Using figures from the 2001 census, together with data from the second National Survey of Sexual Attitudes and Lifestyles, it has been estimated that between 10,000 and 30,000 black and minority ethnic (BME) men who have sex with men (MSM) currently live in Britain.1–3 In 2001, MSM and BME groups were identified as those bearing a disproportionately high burden of sexual ill health and were made the focus of national prevention efforts in the United Kingdom.4 HIV infection is known to disproportionately affect ethnic minorities and MSM. HIV prevalence among BME MSM in England and Wales and the United States is considerably higher than among white MSM.5–7
A large body of research has documented ethnic disparities in access to and quality of health care in developed countries. In the United States inequalities seem to be consistent across a range of outcomes and health care providers and maybe explained by the model of health care provision. A study identified health insurance status as being a key determinant of these disparities with BME groups less likely to be covered and likely to have less comprehensive coverage among those who are covered.7 In other European countries, poorer virological outcomes to combination antiretroviral therapy (cART) have been reported in immigrant populations.8,9 However, in the United Kingdom, the National Health Service provides universal and equitable access to care.10,11 Despite this, there are suggestions that the health care experience of ethnic minorities is poorer with higher levels of dissatisfaction with care,12 longer waits for appointments,13 fewer follow-up services,14 and poorer intermediate outcomes for those with diabetes.14
Ethnic disparities have been described in HIV-positive populations. African Americans have been shown to be persistently less likely to receive cART and to have poorer outcomes.14–17 In the United Kingdom, HIV-positive Africans access HIV services at a later stage of disease than non-Africans.18 Late diagnosis of HIV is more common among BME MSM compared with white MSM.19 In addition, MSM have been shown to have poorer health outcomes when compared with their heterosexual counterparts. As well as undergoing well-documented epidemics of HIV and sexually transmitted infections, they have been shown to have poorer mental health outcomes with increased rates of depression, suicide attempts, drug and alcohol abuse, and eating disorders.20–22
Although HIV-positive MSM and ethnic minority individuals have been studied separately, there remains a paucity of comparative data on the experience of BME and white MSM attending for HIV care in the United Kingdom. We investigated differences in retention in HIV care and uptake of cART and treatment outcomes between different ethnic MSM groups.
Patients were selected from the UK Collaborative HIV Cohort (CHIC) Study, a collaboration of some of the largest HIV clinics in the United Kingdom.23 Participating clinics provide routinely collected data on all patients aged ≥16 years attending for care since 1996, including demographic information, AIDS events, deaths, antiretroviral use, CD4 count, and HIV viral load (VL); the dataset utilized for the present analysis comprises 35,377 patients seen at 12 clinical centers to the end of 2008. Eligible participants were all MSM with known ethnicity seen at least once between January 1, 1996, and December 31, 2008. BME were categorized as follows: black Caribbean; black African, black other; Indian/Pakistani/Bangladeshi; other Asian/Oriental; and other/mixed. Participants whose ethnicity was coded as “other” were excluded.
We identified patients who were engaged into care (defined as having >1 day of follow-up) and compared the characteristics of patients in the different ethnic groups using χ2 tests, Mann–Whitney U tests, and Kruskal–Wallis tests, as appropriate. For these analyses, we considered age at first visit, the first available CD4 count and VL, whether the patient presented with advanced HIV disease (presentation with a CD4 count <200 cells/mm3 or a diagnosis of AIDS within the first 3 months after HIV diagnosis), the year of study entry, duration of follow-up, hepatitis C virus (HCV) and hepatitis B virus (HBV) coinfection. Additional outcomes considered were temporary loss-to-follow-up (>1 year between consecutive CD4 count measurements) and permanent loss-to-follow-up [last visit more than 1 year before the study end date (January 1, 2009) or death]. Of note, “permanent loss-to-follow-up” does not necessarily reflect loss to clinical care, as some patients may have transferred to a clinic that does not participate in UK CHIC. The characteristics of BME and white MSM at the time of starting cART were compared using χ2 test and Mann–Whitney U tests as above; multivariable logistic regression was used to assess the association between ethnic group and differences in initial cART regimen, after adjusting for potential confounders.
