Antiretroviral Therapy (ART) has transformed the landscape of HIV treatment, significantly reducing mortality and improving the quality of life for people living with HIV/AIDS (PLWHA) worldwide. The introduction of ART marked a pivotal moment in the global fight against HIV, offering hope not only for individual survival but also for broader public health goals, such as viral suppression and epidemic control.1,2 However, the full potential of ART can only be realized if individuals are consistently retained in treatment.3–5 Retention in care ensures long-term viral suppression, prevents disease progression, and reduces transmission, yet achieving high retention rates remains a persistent challenge across various regions and demographic groups.6–11

In low- and middle-income countries, where the burden of HIV is disproportionately high, retention to ART is critically influenced by socioeconomic, demographic, and systemic factors. A 2016 meta-analysis estimated retention rates at 12-, 24-, and 36-months in these countries, revealing stark regional disparities.12 While these trends highlight the importance of sustained ART engagement, country-specific analyses are essential to uncover local barriers and tailor effective interventions, particularly in settings like Brazil, which has a longstanding universal ART policy but still faces unique implementation challenges.

Although ART retention in Brazil has been reported in the past by Ministry of Health, it has not fully captured the nuances of treatment retention over time, particularly considering the growing adoption of multi-month dispensing strategies.13 Furthermore, limited research has explored how ART retention varies by demographic factors such as age, sex, race, education, and region, leaving significant gaps in our understanding of retention dynamics in the country.

Understanding trends and disparities in retention to Antiretroviral Therapy (ART) across different sociodemographic groups is crucial for informing public health strategies. The findings of this study can support the Brazilian Ministry of Health and other stakeholders in designing and implementing more effective, equity-focused interventions. By identifying populations at greater risk of disengagement from care, such as specific age groups, racial/ethnic communities, or geographic regions, targeted policies and tailored retention strategies can be developed to improve outcomes and advance progress toward national and global HIV targets. Therefore, this study aims to analyze the trends in ART retention from 2014 to 2022 in Brazil, addressing gaps in current national data.

A total of 3404,469 ARV dispensations were analyzed, with similar distribution among the different cut-offs of time from ART initiation analyzed: 6-months (18 %), 12-, 24- and 36-months (17 % each), 48-months (16 %) and 60-months (15 %). Most of the dispensations analyzed were made to male (68 %), black individuals (43 %), with a predominance of people aged 30- to 49-years (54 %). Regarding the region, 42 % of the ARV dispensations analyzed occurred in the Southeast region, followed by the Northeast (21 %) and South (20 %) regions. In addition, 23 % of ARV dispensations were done to individuals with 8- to 11-years of schooling, followed by 0- to 7-years (22 %) and 12 and more years (17 %). The proportion of missing data was <1 % for all variables, except for race/color (19 %) and education (39 %). The characteristics of PLWHA whose ARV dispensations were analyzed in our study are presented in Table 1.

The proportion of PLWHA retained to ART have progressively decreased from 6- to 60-months after ART initiation in all years analyzed, reaching 87 % for 6-months versus 71 % for 60-months in 2022. At the same time, an increase was observed from 2014 to 2018 (81 % to 87 % for 6-months vs. 64 % to 69 % for 60-months), with relative stabilization from 2019 to 2022, as shown in Fig. 1.

Fig. 1.

Proportion of people retained on ART, according to time of treatment initiation. Brazil, 2014 to 2022.

Higher proportion of individuals retained on ART were observed among men, ranging from 75 % in 2014 to 79 % in 2022, while in women it was 67 % and 73 %, respectively (Fig. 2).

Fig. 2.

Proportion of people retained on ART, according to sex. Brazil, 2014 to 2022.

Between 2014 and 2018, there was an increase in the proportion of people retained on ART in the 15 to 17 (60 % to 74 %) and in the 18 to 24 age group (64 % to 76 %). The proportion of individuals retained remained around 75 % in the 25 to 29-years, 30- to 49-years, and over 50-years age groups throughout the period (Fig. 3).

Fig. 3.

Proportion of people retained on ART, according to age. Brazil, 2014 to 2022.

Important differences were observed in relation to race/color, with lower retention levels observed in the indigenous (from 51 % in 2014 to 67 % in 2022) and black population (from 69 % in 2014 to 76 % in 2022), when compared to the white/yellow population (from 76 % in 2014 to 81 % in 2022), throughout the analyzed period (Fig. 4).

Fig. 4.

Proportion of people retained on ART, according to race/color. Brazil, 2014 to 2022.

Individuals with higher levels of education presented higher levels of retention throughout the period, with 86 % of people retained on ART among those with 12-years of schooling or more, versus 72 % among those with 0- to 7-years of schooling, in 2022. (Fig. 5).

Fig. 5.

Proportion of people retained on ART, by years of study. Brazil, 2014 to 2022.

Some regional differences were observed, with the Southeast, South and Mid-West regions presenting the highest proportions of PLHIV retained to ART (78 %, 79 % and 78 % respectively, in 2022) followed by the Northeast and North regions (74 % and 75 % respectively, in 2022), as shown in Fig. 6.

Fig. 6.

Proportion of people retained on ART, by Region. Brazil, 2014 to 2022.

