This systematic review protocol follows the preferred reporting items for systematic reviews and meta-analysis protocols (PRISMA-P 2015) guidelines (see Additional file 1: PRISMA-P 2015 Checklist).
Eligibility Criteria
The PICOT mnemonic standing for population, intervention, comparative, outcomes and time will guide the definition of inclusion and exclusion criteria.
Study design
Randomized control trials, non-randomized control trials as well as observational studies will be sourced for in this review.
Participants
We will include studies whose Population of interest include adolescents aged 10-19 years, living with HIV and on ART.
Interventions and comparators
The Intervention of interest will be transition interventions for adolescents on HIV treatment (ART). There will be no Comparison as observational studies will be included in this review.
Transition interventions are structured and designed to improve the health outcomes of adolescents and young adults with chronic health challenges pre-, during and post-transfer from child/adolescent healthcare facilities to adult healthcare facilities. Pre-transition intervention happens within the child/adolescent facilities and is run by multidisciplinary healthcare professionals. It starts with disclosure of status and involves transition preparedness, such as addressing what transition is and its barriers, for example, anxiety and fear of transition, which can be measured by using transition readiness scale [22, 27]. This intervention could also focus on providing skills and discussions on adherence to ART, sexually-oriented relationships and coping tactics [8]. These interventions flow into the during transition interventions. The transition interventions targeted at when the transition will occur could include communication between pediatric/adolescent and adult providers. This communication includes adolescents being introduced to and having a full consultation with adult healthcare personnel within the pediatric/adolescent clinic and assistance provided for adolescents to overcome system-level barriers such as insurance changes or social services [27, 28]. These interventions also involve ALHIV being taught self-efficacy and psychosocial skills in peer/family/community support groups as they get transferred to adult facility at this stage.
Post-transition interventions occur in the adult clinics and are more focused on assessing transition outcomes. Transition experiences of ALHIV are gauged. Also, post-transition immunological and virological measurements are taken. To inform better strategies to transition ALHIV and examine if the transition was successful or not [29].
Outcomes
The primary Outcomes of interest will be adherence to ART, retention in care, or viral load suppression. All secondary outcomes such as psychosocial well-being and self-efficacy will also be considered. To ascertain the success of the transition interventions, the study will consider transition outcomes such as percentage/rate of ALHIV recorded as a loss to follow up or retained in care over a period and mortality rate post-transition. It will also consider immunological and virological outcomes such as CD4 count and viral load measurement post-transition as reported in the studies.
Timeframe
The Timeframe for the inclusion of published articles will be 2000 to 2021.
Setting
There will be no restriction on settings or geographical location.
Further inclusion criteria
Grey literature and quantitative studies published in peer-reviewed journals in any language will be included. All languages will be considered to avoid publication bias. The studies should report quantitative results on ART adherence measurement, retention in care, viral suppression, psychosocial well-being or self-efficacy.
Exclusion criteria
Literature reviews, protocols, systematic reviews and meta-analyses will be excluded. Qualitative studies will also be excluded because they do not report quantitative outcomes of interest. Studies that do not focus primarily on transition intervention but just mentioned transition in passing will be excluded.
Information sources and search strategy
We will conduct a comprehensive systematic search using predetermined search strings on electronic databases such as PubMed, Scopus, Web of Science, Ebscohost, CINAHL, Science Direct, Google scholar and the World Health Organization's (WHO's) library database. (See Additional file 2: Draft for the database search strategies). Likewise, the reference list of included studies will be hand-searched to identify additional eligible studies.
Study selection
Two independent reviewers will screen the title and abstracts of the studies identified according to the PICOT criteria: (1) the study population include adolescents (10-19 years) living with HIV; (2) involves a transition intervention (transfer from pediatric to adult HIV care program); (3) reports quantitative (outcome) results of the intervention; and (4) the timeframe for the inclusion of published articles will be 2000 to 2021. Full texts of the eligible studies will be retrieved and reviewed by the two reviewers. A third reviewer will resolve a conflict of judgment between the two reviewers when this is not resolved through discussion. The Preferred Reporting Items stipulate for Systematic Reviews and Meta-Analyses protocols (PRISMA-P) statement [30] will inform the writing of the protocol and the study results will be reported according to the Preferred Reporting Items stipulate for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement [31]. The PRISMA 2020 statement consists of 27-item checklists and a four-phase flow diagram, which will be adhered to for reporting and the four-phase flow diagram for screening studies.
Data extraction
A data extraction form will be designed to gather information on the author and year of publication, study population, intervention description, study design/outcome, outcomes' measurement approach, result, level of evidence and intervention effect. This helps to synthesize relevant information from the individual studies in a standard and transparent way. The information extracted is presented in a table clearly and concisely. The data extraction form will be piloted in at least five transition intervention studies on ALHIV before finalization. To organize the data extraction form, Microsoft Excel software applications will be used.