Logistic regression was used to investigate whether BME status was associated with cART initiation within the 6-month period after each CD4 count assessment, using methods previously described.24 For these analyses, we only included antiretroviral-naive patients under follow-up from January 1, 2000, onwards as by this point cART use had become widespread. cART was defined as any regimen including 1 of a protease inhibitor, nonnucleoside reverse transcriptase inhibitor, abacavir, CCR5, or entry inhibitor, with no restriction on the number of drugs in the combination. Each period of follow-up started on the date of a new CD4 measurement and ended at the initiation of cART, the date of the next CD4 count measurement, or 6 months after the measurement, whichever occurred first. Any CD4 count measured within 30 days of a previous one was considered to have been measured for confirmatory purposes. Where this occurred, the value used in analyses was the average of the 2 values; a subsequent measurement within the same 30-day period was considered to be a “new” measurement (on the basis that clinicians are unlikely to take more than 1 confirmatory measurement) and was retained in the dataset. Logistic regression models were fitted using the general linear models procedure in SAS (v9.13) with generalized estimating equations to allow for the multiple measurements from each subject. Potential confounding factors considered in these analyses were age, HIV VL, CD4 count, years from first entering HIV care, calendar year, previous AIDS, and clinical center. This analysis was also repeated using the finer ethnic groupings.
The following post-cART outcomes were considered among the subgroup of patients who had initiated cART before December 31, 2007 (to allow all patients the opportunity for at least 1 year of follow-up after initiating cART) as follows: (1) time to initial virological suppression <50 copies per milliliter (initial virological response); (2) HIV RNA level and CD4 cell count change from pre-cART levels at 12 months after starting cART; and (3) the development of a new AIDS event/death during the first year. For analyses of time to virological suppression, patient follow-up was right censored on the date of the last clinic visit or at the end of the first year, if the patient had not experienced virological suppression by this time. Virological suppression rates in white and BME MSM were compared using Cox proportional hazards regression with adjustment for the covariates described earlier. Analyses of clinical outcomes in the first year after starting cART were performed using a similar approach. The nearest HIV RNA level and CD4 count to 12 months after starting cART was used for analyses of outcome (2) as long as these had been measured within 3 months of the time of interest; the values were compared in the 2 groups using logistic regression (HIV RNA ≤50 copies/mL) or multiple linear regression (CD4 count).
Overall, after excluding 1513 MSM with unknown or “other” ethnicity, we identified 16,428 MSM with known ethnicity in the UK CHIC database of whom 1883 were BME. Of these patients, 404 (2.5%) did not engage in care [338 (2.3%) of the white MSM, and 66 (3.5%) of the BME MSM, P = 0.002, χ2 test] and were excluded from subsequent analyses. Patients with unknown or “other” ethnicities generally entered the cohort at an earlier stage (41.6% entering in 1996 compared with 32.8% of those with known ethnicity), but differences between the 2 groups in terms of total duration of follow-up (medians of 5.9 vs. 6.0 years in those with missing/“other” and known ethnicities, respectively), initial CD4 counts (380 vs. 393 cells/mm3, respectively) and initial VLs (4.35 vs. 4.47 log10 copies/mL) were small and of limited clinical relevance.
Among the 16,024 MSM who did engage in care, 1817 (11.3%) were BME with the remaining 14,207 (88.7%) of white ethnicity. Selected characteristics of this cohort, overall and stratified by ethnic group, are shown in Table 1. Overall, compared with white MSM, BME MSM tended to be younger, had presented with a lower CD4 count but lower VL, were less likely to be coinfected with HCV but more likely to be co-infected with HBV. BME MSM had joined the study in more recent calendar years resulting in a shorter duration of follow-up compared with white MSM. Although temporary loss-to-follow-up was less common in BME MSM compared with white MSM, permanent loss-to-follow-up was more common in the former group.
