Adolescents aged 10–19 years old living with HIV are a particularly vulnerable group because of developmental, social and healthcare-related challenges (Baltag, Guthold & Sawyer 2020; World Health Organization 2024). In 2023, over one million adolescents aged 15–19 years old were living with HIV globally, with approximately 840 000 in sub-Saharan Africa (SSA) (UNICEF 2024). South Africa bears a significant burden, with adolescent girls disproportionately affected because of biological and socio-economic vulnerabilities (Govender et al. 2022; Human Sciences Research Council 2023).

Despite expanded access to antiretroviral therapy (ART), adolescents living with HIV (ALHIV) persist in experiencing markedly inferior treatment outcomes than adults, a trend observed across multiple studies and settings (Enane, Vreeman & Foster 2018; Tarantino, Lowery & Brown 2021). Adolescents consistently demonstrate lower rates of viral load suppression (VLS) compared to adults, with some studies reporting that only 74% of adolescents achieve VLS (< 1000 copies/mL) and just 41% reach full VLS (< 50 copies/mL), falling short of the Joint United Nations Programme on HIV/AIDS (UNAIDS) 95-95-95 targets (UNAIDS 2021). Furthermore, while younger adolescents may initially show promising immunological responses after starting ART, they are more likely to experience virological failure and detectable viral loads after 12 months of treatment initiation (Ferrand et al. 2016; Han et al. 2021). Retention in care (RiC) also presents a substantial challenge, as ALHIV, particularly those aged 15–19 years old, are at higher risk of loss to follow-up (LTFU) compared to adults, further undermining treatment success (Han et al. 2021; Molopa et al. 2024). These disparities are driven by multiple factors, including suboptimal adherence linked to developmental challenges, stigma, and mental health issues, as well as gaps in adolescent-friendly service delivery (Cluver et al. 2018; Crowley, Van der Merwe & Skinner 2021; Mayman & Van Wyk 2022; Mushy et al. 2024; World Health Organization 2013).

For instance, clinics that do not provide tailored support or optimal ART regimens, such as dolutegravir (DTG)-based therapies, report higher rates of viral non-suppression among ALHIV (Mushy et al. 2024).

Programmatic interventions incorporating peer support, psychosocial care, and community-based case management have shown promise in providing both RiC and VLS among adolescents (Rakhmanina, Foster & Agwu 2024). Collectively, these findings highlight the urgent need for developmentally appropriate, comprehensive HIV care strategies that address the unique barriers faced by adolescents in order to close the treatment outcome gap with adults.

The coronavirus disease 2019 (COVID-19) pandemic, beginning in early 2020, exacerbated these challenges. Lockdowns and service disruptions threatened ART access and continuity, especially in resource-limited settings in Africa (Mayman & Van Wyk 2022; Pillay et al. 2021). Adolescents encountered new obstacles, including clinic closures, mobility restrictions, increased mental health stressors, and diminished social support, all of which could negatively impact ART adherence and RiC (Mayman & Van Wyk 2023; Pillay et al. 2021).

While numerous studies have investigated the impact of COVID-19 on general HIV service delivery, relatively few have specifically examined its effects on adolescents, and even fewer have utilised disaggregated data within local South African contexts (Cluver et al. 2020; Mayman & Van Wyk 2023). In the Western Cape, especially within the Khayelitsha and Eastern Substructure (KESS), HIV prevalence remains disproportionately high, compounded by pronounced socioeconomic vulnerabilities, including poverty, unemployment, and informal housing, that exacerbate barriers to healthcare access (Statistics South Africa 2011; Stinson et al. 2017). During the COVID-19 pandemic, adolescents in this region experienced notable disruptions in care, characterised by delayed clinic visits, diminished psychosocial support, and challenges in accessing medication (Mayman & Van Wyk 2022). Despite these obstacles, differentiated service delivery (DSD) models implemented locally, such as multi-month ART dispensing and youth-friendly clinics, contributed to sustaining some continuity of care (Hightow-Weidman et al. 2020; Maskew et al. 2022).

Nonetheless, there remains an urgent need for adolescent-centred, contextually responsive interventions that address the complex interplay between structural inequalities and treatment outcomes (Crowley et al. 2021; Hlophe, Van Wyk & Mukumbang 2023). Consequently, a thorough understanding of the pandemic’s impact on adolescent HIV care in high-burden, resource-limited settings such as KESS is critical to inform strategies for health system resilience and targeted policy development (Mayman & Van Wyk 2023; Van Wyk, Kriel & Mukumbang 2020).

