.18 (0.13, 0.23)	< 0.0001	0.19 (0.09, 0.29)	0.0001
Study quality: moderate (vs. high)	0.13 (0.07, 0.19)	< 0.0001	0.12 (0.02, 0.21)	0.02
Pre-entry testing (vs. post UK arrival)	-0.06 (-0.26, 0.13)	> 0.10	0.00 (-0.10, 0.11)	> 0.10
At entry testing (vs. post UK arrival)	0.03 (-0.10, 0.16)	> 0.10	-0.02 (-0.11, 0.08)	> 0.10
HBV prevalence
Midpoint of study (year)	0.00 (-0.01, 0.01)	> 0.10	0.00 (-0.01, 0.01)	> 0.10
Study quality: low (vs. high)	0.06 (-0.11, 0.23)	> 0.10	0.01 (-0.20, 0.23)	> 0.10
Study quality: moderate (vs. high)	-0.01 (-0.09, 0.08)	> 0.10	-0.01 (-0.11, 0.09)	> 0.10
Pre-entry testing (vs. post UK arrival)	0.00 (-0.15, 0.14)	> 0.10	-0.01 (-0.17, 0.16)	> 0.10
At entry testing (vs. post UK arrival)	0.05 (-0.04, 0.15)	> 0.10	0.05 (-0.08, 0.19)	> 0.10
HCV prevalence
Midpoint of study (year)	-0.01 (-0.01, 0.00)	< 0.0001	-0.01 (-0.01, 0.00)	0.0009
Study quality: low (vs. high)	-0.03 (-0.15, 0.09)	> 0.10	0.01 (-0.10, 0.12)	> 0.10
Study quality: moderate (vs. high)	-0.05 (-0.12, 0.03)	> 0.10	-0.01 (-0.07, 0.05)	> 0.10
Pre-entry testing (vs. post UK arrival)	-0.01 (-0.09, 0.08)	> 0.10	0.00 (-0.04, 0.05)	> 0.10
At entry testing (vs. post UK arrival)	-0.03 (-0.11, 0.05)	> 0.10	0.00 (-0.06, 0.06)	> 0.10

Risk factor analysis: prevalence estimates stratified by risk factors

Tables summarising infection prevalence estimates stratified by risk factors reported in included studies can be found in Supplementary Material Table S5 (IGRA positivity), Table S6 (active TB infection), Table S7 (HIV infection), Tables S8 and S9 (HBV infection) and Tables S10 and S11 (current and past HCV infection, respectively). There was considerable heterogeneity in risk factors presented. Prevalence was stratified by sex/gender and age in four IGRA [13, 48, 49, 58], one active TB [56], one HIV [48] and three HBV/HCV studies [48, 55, 69]. Prevalence stratified by region/country of birth/origin or nationality was reported by three IGRA [13, 58, 59], two active TB [44, 56], one HIV [44], five HBV [44, 59, 69, 72, 73] and two HCV studies [44, 69].

Other risk factors explored were TB incidence in country of origin [13, 58], BCG vaccination status [13, 58], contact with a TB case [56, 58], time since UK arrival [58, 69] and alcohol use [58, 68], while the following risk factors were reported by just a single study: ethnicity [55], travel to a TB-endemic country [58], employment status [58], smoking status [58], history of imprisonment [58], TB prevalence in country of screening [56] and migrant type [57]. Additionally, Kelly et al. reported prevalence amongst migrants according to the following risk factors for HBV infection: blood transfusion, surgery and dental work abroad, vaccination and hepatitis vaccination (which were not further defined in the article), past history of jaundice, family history of liver diseases, body/ear piercing and recreational/illicit substance use (Table S8 [68]).

Figure S2 (Supplementary Material) presents a heat map summarising prevalence of infection/test positivity for each infection, stratified by sex/gender (as reported by each original publication) and age. There were no obvious patterns of differences in infection prevalence by sex, other than one study finding higher active TB prevalence among women [56] and men having higher HBV prevalence [55, 69]. Trends in infection prevalence by age are hard to compare across studies because of the different age bands used by different studies. Despite IGRA testing being recommended for patients only up to 35 years in the UK [74], studies did report prevalence estimates for older age groups, with prevalence being higher than for younger groups [48, 49, 58]. Active TB prevalence increased with age [56].

For the three studies stratifying prevalence by age, HBV prevalence appeared slightly higher in the middle age groups (30 years up to around 60–65 years) [48, 55, 69]. No overall pattern was apparent for HCV.

Figure S3 (Supplementary Material) illustrates the relationship between study sample size and the number of risk factors by which prevalence was stratified for the included study population. It suggests that there is no relationship between sample size and number of risk factors for infection presented by studies, and that the majority of studies attempted no risk factor analysis for the study population included in our analysis.