Characteristics of Specific BME
Among the BME MSM, 384 (21.3%) were black Caribbean, 237 (13.0%) black African, 271 (14.9%) black other, 139 (7.6%) Indian/Pakistani/Bangladeshi, 254 (13.9%) other Asian/other, and 532 (29.2%) other/mixed. CD4 count at first presentation was 367, 340, 383, 425, 340, and 409 cells per cubic millimeter in black Caribbean, black African, black other, Indian/Pakistani/Bangladeshi, other Asian/other and other/mixed groups, respectively (P = 0.0001), with the proportion defined as presenting with advanced HIV disease differing significantly as a result. Follow-up was longest among the black ethnic groups (medians of 4.8–6.7 years) and shortest in those of Indian/Pakistani/Bangladeshi background (median 3.9 years), reflecting the differing epidemic patterns in the United Kingdom. Coinfection with HCV was highest among the other/mixed (12.4%) and Indian/Pakistani/Bangladeshi (10.1%) groups, and lowest in the black ethnic groups (5.9%–8.2%, (P = 0.001)), whereas trends for HBV coinfection were in the opposite direction. Although temporary loss-to-follow-up was most common in the black other (33.6%) and black Caribbean (31.1%) groups, permanent loss-to-follow-up was highest in the other/mixed (29.1%) group and lowest in the Indian/Pakistani/Bangladeshi group (14.4%).
Initiation of cART
In total, 6338 participants initiated cART for the first time from January 1, 2000, to December 31, 2008. Among the subgroup that initiated cART (Table 2), BME MSM were younger (medians of 34 vs. 37 years, P = 0001) and had a slightly lower CD4 count (206 vs. 224 cells/mm3, P = 0.001) but a lower HIV VL (4.8 vs. 5.0 log10 copies/mL, P = 0003) than white MSM at time of initiation. Consistent with being diagnosed more recently BME MSM tended to start cART in later calendar periods. There were no large differences in terms of the initial cART regimens received.
Delayed Initiation of cART
After removing confirmatory CD4 counts, there were 59,132 CD4 counts in the dataset amongst ART-naive people (51,645 from white MSM, 7487 from BME MSM). Overall, 5614 (9.5%) CD4 count measurements were followed by initiation of cART; 4978 (9.6%) of those in white MSM compared with 636 (8.5%) of those in BME MSM (Table 3). In unadjusted models, BME were 13% less likely to initiate cART after a CD4 count measurement [OR: 0.87 (0.79, 0.95), P = 0.003]. Within each CD4 strata, BME were slightly less likely to initiate cART after a CD4 count measurement; after adjusting for CD4 strata, patients from BME continued to be less likely to initiate cART after each CD4 count measurement [adjusted OR: 0.83 (0.75, 0.92), P = 0.0002]. The left-hand side of Table 3 shows the results from the final multivariable model after additional adjustment for age, VL, years since first accessing HIV care, calendar year, previous AIDS diagnosis, and the patient's clinical center at the time of the CD4 count measurement. BME continued to be less likely to initiate cART in this model [adjusted OR: 0.83 (0.75, 0.92), P = 0.0002]. When the analysis was repeated to include specific BME subgroups (right-hand side of Table 3), rates of initiation of cART were lower in all ethnic groups (with the exception of other Asian/oriental) than in white MSM.
Of the 6338 patients who initiated cART, 4589 did so before the cut-off of December 31, 2007, and were included in analyses of post-cART outcomes (Table 4). Overall, 4356 (94.9%) of patients achieved virological suppression (≤50 copies/mL) at some stage after starting cART; by 3, 6, 9, and 12 months after starting cART, 31.4%, 69.2%, 81.2%, and 85.6% of white MSM, respectively, and 34.7%, 72.5%, 80.8%, and 84.7% of BME MSM, respectively, had achieved virological suppression (P = 0.67, log-rank test). The median times to virological suppression were 4.2 and 3.9 months in white and BME MSM. Adjustment for potential confounders had no impact on the apparent lack of difference in outcome between white and BME MSM (Table 4). By 12 months after starting cART, 3247 of 3957 patients with available measurements (82.1%) had a VL ≤50 copies per milliliter; again, there were no differences between white and BME MSM in the proportion attaining this level (82.3% vs. 80.4%, respectively, P = 0.38, χ2 test) either before or after controlling for potential confounders or between the different ethnic groups. Similarly, among the 3860 individuals with a CD4 count measurement at 12 months, there was no evidence of a difference in the CD4 count increase from pre-cART levels between white [median (IQR) change: 186 (98–287) cells/mm3] and BME [182 (94–279) cells/mm3] MSM (P = 0.68). In total, 274 MSM (6.0%) experienced a new AIDS event (n = 243) or death (n = 49) in the first year, with rates being similar in white (212 events, 5.2%) and BME (31 events, 6.2%) MSM (P = 0.38, χ2 test).