This study aimed to compare treatment outcomes, including VLS, ART adherence, and RiC, among ALHIV before and during the COVID-19 pandemic. The authors hypothesised that the pandemic negatively impacted these outcomes.

Table 1 presents the comparison of sociodemographic, clinical and treatment characteristics of ALHIV on ART in the period pre-COVID-19 and during the COVID-19 pandemic. The cohort assessed during the COVID-19 pandemic exhibited a moderately higher proportion of females (57.0% before COVID-19 versus 66.3% during COVID-19) and a lower representation of adolescents aged 10–14 years old (32.5% versus 41.4%) relative to the pre-pandemic cohort. These demographic differences are detailed in Table 1. A marginal decrease was observed in the proportion of ALHIV with CD4 counts ≥ 350 cells/mm³ (30.8% during COVID-19 versus 31.1% pre-COVID-19). Notably, adherence to ART demonstrated a substantial decline from 96.4% before the pandemic to 70.0% during the COVID-19 period (p < 0.001). RiC similarly decreased from 80.3% to 76.3% (p < 0.001).

The recorded prevalence rates for VLS (defined as viral load less than 1000 copies/mL) – were reduced from 82.1% pre-COVID-19 to 64.8% during COVID-19; although not statistically significant (p = 0.065).

Furthermore, full VLS (defined as a viral load below 50 copies/mL) decreased from 70.8% to 53.7% (Table 1). Figure 1 and Figure 2 visually depict these trends, illustrating the decline in VLS rates before and during the pandemic, while also pointing out that a notable proportion of adolescents continued to achieve full VLS despite the challenges posed by the COVID-19 pandemic.

Multivariate regression analysis was conducted on identified factors independently associated with VLS during the COVID-19 period. The results of both crude odds ratio (COR) and AOR are summarised in Table 2. Higher CD4 counts were strongly predictive of VLS. This is demonstrated by the finding that adolescents with higher CD4 counts had significantly lower odds of viral non-suppression, with those in the ≥ 350 cells/mm³ category having a notably reduced odds (AOR = 0.140; 95% CI: 0.092–0.212) compared to adolescents with CD4 counts < 200 cells/mm³. Antiretroviral therapy regimen was also a significant predictor: individuals on second-line (AOR = 3.237; 95% CI: 2.346–4.467) and other regimens (AOR = 3.513; 95% CI: 2.372–5.202) were more likely to achieve VLS relative to those on first-line ART.

Optimal ART adherence emerged as a key determinant of VLS, with adherent adolescents demonstrating significantly higher odds of VLS (AOR = 0.321; 95% CI: 0.204–0.505). Although RiC was significantly associated with VLS in unadjusted analyses, this association was not statistically significant in the multivariate model (AOR = 1.148; 95% CI: 0.680–1.938).

It is important to note that the sample size for the regression analyses was reduced because of the exclusion of viral load and CD4 count results that either fell outside the defined COVID-19 period (01 March 2020–31 December 2021) or lacked valid test dates. This data cleaning step is detailed in the note beneath Table 2.

This study compared treatment outcomes of ALHIV receiving ART before and during the COVID-19 pandemic in KESS of Cape Town, South Africa. By analysing data from both periods, the authors aimed to assess the impact of the pandemic on key indicators, including ART adherence, RiC, and VLS. Our objective was to elucidate the extent to which pandemic-related disruptions influenced adolescent HIV care within this high-burden urban context.

The findings reveal a marked deterioration in measured treatment outcomes during the COVID-19 pandemic period, with ART adherence experiencing the most substantial decline. Specifically, ART adherence, as measured through pharmacy refill records, decreased from 96.4% pre-pandemic to 70.0% during the pandemic (p < 0.001). This significant reduction is consistent with existing literature that highlights the compounded challenges introduced by the pandemic, such as healthcare facility closures, restricted mobility because of lockdown regulations, limited transportation, and increased psychosocial stressors (Mayman & Van Wyk 2023; Pillay et al. 2021). Adolescents, already at heightened risk because of their developmental stage, may have faced particular difficulty maintaining consistent ART adherence under these adverse circumstances.

Retention in care also declined, albeit to a lesser extent, from 80.3% to 76.3% (p < 0.001). Although this reduction was less pronounced than that observed for ART adherence, even modest disruptions to care continuity are concerning for a population requiring lifelong ART. The observed decline likely reflects known barriers to sustained engagement in HIV care during the pandemic, including fears of COVID-19 infection, reduced clinic hours, and staff shortages (Jardim, Zamani & Akrami 2022; Mathamo et al. 2022).

