This is a retrospective longitudinal cohort study conducted among HIV-infected adolescents in Tanga, Tanzania. Results showed that the overall interruption in treatment rate was high (26.2%), much higher than required to meet the 95-95-95 targets. Adolescents aged between 15 and 19 years, male sex, those residing from Mheza district, advanced HIV disease, being on ART for long time and those placed in non DTG based drug had significantly higher risk for interruption in treatment. Our study population of HIV-infected adolescents in Tanga was higher enough for whom detailed longitudinal clinical and demographic were available.

The interruption in treatment rate observed in this study population were high and somewhat consistent with other studies carried out in some African countries and even outside the continent. Other estimates of interruption in treatment among HIV-infected adolescents in Sub-Saharan African countries including Cameroon, Kenya, Tanzania and South Africa have been reported ranging from 13.4 to 42.2% [4, 14,15,16]. However, compared to our study, low propotion of interruption in treatment of 4.3% was reported in Asian regional cohort study incoprporating six countries in Asia, 9.2% among Indian adolescents [17] and 8.4% in Zimbabwe [18]. This wide range of reported proportion of interruption in treatment could be due to differences in the study population and the disparity of follow up time in various studies. Viral load suppression among adolescents during the time of study, fear of discrimination and disclosure is anticipated to cause dropout to treatment and care among our study population. Strengthening the comprehesive tracking system and establishment of adolescents HIV support groups are highly encouraged.

The current study found that the determinants of interruption in treatment were generally similar to those in other African settings. Adolescents aged 15–19 years had a higher risk of interruption in treatment from ART uptake. The finding was inline with previous studies conducted in Ethiopia, South Africa, and sub-Saharan Africa [19,20,21]. Fear of stigma, peer pressure, growing up independent and discrimination might be a contributor to interruption in treatment among adolescents [8, 22]. In our context, we suspect but cannot confirm without further research that the significantly greater treatment interruption observed among adolescents of the said age group could be caused by their being busy with work or feeling shy about attending the clinics. Clearly, treatment facilities need to focus more on preventing IIT among adolescents to increase their chances of survival and gain them the greatest benefit from treatment and conduct age-specific interventions to reduce interruption in treatment among adolescents in our resource settings.

We also identified increased risk of interruption in treatment among male adolescents in our study which was consistent with some previous studies conducted in Tanzania, Ethiopia and Malawi [23,24,25]. Nevertheless, a study conducted in 15 ART programs in Africa, Asia, and South America reported no association between gender and interruption in treatment [26]. However, previous findings in Uganda and a study conducted on MTCT-Plus programs in 9 different countries showed male adolescents were protective towards interruption in treatment [27, 28]. The reason for gender gap of interruption in treatment among male adolescents in our study might be attributed to challenges of obidiency, busy with recreational activities, stigma, undocumented transfer to other HIV care clinics and death. This findings advocates for strengthening the linkage to HIV care and counseling, implement adolescent-friendly service approach and necessitates further investigation of cultural and behavioural differences among male adolescents for their better clinical health outcome.

Our study found that sicker patients were more likely to interrupt treatment. Increased risk of interruption in treatment with advanced WHO clinical staging (stage IV) observed in our study was consistent with previous study conducted in Tanzania, India, Kenya and Nzimbabwe [14, 16,17,18], that adolescents with more advanced disease stage understand the benefit of regular clinic visit, however they can’t stick to patient follow-up schedule may be due to their weak conditions or undocumented transferred out or death. It is important for measures to be put in place that improve retention rates among patients who display advanced symptoms of HIV despite taking ART, since results from populations with high interruption in treatment may be biased. Adolescent patients who are lost during follow-up may be less healthy than the patients who remain in the program and therefore more likely to die, leading to underestimation of death rates. Alternatively, the sickest adolescents on ART may be more highly motivated to sustain regular follow-up while the healthiest, asymptomatic adolescent patients may have extended periods between follow-up visits, given little perceived need. Such patients may be less likely to die and may subsequently return to the follow-up. At a program level, interruption in treatment can make it difficult to evaluate outcomes of treatment and care [12, 29]. High rates of interruption in treatment also diminish treatment options and substantially limit the effectiveness of ART strategies [12]. This findings calls for early test and treat services among adolescents in our resource-constrained settings. However, no association was observed in a study conducted in Kenya between interruption in treatment and advanced WHO stage [16].

This study, like other studies, has found that adolescence placed in second-line ART regimen had reduced risk of of interruption to treatment, our findings agree with studies conducted in Tanzania and Ethiopia [14, 30]. However, our results contradict other studies conducted in Myanmar and Nigeria [31, 32], which reported a likelihood risk of interruption in treatment among patients on a second-line regimen. A study conducted in Nigeria stated an increased risk of interruption in treatment might be caused by adverse effects obtained from second-line drugs [32]. Adolescents in the second-line being protective towards interruption in treatment in our study could be expalined by intensive counseling and close followup given to this group after first line treatment failure by health care givers.

Studies in South Africa and USA show that among adolescents who virally unsupressed the risk of interruption in treatment is high [15, 33]. However, an interesting findings in our study show that viral load suppression failure was not found to be a significant risk factor for interruption in treatment among adolescence in Tanga region. This perhaps is due the fact that adolescents with poor viral suppression has special management package named enhanced adherence counseling where each unsupressed client is paired with the counselor and peer for close followup to support adherence to clinic visits and ART uptake.

Other noteworthy findings were increased risk of interruption in treatment among adolescence who have been placed in non-DTG based regimen of which we have observed limited report of related studies and among adolescents who have been on ART for long time (at least one year from enrollment or ART initiation date). This finding contradict with other studies in African countries. For instance, a study conducted in South Africa show that shorter time on ART (≤ 12 months) was independently associated with higher interruption in treatment [15]. Moreover, prolonged period on ART was protective among adolescents in a retrospective cohort study conducted in Northwest and Southwest of Cameroon [4]. High rate of interruption in treatment in our study might be due attributed to patients who have been on treatment for long period, hence some feel they are healthier or they are get tired of taking drugs or they are in advance disease progression. This calls for a need for healthcare providers to keep track of long term clients on ART in order to reduce likelihood of transmission of HIV in the community.

It is important to note that clinics may tend to invest their limited resources in educating patients who are initiating ART about adherence, rather than in efforts to track patients who have interrupted treatment in their communities and return them to care, further exascerbating the impact of non-retention. Finally, in order to reach the recommended 95-95-95 goals to end the AIDS epidemic by 2030, it is essential that adolescent patients who are initiated with ART remain in care and virally suppressed, the final step of the treatment/prevention cascade [34]. With the goal of all HIV-positives adolescent being on treatment, without adequate retention, there is substantial concern about the emergence of unresolveable drug resistance [35]. It is important from a clinical and programmatic perspective to ensure that HIV care and treatment serive providers are aware of this high risk patient group of adolescents and put in place good and relevant strategies to track and retain them.

Major strengths of this study is the relatively large sample size with inclusion of all the 86 health facilities in the region, providing us substantial power to obtain accurate and reliable estimate over a large number of potential risk factors, adjusted for one another, and the sufficient follow-up time of 24 months. Additionally, the longitudinal nature of this analysis provided an opportunity to assess rate and time of interruption in treatment and associated predictors. The major limitation is that our analysis based on the secondary data from the existing database. It is possible that not all clinical data were correctly captured and that some data were not captured at all, resulting in missing data which however was handled during data analysis but this is a general limitation of using routine data.