Our study demonstrates important differences between uptake of HIV treatment services in MSM according to ethnic group, but similar treatment outcomes. When compared with white MSM, HIV-positive BME MSM were more likely to fail to engage in care after diagnosis and were also more likely to be permanently loss-to-follow-up after engaging in care. Reassuringly, these disparities, although statistically significant, were slight with more than 96% of both groups engaging in care after their HIV diagnosis and around a quarter being permanently lost-to-follow-up. We report marked heterogeneity between ethnicities across a number of parameters, including CD4 count at first presentation, presentation with advanced disease, and coinfection with HCV or HBV. Of those who did engage with care, MSM of all BME groups were significantly less likely to start cART at any given CD4 count compared with their white counterparts particularly those of black ethnicities. However, once on cART, there were no differences in virological, immunological, and clinical outcomes.
These findings are consistent with those from several cohorts in North America and Europe, which reported non-white ethnicity to be associated with higher rates of loss-to-follow-up.25,26 Gerver et al27 reported that black Caribbean and African MSM were more likely to be lost-to-follow-up than white MSM in a South London cohort.24 However, Elford et al28 reported no difference in uptake of cART according to ethnicity and sexual orientation. However, this was a self-reported questionnaire study and could have been confounded by recall bias.
Despite the widespread availability of free health care and cART in the United Kingdom, we found that BME MSM were less likely than white MSM to initiate cART. Given that country of birth, migration status, and transferring care to a non-UK CHIC clinic were not recorded, these differences maybe smaller than reported as patients may have migrated before cART initiation.
It has been demonstrated that BME and MSM populations are at higher risk of poorer health outcomes.14,29 More recent studies have demonstrated that utilization of primary health care and outcomes for hypertension, raised cholesterol, and diabetes were similar across ethnic groups in England.10 Reassuringly our study demonstrates that ethnic disparities amongst MSM in access to HIV services and outcomes, for which treatment is largely provided in hospital-based HIV treatment centers, seem to be minimal in the UK CHIC cohort.
Of note, the largest ethnic group was that of other/mixed ethnicity (29.2%). This group had the highest permanent loss-to-follow-up and HIV/HCV-coinfection rates. Further research is warranted into identifying these individuals as they seem to be at high risk of acquiring hepatitis C which may be a surrogate marker for high-risk sexual behavior.30 Ethnicity is based on a combination of self-report and clinician-report and may vary both between and within clinics. Our classification of ethnic groups is based upon the current classification used by the Office of National Statistics.31 However, the finding that just under a third of BME men were categorized as “Other non-white” (excluded from our analyses) suggests a lack of sophistication in the classification categories.
There are several limitations to our study. We do not collect information on the offer of cART, only on whether it was actually initiated; thus, it is possible that individuals who delayed cART had previously declined an offer of treatment. We also do not collect the reasons for starting or not starting cART, and so cannot investigate possible reasons for any delay in some ethnic groups. Our findings are only applicable to MSM populations. Ethnic variance in HIV care uptake and utilization may also be different among heterosexuals. Finally, the limitations of this study are those inherent to observational studies, including concerns surrounding unmeasured confounding and missing or inconsistent data.
This study demonstrates that despite BME MSM being a “minority within a minority” for those HIV infected there are few ethnic disparities in access to and treatment outcomes in our setting. This is reassuringly different from the Institute of Medicine review of ethnic inequalities in the US health care system. The findings from the analyses presented here suggest that the provision through the National Health Service of publicly funded HIV care with universal access has resulted in equitable utilization and outcomes of HIV care across different ethnic MSM groups. Nevertheless we have shown that there are disparities in uptake of cART amongst some ethnic groups and the reasons behind this warrant further study.
The authors thank Dr John White for reviewing the article.