Our findings revealed a significant and progressive decrease in ART retention over time, reaching concerning levels at 60-months of treatment. This pattern is consistent with international evidence. A systematic review and meta-analysis conducted in low- and middle-income countries demonstrated a similar decline in ART retention over time ‒ 78 %, 71 %, and 69 % at 12-, 24-, and 36-months, respectively ‒ corroborating observations from other studies that maintaining high levels of ART retention over time remains a challenge in many countries.7,12,14,15

The persistently low levels of retention at 60-months observed throughout the study period is a matter of concern, given the chronic nature of HIV infection and the requirement for lifelong treatment. Suboptimal retention increases the risk of virologic failure, drug resistance, and HIV-related morbidity and mortality.6–9

This study observed a relative stability of retention rates from 2018 to 2022, while many countries experienced a negative impact of the SARS-CoV-2 pandemic across HIV continuum of care cascade, including retention, adherence, and viral suppression.16,17 In Brazil, this stability may be explained by strategies implemented by the Ministry of Health aimed at mitigating possible negative impacts of the pandemic on HIV care, such as the expansion of multi-month ARV dispensation and the implementation and expansion of telemedicine for the ARV prescription and follow-up of PLWHA.18–20 The introduction of integrase inhibitor-based regimens in previous years, with fewer side effects and improved tolerability, may also have contributed to the maintenance of retention during this period.21–23

We also found that men had higher retention rates than women. This finding aligns with studies conducted in Indonesia and Canada, in which women demonstrated lower retention and adherence, potentially due to factors such as social stigma, gender-based barriers to care, lack of family or partner support, and mental health challenges.24–26

Some studies have identified lower levels of adherence to ART among women, which may be related to factors such as marital approval, stigmatization, and mental health issues.26 However, other studies have observed that men were more likely to be lost to follow-up, indicating that gender disparities in ART retention may be context-specific and require further investigation.27–29

Although progress has been made over time, lower retention levels were observed among the population aged 15- to 17-years, particularly between 2014 and 2016. This finding is consistent with other national and international studies, which have highlighted the unique challenges faced by adolescents in adhering to lifelong treatment regimens.30–35

Racial and ethnic disparities were also evident in our study. Black and Indigenous population had lower retention rates compared with White and Asian populations. This pattern echoes previous findings in Brazil, where these populations are more likely to experience late HIV diagnosis, poor retention, and unsuppressed viral load.36 Structural barriers and inequities in access to healthcare services may contribute to these outcomes.33,37,38

Geographical disparities were also observed, with the North and Northeast regions presenting the lowest retention rates. Despite Brazil’s long-standing policy of universal access to ART, services remain unevenly distributed, with a greater concentration in urban centers, particularly in the South and Southeast regions.39–41

Education attainment emerged as an important factor for retention. Individuals with higher levels of education were more likely to remain in treatment, supporting existing evidence that lower education levels are associated with lower adherence rates and retention in ART, as well as higher levels of loss to follow-up.42,43

A methodological strength of this study was the development and application of an approach to assessing ART retention using national-level data from Brazil’s ART dispensation database. This method enabled us to define retention status at various time points after ART initiation while accounting for patients receiving multi-month dispensation. As an opportunity for improvement, the Brazilian Ministry of Health may revise its methods for analyzing ART retention based on the approach used in our study.

Nonetheless, some limitations should be noted. Retention was inferred from pharmacy refill data and may not fully capture true engagement in care, adherence, or virologic suppression. The descriptive nature of this study and the absence of formal statistical testing or trend modeling did not allow us to determine whether the observed changes in retention were attributable to specific programmatic interventions. The inclusion criterion of at least one dispensation may have captured individuals who initiated ART but never meaningfully engaged in care, potentially influencing early retention estimates. We were unable to exclude HIV/AIDS-related deaths during the study period. However, a previous population-based study conducted in Brazil showed that out of the 10 to 12 thousand annual cases of AIDS-related deaths in the country, almost half occurred among PLWHA who had not started ART.44 Considering the magnitude of the population analyzed in this study, no considerable impact on results is expected due to not excluding AIDS-related deaths from the analysis. Additionally, there were high levels of missing data for race/ethnicity and education, which may bias subgroup analyses and likely lead to underestimation of disparities; therefore, findings related to these variables should be interpreted with caution. This limitation underscores the importance of strengthening the completeness and quality of health information systems, as accurate and comprehensive sociodemographic data are essential to adequately characterize demographic and clinical profiles, identify inequities, and inform more targeted and effective public health and clinical interventions. Although plausible, the interpretations proposed to explain the observed trends ‒ such as the impact of pandemic mitigation strategies and the introduction of ARV regimens ‒ remain speculative and cannot be formally tested within the design of the present study.

We also used aggregated data and assessed ART retention at fixed time points, which limited our ability to evaluate continuous retention dynamics. As a study based on secondary data, our findings raise important hypotheses but cannot establish causality. Further research is needed to better understand the factors driving retention in the Brazilian context.

By introducing an innovative methodology based on delays in ART dispensation and applying it to a national database, this population-based study contributes new evidence on ART retention trends at multiple time points after ART initiation in Brazil. Our findings build upon previous research on ART adherence and retention, emphasizing the challenges of maintaining long-term ART engagement even in a setting with universal access. Importantly, they reveal persistent disparities in retention among vulnerable populations, underscoring the need for targeted and equitable public health interventions to sustain engagement in HIV care.

This was the first study to analyze ART retention trends at multiple time points using national-level data in Brazil. We observed a gradual decline in retention over time, with marked disparities across sociodemographic groups. Although retention remained relatively stable during the COVID-19 pandemic, long-term retention remains a key challenge. Our methodology offers a valuable tool for routine monitoring of ART retention and reinforces the need for equity-focused strategies to improve retention. While this study provides important population-level insights, longitudinal cohort studies are needed to assess individual care journeys and better understand the factors influencing sustained engagement in HIV care. Qualitative research may help uncover structural and psychosocial barriers to retention. Expanding differentiated care, strengthening community support, and addressing social determinants of health are critical to sustaining ART engagement and advancing HIV epidemic control.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

The data that support the findings of this study are available from the corresponding author upon reasonable request.