People Born in Non–Main English Speaking Countries Are Less Likely to Start HIV Treatment Early in Australia: A National Cohort Analysis, 2014–15

To the Editors: INTRODUCTION In Australia, people born in Sub-Saharan Africa, South-East Asia, and North-East Asia have HIV diagnosis rates 2–4 times higher than people born in Australia,1 and are more likely to be diagnosed late with a CD4 count <350 cells/μL.1 Late presentation is associated with a significantly increased incidence of AIDS or death.2 Once diagnosed, overseas-born people living with HIV (PLHIV) can face numerous linguistic, cultural, and financial barriers to linkage and retention into HIV treatment and care.3 Prompt diagnosis and treatment of HIV is beneficial both for individuals and for reducing onward transmission in the community.4,5 Many studies in high-income countries have focused on late diagnosis among migrant PLHIV, but few on clinical pathways and outcomes after diagnosis with these populations. In Australia, there has only been one study comparing Australian-born HIV patients with those born overseas in high- or low/middle-income countries, which found no differences in routine monitoring of HIV infection, loss to follow-up or outcomes such as undetectable viral load,6 but used a patient cohort who were treatment-experienced rather than newly diagnosed, with 30% excluded because country of birth information was missing.6 A Canadian study also found foreign-born PLHIV were as likely as Canadian-born non-Aboriginal PLHIV on antiretroviral treatment (ART) to achieve virological suppression, but did not investigate differences between groups in the likelihood of being on treatment once diagnosed.7 Research to date in other high-income countries used patient data before 2014, before guidance was released internationally and in Australia recommending that ART be initiated in all adults living with HIV.8,9 As of November 2013, all Australian citizens and residents diagnosed with HIV can access subsidized ART through the Pharmaceutical Benefits Scheme (PBS) irrespective of CD4 count.10 The aim of this study—the first we are aware since universal ART guidelines were introduced—was to estimate the proportion of patients born in non–main English speaking countries and newly diagnosed with HIV in Australian sexual health clinics who had initiated treatment 6 months after diagnosis, compared with other patients. METHODS We used deidentified data from 43 publically funded sexual health services participating in the Australian Collaboration for Coordinated Enhanced Sentinel Surveillance of Blood Borne Viruses and Sexually Transmitted Infections (ACCESS) network in New South Wales, Western Australia, Victoria, Queensland, and the Northern Territory.11 All patients diagnosed with HIV between January 1, 2014 and June 30, 2015 were included and categorized as “culturally and linguistically diverse (CALD)” or “other.” CALD patients were those born in a country where English is not the main language spoken, in line with definitions used by government health and statistics bodies12,13 and scientific literature showing that CALD populations have specific health care needs.14 Other patients were born in Australia or a main English-speaking country, ie, United Kingdom, Republic of Ireland, New Zealand, Canada, and the United States.12 Although designated as main English-speaking, South African–born patients were included in the CALD category because evidence suggests that they can face similar barriers to other CALD patients in accessing HIV services after migration to a high-income country.15,16 Patients were followed up for a maximum of 183 days (6 months) after diagnosis and assessed as having commenced treatment early if they had recorded ART initiation before this time. Patients were excluded from the primary analysis if they were recorded in the treatment file as accessing HIV management at another clinic. Demographic and clinical variables were included in the analysis.17 Missing values of viral load at diagnosis were imputed using ordinal logistic regression. Data on year of arrival variable were available but collinear with CALD patient status and excluded from further analysis. The primary outcome was the proportion of patients in each group on treatment 6 months after diagnosis. The demographic profiles of CALD and other patients were compared using Pearson χ2 test for categorical variables and Wilcoxon rank-sum test for median values. We explored differences in time to treatment by CALD status using Kaplan–Meier curves and log-rank tests. Univariate and multivariate Cox proportional hazards models stratified on treatment centre were used to assess unadjusted and adjusted associations with being on treatment 6 months after diagnosis. A model was also developed adjusting for the competing risk of loss to follow-up, and sensitivity analyses conducted: (1) including men who have sex with men (MSM) only, (2) excluding patients with incomplete follow-up (more than 6 months between clinic visits), and (3) including patients managed at other clinics, with these patients assigned late treatment status unless otherwise