Figure 4 summarises prevalence of infection/test positivity stratified by country of birth/origin reported by studies, demonstrating that few studies attempted to stratify in this way. Fig. S4 (Supplementary Material) shows number of country of birth/origin-stratified prevalence estimates reported by included studies for the most common non-EU countries represented in the UK migrant population, alongside the infection burden in those countries of origin. Even for migrant groups from high-burden countries that make up a substantial share of UK migrants, country-specific prevalence estimates are often lacking. For example, no study reported Nigeria-specific infection prevalence estimates, despite Nigeria’s high infection burden and its position as the third most common non-EU country of origin among UK migrants. Care must be taken not to overinterpret Fig, 4 due to the small sample sizes of some of the strata. Nevertheless, it illustrates trends by region of birth/origin such as high levels of active TB and TB exposure among migrants from Eastern Africa and South-eastern Asia, and low prevalence in Western Africa. Southern Asian country of birth/origin estimates were generally high for HCV (three of five countries having high prevalence estimates), while Eastern, Middle and Northern African countries of birth/origin included many high prevalence levels for all infections.

Fig. 4.

Heat map summarising risk of infection/test positivity for IGRA test yield, active TB, HIV, HBV and HCV infection, stratified by country of birth/origin reported by included studies. Prevalence is shown as % (number infected or testing positive over total screened) for IGRA test yield and HIV, HBV and HCV infection, and prevalence per 100,000 (number infected or testing positive over total screened) for active TB infection. Countries are arranged by United Nations geoscheme. Multiple estimates in one cell represent findings from independent studies. Further details of included studies [44, 56, 59, 69, 72, 73] are shown in Supplementary Material Tables S5-10, which also includes prevalence and test yield estimates for other risk factors. Colour scheme represents the categorisation described in the Methods section. HIV: red = extremely high prevalence (> 0.5%), orange = high (0.2–0.5%), green = low (< 0.2%) [34, 35]. HBV: red = high prevalence (≥ 5%), orange = intermediate (2-4.9%), green = low (< 2%) [36]. HCV: red = high prevalence (> 0.36%), orange = intermediate (> 0.18–0.36%), green = low (≤ 0.18%). IGRA: red = high yield (> 15%), orange = intermediate (> 7.5–15%), green = low (≤ 7.5%). Active TB: red = extremely high prevalence (≥ 150 per 100,000 population), orange = high (40–149 per 100,000), green = low (< 40 per 100,000) [40]. Multiple estimates within one cell have not been pooled because of their heterogeneity. However, simple pooling of estimates was conducted to determine colour categories for these cells

Figure S5 (Supplementary Material) shows infection prevalence estimates for HIV, HBV and HCV, comparing prevalence among migrants with prevalence in their country of origin. The small samples for each country of origin stratum have resulted in wide 95%CI and uncertainty bounds for country prevalence are not shown, making interpretation difficult. However, it is notable that migrant prevalence estimates show no consistent pattern of being above or below prevalence in country of origin.

HIV prevalence estimates were higher for migrants from African regions than elsewhere but, as for TB/IGRA estimates, all strata were informed by a single study only [44]. For viral hepatitis, multiple studies presented several strata (two or three) for prevalence estimates. HBV and HCV estimates were heterogeneous between countries from the same region, particularly Southern Asia and Eastern Africa. HBV prevalence was higher among migrants from North, Middle and Eastern Africa (4.78–8.77% Ethiopia [44, 59], 12.50% South Sudan [44], 5.82–7.41% Sudan (44, 59), 5.81% DRC [44]) and Afghanistan (1.75–7.85% [44, 59]). Small sample sizes for some HCV strata restrict interpretation, but HCV prevalence was highest among migrants from Southern Asia, particularly Pakistan (0.47–2.73% [55, 69]) and Afghanistan (1.75% [44]).

Discussion

We have systematically reviewed and summarised evidence on the prevalence of IGRA positivity and prevalence of active TB, HIV, HBV and HCV, in addition to risk factor-stratified prevalence estimates for migrants to the UK. Our review highlights the overall high levels of IGRA positivity and prevalence of all infections other than HCV, where prevalence amongst migrants is low and appears to be declining over time, consistent with observed decreases in chronic hepatitis C prevalence in the UK following expanded testing and access to curative direct-acting antiviral treatment, with chronic HCV prevalence in England having fallen by 57% between 2015 and 2023 [75]. IGRA positivity and HBV prevalence were particularly high. It also illustrates high levels of heterogeneity between studies, including a wide variety of migrant types who will have different patterns of risk, from UASC to specific subcategories of migrants, such as those attending sexual health services, and approaches to testing (including timing, eligibility and venue), all of which limit comparisons and generalisations. This could be explained by substantial variations in study populations and in methods of recruitment. While IGRA positivity estimates were relatively homogeneous, active TB prevalence showed great heterogeneity. For example, very low prevalence was reported for pre-entry migrant screening with diagnosis based on symptoms and chest X-ray [45], compared to UK-based testing, which may employ more sensitive diagnostic methods or a broader definition of active TB [48].