1. Johnson AM, Mercer CH, Erens B, et al.. Sexual behaviour in Britain: partnerships, practices, and HIV risk behaviours. Lancet. 2001;358:1835–1842.
2. Mercer CH, Fenton KA, Copas AJ, et al.. Increasing prevalence of male homosexual partnerships and practices in Britain 1990-2000: evidence from national probability surveys. AIDS. 2004;18:1453–1458.
4. Department of Health. National Strategy for Sexual Health and HIV. London, United Kingdom:Department of Health; 2001.
5. Dougan S, Elford J, Rice B, et al.. Epidemiology of HIV among black and minority ethnic men who have sex with men in England and Wales. Sex Transm Infect. 2005;81:345–350.
7. Smedley BD, Stith AY, Nelson AR. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Wahington, DC:Institute of Medicine of the National Academies; 2003.
8. Dray-Spira R, Spire B, Heard I, et al.. Heterogeneous response to HAART across a diverse population of people living with HIV: results from the ANRS-EN12-VESPA Study. AIDS. 2007;21:S5–S12.
9. Nellen JF, Wit FW, De Wolf F, et al.. Virologic and immunologic response to highly active antiretroviral therapy in indigenous and nonindigenous HIV-1-infected patients in the Netherlands. J Acquir Immune Defic Syndr. 2004;36:943–950.
10. Nazroo JY, Falaschetti E, Pierce M, et al.. Ethnic inequalities in access to and outcomes of healthcare: analysis of the Health Survey for England. J Epidemiol Community Health. 2009;63:1022–1027.
12. Airey C, Bruster S, Erens B, et al.. National Surveys of NHS Patients: General Practice 1998. London, United Kingdom:NHS Executive; 1999.
13. Gillam SJ, Jarman B, White P, et al.. Ethnic differences in consultation rates in urban general practice. BMJ. 1989;299:953–957.
14. Millett C, Gray J, Saxena S, et al.. Ethnic disparities in diabetes management and pay-for-performance in the UK: The Wandsworth Prospective Diabetes Study. PLoS Med. 2007;4:1087–1093.
15. Gebo KA, Fleishman JA, Conviser R, et al.. Racial and gender disparities in receipt of Highly Active Antiretroviral Therapy persist in a multistate sample of HIV patients. J Acquir Immune Defic Syndr. 2005;38:96–103.
16. Yehia B, Fleishman J, Gebo K, et al.. Disparities in HAART receipt for HIV infected adults in care (2002–2008). Presented at: Conference on Retroviruses and Opportunistic Infections; Boston, MA; February 27-March 2, 2011.
17. Oster AM, Wiegand RE, Sionean C, et al.. Understanding disparities in HIV infection between black and white MSM in the United States. AIDS. 2011;25:1103–1112.
18. Sullivan AK, Curtis H, Sabin CA, et al.. Newly diagnosed HIV infections: review in UK and Ireland. BMJ. 2005;330:1301–1302.
20. Bolton SL, Sareen J. Sexual orientation and its relation to mental disorders and suicide attempts: findings from a nationally representative sample. Can J Psychiatry. 2011;56:35–43.
21. Chakraborty A, McManus S, Brugha TS, et al.. Mental health of the non-heterosexual population of England. Br J Psychiatry. 2011;198:143–148.
22. Feldman MB, Meyer IH. Eating disorders in diverse lesbian, gay, and bisexual populations. Int J Eat Disord. 2007;40:218–226.
23. The UK Collaborative HIV Cohort Steering Committee. The creation of a large UK-based multicentre cohort of HIV-infected individuals: The UK Collaborative HIV Cohort (UK CHIC) Study. HIV Med. 2004;5:115–124.
24. Stöhr W, Dunn DT, Porter K, et al.. CD4 cell count and initiation of antiretroviral therapy: trends in seven UK centres, 1997–2003. HIV Med. 2007;8:135–141.
25. Lanoy E, Mary-Krause M, Tattevin P, et al.. Predictors identified for losses to follow-up among HIV-seropositive patients. J Clin Epidemiol. 2006;59:829–835.
26. Cohen CJ, Iwane MK, Palensky JB, et al.. A national HIV community cohort: design, baseline, and follow-up of the AmFAR Observational Database. American Foundation for AIDS Research Community-Based Clinical Trials Network. J Clin Epidemiol. 1998;51:779–793.