indicated on file. Only variables significant at the 5% in the level in the univariate analysis were included in the multivariate model. All events were censored on December 31, 2015. Stata statistical software, version IC 14 (StataCorp, College Station, TX) was used for all analyses. Ethics The ACCESS Project received ethical approval from the human research ethics committees of Central Australia, St. Vincent's Hospital (Sydney), Cairns Hinterland Health Service District, Menzies School of Health Research, South Metropolitan Area Health Service District, Gold Coast Health Service District, the Alfred Hospital, the Princess Alexandra Hospital, and Townsville Health Service District. RESULTS Between January 1, 2014 and June 30, 2015, 290 patients were diagnosed with HIV at participating sexual health services—104 CALD and 186 other patients. CALD patients were younger, more likely to be female, reside in major cities, be heterosexual, and had a lower median CD4 count at diagnosis compared with other patients (Table 1).TABLE 1.: Association of Demographic and Clinical Factors With Being on Treatment 6 Months After HIV Diagnosis*†TABLE 1-A.: Association of Demographic and Clinical Factors With Being on Treatment 6 Months After HIV Diagnosis*†By 6 months follow-up, the Kaplan–Meier estimates of being on treatment were 44.2% (95% CI: 35.3 to 54.2) for CALD patients and 58.6% (95% CI: 51.6 to 65.7) for other patients (log-rank test P = 0.011). In the univariate Cox proportional hazards model, being a CALD patient, being MSM, and having a higher CD4 count at diagnosis were associated with a lower probability of having started treatment 6 months after diagnosis (Table 1). In the multivariate model, after adjusting for CD4 count at diagnosis and MSM status, CALD patients were still 47% less likely to have initiated treatment by 6 months after diagnosis (Table 1). When considering MSM only, CALD patients were 41% less likely to be on treatment at 6 months than other patients after adjusting for CD4 count (aHR 0.59, 95% CI: 0.38 to 0.93, P = 0.023). A model adjusting for competing risk of loss to follow-up (aHR 0.55, 95% CI: 0.37 to 0.82, P = 0.003) and sensitivity analyses excluding patients with incomplete follow-up (aHR 0.64, 95% CI: 0.43 to 0.95, P = 0.028) or including patients with recorded treatment at other clinics (aHR 0.52, 95% CI: 0.35 to 0.77, P = 0.001) produced similar results. DISCUSSION This is the first study we are aware of in a high-income country, after universal treatment recommendations, to consider differences in time between diagnosis and treatment initiation for HIV-positive persons born in a non–main English speaking country compared with persons born in Australia or another main English-speaking country. CALD patients were significantly less likely than other patients to have started ART at 6 months after diagnosis, irrespective of CD4 count. These findings are consistent with previous studies in high-income countries concerning ethnic differences in HIV treatment initiation.18–20 Income was an important correlate in some of these studies, but its' effect is likely to be attenuated in Australia, where sexual health services provide health care such as HIV testing to anyone free of charge and ART is subsidized for Australian citizens and permanent residents under the PBS. Some PLHIV in Australia, such as temporary visa holders, are ineligible for treatment subsidies; however, at least some are able to access ARTs through compassionate ACCESS schemes or state-based public health services including hospitals.21 Other factors causing delay in treatment initiation among HIV-positive CALD patients could include difficulty coming to terms with the diagnosis, compounded by the stigma which still surrounds HIV in some communities, and concern about what an HIV diagnosis might mean for their visa status.14,21 Linguistic barriers and a lack of access to interpreter services can make it more difficult for health care providers to communicate to migrant patients, particularly those with limited English, about the individual and community benefits of starting ART early. Finally, indirect costs for example around transport and low levels of knowledge about the Australian health care system can also be impediments.3,22 This study has a few limitations. One, the sample size was relatively low and may have reduced the ability to accurately detect differences between migrant and nonmigrant patients and adjust for factors such as age, sex, exposure, and region of birth. Two, our study focused on sexual health clinics and findings may not be generalizable to CALD patients who receive care in other settings. In summary, HIV-positive CALD patients at sexual health services are less likely than other patients to initiate treatment within 6 months of diagnosis. Reducing any financial barriers to access, increasing health literacy among CALD migrants, and health professionals' ability to communicate—in a culturally and linguistically appropriate way—the benefits of starting treatment early may be needed to reduce this gap.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.