Outcomes of interest for adolescents on antiretroviral therapy
Extant literatures [32-36] have identified the three primary outcomes of interest among ALHIV to be adherence to ART, retention in care and viral load suppression.
Adherence
Adherence can be defined as the process by which HIV patients take their medication (antiretroviral drugs) as prescribed [37]. Adherence to ART of at least 95% (for a daily single-pill regimen - 14 non-sequential missed doses over a year) is recommended as optimal adherence. At the same time, anything below is considered a suboptimal/poor adherence to ART [38].
Method of adherence measurement
Varied adherence to ART measurements exists in the literature, and these include viral load, self-report, pill count, home record, pharmacy refill, healthcare provider assessment, or electronic monitoring [39]. Data on the listed method of adherence measurement will be considered in this review.
Viral load suppression
According to [40], viral load suppression is defined as “suppressing or reducing the function and replication of a virus.” Hence, viral load measurement is used to monitor treatment effectiveness in HIV patients on ART [36]. According to the Abbott platform, an undetectable viral load occurs when viral load reads <40 copies/ml and a detectable viral load at ≥ 40 copies/ml [41]. Virological failure occurs when viral load is >1000 copies/ml after two consecutive viral load tests in 3 months, which involves having adherence support and being on ART for at least 6 months [42].
Retention in care
The World Health Organization defines retention in care as “the continuous engagement from diagnosis in a package of prevention, treatment, support and care services [36].” Retention in care can be measured by considering missed visits, appointment adherence, visit constancy and gaps in care, as indicators [43]. Low retention in care has been associated with non-viral suppression and a high mortality rate [44-46].
Risk of bias assessment
Two researchers will screen and evaluate the quality of each selected research independently to ensure accuracy and generalizability [47]. Checklists such as the Cochrane Collaboration tool RoB 2 will be used to assess the risk of bias in included randomized control trial studies [48]. The Newcastle–Ottawa Scale checklist will be used to assess the risk of bias in included prospective observational studies [49]. The Risk of Bias in Non-randomized Studies - of Interventions (ROBINS-I) will be used to assess the risk of bias in included non-randomized intervention studies [50].
Effect measures
We will consider effect size measurements such as the relative risk, risk difference and confidence intervals (CI) from the individual included studies and for each of the outcome of interest. This will inform the overall magnitude of effect and evidence confidence.
Data Analysis
The data extracted in step 5 will be imported to SPSS statistical package for analysis. If there is high heterogeneity in the included studies such that a meta-analysis is impossible, the present study will adopt a narrative synthesis. The narrative synthesis will structurally summarize and describe the characteristics of the studies: study population, intervention description and study design/outcome, their findings and quality [47]. The included studies' findings will highlight their direction and effect size, effect consistency and the strength of evidence for the effect [47]. While for the quality, the study will only include studies with low or moderate risk of bias to improve the trustworthiness of the synthesis [51]. Studies included will be compared for similarities and differences, while possible trends will also be described.
If meta-analysis is possible, it will be done using Stata statistical package and will be focused on randomized control trials. The study will use a random-effects meta-analysis because, according to [47], "it allows the true effects underlying the studies to differ and thus accounts for unexplained heterogeneity between studies." The study will pool estimates to obtain a summary estimate, with 95% CI. The χ2 test will be performed to test for heterogeneity, and I2 statistic to assess the degree of homogeneity [52]. If the study finds more than ten studies included, we will use a funnel plot and an Egger test to assess publication bias [53]. If there are sufficient data, we will consider doing a subgroup analysis on age intervals, sex, study designs and model of transition interventions.
Certainty assessment
Two reviewers will individually ascertain the overall quality of evidence for the primary outcomes of interest. The strength and certainty of evidence from each study included will be evaluated according to the GRADE system [54]. The GRADE system provides the platform to assess the quality of evidence for each outcome on five different domains which are: risk of bias, consistency; directness; precision; and publication bias. The presence of risk of bias (study limitations), inconsistency, indirectness, imprecision or publication bias leads to lowering the quality of evidence while large magnitude of effect, opposing plausible confounders and dose-response gradient can increase the quality of evidence [54]. Confidence in the effect estimate for each outcome will be ascertained independently by the two reviewers as high, moderate, low or very low. High denotes sound confidence in the true effect lying close to that of the estimate of the effect. Moderate denotes moderate confidence in the true effect lying close to that of the estimate of the effect i.e., there is a possibility for the true effect to be substantially different from the estimate effect. Low denotes limited confidence in the true effect lying close to that of the estimate of the effect i.e., true effect may be substantially different from estimate effect. Very Low denotes very little confidence in the true effect lying close to that of the estimate of the effect as the true effect is very likely to be different from the estimate effect [55]. A third reviewer will resolve a conflict of judgment between the two reviewers when this is not resolved through discussion. A detailed summary of findings table will be included in the actual review paper.