27. Gerver SM, Chadborn TR, Ibrahim F, et al.. High rate of loss to clinical follow up among African HIV-infected patients attending a London clinic: a retrospective analysis of a clinical cohort. J Int AIDS Soc. 2010;13:29.
28. Elford J, Ibrahim F, Bukutu C, et al.. Uptake of antiretroviral treatment among people living with HIV in London: ethnicity, gender and sexual orientation. Sex Transm Infect. 2001;84:176–178.
29. Friedman MS, Marshal MP, Ron Stall R, et al.. Gay-related development, early abuse and adult health outcomes among gay males. AIDS Behav. 2008;12:891–902.
30. Turner J, Rider A, Imrie J, et al.. Behavioural predictors of subsequent hepatitis C diagnosis in a UK clinic sample of HIV positive men who have sex with men. Sex Transm Infect. 2006;82:298–300.
31. Ethnic Group Statistics. A guide for the collection and classification of ethnicity data. Available at: www.ons.gov.uk
. Accessed May 31, 2010.
UNITED KINGDOM COLLABORATIVE HIV COHORT STUDY GROUP
Writing Committee: Gulshan Sethi, Simon Edwards, Suneeta Soni, Richard Gilson, Jane Anderson, Adrian Palfreeman, Frank Post, Valerie Delpech, Martin Fisher, Teresa Hill, John Walsh, Brian Gazzard, Loveleen Bansi, Clifford Leen, David Dunn, Margaret Johnson, Andrew N. Phillips, Chloe Orkin, Mark Gompels, Jonathan Ainsworth, and Caroline A. Sabin for the UK Collaborative HIV Cohort (CHIC) Study.
Steering Committee: Jonathan Ainsworth, Jane Anderson, Abdel Babiker, Loveleen Bansi, David Chadwick, Valerie Delpech, David Dunn, Martin Fisher, Brian Gazzard, Richard Gilson, Mark Gompels, Teresa Hill, Margaret Johnson, Clifford Leen, Mark Nelson, Chloe Orkin, Adrian Palfreeman, Andrew Phillips, Deenan Pillay, Frank Post, Caroline Sabin (PI), Memory Sachikonye, Achim Schwenk, John Walsh.
Central Coordination: Royal Free National Health Service Trust and RFUCMS, London (Loveleen Bansi, Teresa Hill, Susie Huntington, Andrew Phillips, Caroline Sabin); Medical Research Council Clinical Trials Unit (MRC CTU), London (David Dunn, Adam Glabay).
Participating Centers: Barts and The London National Health Service Trust, London (C. Orkin, N. Garrett, J. Lynch, J. Hand, C. de Souza); Brighton and Sussex University Hospitals National Health Service Trust (M Fisher, N Perry, S Tilbury, D Churchill); Chelsea and Westminster Hospital National Health Service Trust, London (B Gazzard, M Nelson, M Waxman, D Asboe, S Mandalia); Health Protection Agency–Center for Infections London (HPA) (V Delpech); Homerton University Hospital National Health Service Trust, London (J Anderson, S Munshi); King's College Hospital National Health Service Foundation Trust, London (F Post, H Korat, C Taylor, Z Gleisner, F Ibrahim, L Campbell); Mortimer Market Center, London (R Gilson, N Brima, I Williams); North Middlesex University Hospital National Health Service Trust, London (A Schwenk, J Ainsworth, C Wood, S Miller); Royal Free National Health Service Trust and UCL Medical School, London (M. Johnson, M. Youle, F. Lampe, C. Smith, H. Grabowska, C. Chaloner, D. Puradiredja); St. Mary's Hospital, London (J. Walsh, J. Weber, F Ramzan, N Mackie, A Winston); The Lothian University Hospitals National Health Service Trust, Edinburgh (C. Leen, A. Wilson); North Bristol National Health Service Trust (M. Gompels, S. Allan); University of Leicester National Health Service Trust (A. Palfreeman, A. Moore); South Tees Hospitals National Health Service Foundation Trust (D. Chadwick, K. Wakeman).