Although IGRA positivity was high in several studies, the corresponding rates of active TB were relatively low. This pattern is consistent with a substantial underlying burden of latent TB infection, with only a small proportion progressing to active disease. It is also important to note that IGRA is an imperfect test and is an indirect marker of Mycobacterium tuberculosis exposure rather than definitively indicating latent infection, so to maximise its positive predictive value, IGRA-based latent TB infection screening is reserved for those at sufficiently high risk of progressing to disease [76]. Uptake and completion of latent TB infection treatment among migrant populations may be limited by barriers such as access to care, fear of immigration consequences, language challenges, and competing social needs, although Baggaley et al. reported high acceptance of testing and treatment completion amongst migrants registering with primary care [48]. However, migrants less likely to register with a GP will be missed by such testing approaches and consequently, many individuals with latent TB infection remain untreated and at persistent risk of reactivation over time. Understanding these gaps in the latent TB infection care cascade is essential when interpreting IGRA prevalence and considering its implications for TB prevention strategies.

Studies are likely to have reported on a limited number of risk factors because of restricted sample sizes. However, the choice of risk factors varied widely and there were inconsistent category boundaries (for example, age group cut-offs), making comparisons difficult.

There were also gaps in terms of reporting infection prevalence for migrants from some countries with high numbers living in the UK, and which have high infection prevalence in their country of origin. We would have expected to see reporting of more risk factors with increasing study sample size, but that did not appear to be the case. The risk factor analysis revealed that very few studies explored differences in prevalence beyond basic demographic factors (age, sex, ethnicity, country of origin/birth), such as reporting of risk behaviours (e.g., for BBVs, injecting drug use, various sexual risk behaviours). This is likely due to limitations in data collection, particularly for studies analysing routine data. Reporting prevalence stratified by additional risk factors would help identify which migrant groups are at greatest risk of infection, informing targeted interventions such as raising testing awareness among higher-risk groups.

The prevalence estimates we have compiled in this review each offer a snapshot of prevalence amongst migrants. However, migrants are a highly heterogeneous, dynamic population; countries of origin of UK-bound migrants change over time, and prevalence of infection in these countries may also change. Over the past two decades, the demographic profile of migrants to the UK has shifted substantially. Following EU enlargement in 2004, migration was dominated by arrivals from Central and Eastern Europe (e.g., Poland, Romania, Lithuania), where the prevalence of TB, hepatitis B, hepatitis C, and HIV is comparatively low. However, from the mid-2010s onwards, migration increasingly originated from South Asia (particularly India, Pakistan and Bangladesh), sub-Saharan Africa (notably Nigeria, Ghana, Somalia, and Eritrea) and the Middle East. These regions have considerably higher burdens of TB and viral hepatitis, and in some settings higher HIV prevalence as well. As a result, prevalence estimates reported in studies spanning 2004–2025 may partly reflect these changing migrant origins rather than genuine temporal changes in infection risk. The considerable heterogeneity between studies may have masked any changes in prevalence over time, with the exception of decreasing HCV prevalence.

These dynamics mean there is a need for continual monitoring of key infections such as TB and BBVs among this and other higher risk populations, as circumstances potentially change. In addition, surveillance should encompass the prevalence of other infections that disproportionately affect migrant populations, including helminthiases, together with the proportion of infections that exhibit drug resistance, to inform appropriate public health responses. More generally, the disruption to health services arising from drastic and abrupt cuts in global health funding threatens to reverse progress with infectious disease prevention programmes [77, 78], so increases in prevalence are possible. Furthermore, an HCV modelling analysis among another BBV high-risk population (people who inject drugs) has shown the importance of maintaining test and treat provision as we approach the elimination threshold, to prevent a U-turn in prevalence [79].