Viral load suppression, a critical indicator of treatment efficacy, also worsened during the pandemic. Full VLS (< 50 copies/mL) decreased from 70.8% to 53.7%, and overall VLS (< 1000 copies/mL) fell from 82.1% to 64.8%. While the decline in full VLS did not reach statistical significance (p = 0.065), the 17.1% percentage-point reduction is clinically meaningful and indicates that fewer adolescents achieved optimal virological control during the pandemic period. Notably, these findings contrast with reports from other SSA settings where VLS remained stable, potentially because of more robust implementation of DSD models (Carpenter et al. 2024; Mugo et al. 2024). The findings from our study are displayed in Table 1; Figure 1 and Figure 2.

Multivariate regression analysis identified several factors associated with improved virological outcomes, which are likely protective factors against viral non-suppression. Higher latest CD4 counts were strongly predictive of VLS, as adolescents with CD4 counts ≥ 350 cells/mm³ had significantly lower odds of viral non-suppression (AOR = 0.140; 95% CI: 0.092–0.212) compared to those with CD4 counts < 200 cells/mm³. This aligns with findings from other studies in SSA, where higher immunological status has consistently been associated with better VLS (Ferrand et al. 2016; Han et al. 2021). Similarly, optimal ART adherence was a key determinant, consistent with previous evidence demonstrating that ART adherence is critical for achieving VLS among ALHIV (Cluver et al. 2020; Mayman & Van Wyk 2023). Conversely, being on a second-line or other ART regimen was associated with increased odds of viral non-suppression, reflecting the challenges often seen with treatment-experienced ALHIV, including prior treatment failure and more complex ART regimens (Mugo et al. 2024). While RiC was significant in bivariate analysis, it was not retained in the multivariate model, suggesting that its effect may be mediated through ART adherence or other clinical factors, a pattern observed in prior studies from South Africa, which has demonstrated protective effects of good, consistent adherence and the use of well-tolerated regimens, such as those based on dolutegravir (DTG), on virological outcomes (Elashi & Van Wyk 2022; Okonji et al. 2021; Western Cape Government Department of Health 2018).

Despite the introduction of DSD strategies during the pandemic, including multi-month dispensing and community-based ART delivery, our findings suggest that these adaptations did not fully mitigate the negative impact of the pandemic on adolescent HIV care. Vulnerable subgroups, such as younger adolescents 10–14 years old, males, and those from socioeconomically disadvantaged backgrounds, may not have benefitted equally from these interventions. This underscores the need for adolescent-responsive care models that are sensitive to age, gender, and social context (Mathamo et al. 2022; Van Wyk & Davids 2019).

Several limitations should be acknowledged. Firstly, the reliance on routinely collected electronic health data introduces the potential for measurement bias because of missing or incomplete records. Notably, CD4 count, and viral load data were more frequently incomplete during the pandemic, likely reflecting reduced laboratory capacity and fewer clinic visits, which may have reduced the power to detect significant differences. Secondly, the use of pharmacy refill data as proxy for adherence to ART may overestimate actual medication intake, as it does not confirm actual medication ingestion. Thirdly, the retrospective observational design of this study limits causal inference; while temporal associations with the pandemic are evident, unmeasured confounders such as household stressors, mental health, and social support may have influenced outcomes. Lastly, the study’s focus on a single urban district may limit the generalisability of these findings to other regions or rural settings with different healthcare infrastructures.

This analysis provides important insights into the ways in which a global public health emergency can disrupt adolescent HIV treatment outcomes in a high-burden urban setting. The observed declines in ART adherence, RiC, and VLS point out the need for adolescent-responsive, resilient care models that can sustain engagement and virological control during future health system shocks. Strengthening these systems is essential to safeguard the health and well-being of ALHIV in the face of ongoing and emerging challenges.

During the COVID-19 pandemic, declines were observed in ART adherence, RiC, and VLS among ALHIV. Although a substantial portion of adolescents maintained full VLS, the overall reductions underscore the heightened vulnerability of this population’s treatment outcomes during periods of systemic disruption.

Factors positively associated with achieving VLS included initiation and maintenance on a first-line ART regimen, consistent ART adherence, and higher CD4 cell counts. These findings emphasise the imperative for adaptive, adolescent-responsive models of HIV care that are resilient to public health emergencies. To safeguard treatment outcomes for adolescents in future crises, it will be essential to strengthen DSD approaches, enhance psychosocial support mechanisms, and address persistent structural barriers to care.