A multifaceted approach, providing multiple opportunities to test, is likely to be more successful than testing and treating these infections in silos, and has been shown to be acceptable to migrant populations when registering with primary care [48, 80]. The COMBAT-ID study reported comparatively little coinfection (0.06% tested HIV-HBV co-positive, 0.14% HIV-IGRA co-positive and 0.80% HBV-IGRA co-positive, with no coinfections involving HCV). BBV prevalence was substantial amongst those IGRA-negative and actually higher for all three BBVs for those not IGRA tested compared to those tested (IGRA testing is recommended only for patients up to 35 years in the UK [74]). HIV, HBV and HCV prevalence were 0.70%, 3.99% and 0.00% among IGRA-positive patients respectively, compared to 0.12%, 1.94% and 0.06% among IGRA-negative patients and 0.85%, 4.86% and 0.37% amongst patients not IGRA tested [48]. This suggests that a multiple infection rather than sequential approach to IGRA and BBV infection testing is appropriate, so BBV infections are not missed.

There are some limitations to our analysis. We may have missed unpublished studies and many prevalence estimates stratified by risk factors could be derived by re-analysing existing datasets, but that was beyond the scope of the current analysis. We included prevalence estimates from up to 21 years ago, but older studies may not reflect the situation now, given the dynamic, heterogeneous composition of migrants to the UK and changes in prevalence of infection and risk factors for infection acquisition in migrants’ countries of origin and along migration routes to the UK. Similarly, although we could have included infection prevalence estimates for migrants to other high-income countries, substantial differences in migrant population composition, such as variation in countries of origin, even among Western European destinations, mean that prevalence estimates from one country cannot be generalised to others without verification that the migrant populations are comparable.

Given the small number of studies available for certain outcomes, the meta-regression findings should be considered hypothesis-generating rather than confirmatory. As meta-regression was conducted using study-level data, observed associations represent ecological relationships and should not be interpreted as individual-level or causal effects. The categorisation of IGRA positivity and HCV prevalence into qualitative descriptors (e.g., low, intermediate and high HCV prevalence levels) was author-defined for descriptive purposes in the absence of universally accepted thresholds, and these categories should not be interpreted as validated clinical or policy cut-offs. Our risk factor analysis was limited by small sample sizes in many cases; many papers reporting relevant data had different research aims, meaning data were stratified in a different way which could not fulfil our inclusion criteria. Further research may include revisiting these datasets to analyse risk factors using standardised categories, data access permitting.

Conclusions

While prevalence estimates for each infection showed high heterogeneity, to be expected given the wide variation in study designs and study populations, prevalence of these infections, particularly IGRA yield and HBV, is high. This underscores the need to maintain effective monitoring, testing and treatment for key infections among migrant populations. Further work is needed to determine which testing strategies are most cost-effective, acceptable to migrants, and appropriately targeted by risk of infection, particularly in the context of a rapidly changing epidemiological and demographic landscape. To generate this evidence, more detailed reporting of infection prevalence by migrant subgroup and relevant risk factors is needed.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

95%CI

95% confidence interval

BBV

Blood-borne virus

DBS

Neonatal dried blood spot testing

E

East

ED

Emergency Department

ESOL

English for Speaker of Other Languages course

exc

excluding

FSW

Female sex worker

GP

General practice/primary care

GUM

Genitourinary medicine or sexual health clinic

HBV

Hepatitis B virus

HCV

hepatitis C virus

HIV

Human immunodeficiency virus

HPU

Health Protection Unit

IAC

Initial Accommodation Centre for UK asylum seekers

IGRA

Interferon gamma release assay

IOM

International Organization for Migration

IQR

Interquartile range

MeSH

Medical Subject Heading

MSW

Male sex worker

NICE

National Institute for Health and Care Excellence

NS

Not stated

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

SSA

Sub Saharan Africa

TB

Tuberculosis

UASC

Unaccompanied asylum seeking children

UKHSA

United Kingdom Health Security Agency

y

years

Author contributions

RFB conceived the study and search strategy. RFB, LS, ZL and PB performed database searches. RFB, CMH, LS, ZL and PB performed data extraction and critical appraisal of studies. RFB devised and carried out the current analysis and wrote the first draft of the manuscript with input from CMH, LS, ZL, PB and MP. All authors reviewed the manuscript, revised it critically for important intellectual content, and approved the final version prior to submission. The corresponding author, RFB, attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.

Funding

This work was supported by the National Institute for Health and Care Research (NIHR) [RFB: Advanced Fellowship NIHR302494; RFB and MP: Leicester NIHR Biomedical Research Centre (BRC); MP: NIHR Applied Research Collaboration East Midlands (ARC EM)]. The views expressed in this publication are those of the authors and not necessarily those of the NIHR or the Department of Health and Social Care.

Data availability

All data generated or analysed during this study are included in this published article and its supplementary information files.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

MP has received grants and personal fees from Gilead Sciences and personal fees from QIAGEN, outside the submitted work. All other authors report no conflicts of interest.