The framework adopted for conducting the proposed review is by Arksey, H., and O’Malley, [
18]. Briefly, the framework involves (I) identifying the research question, (II) identifying relevant studies, (III) study selection and (IV) charting the data, and (V) collating, summarizing, and reporting the results.
I.
Identifying the research question
The research question is: What is known from existing literature about the dynamics of HIV transmission in men in South Africa?
Eligibility of research question
The study will use an amended PICOS (population, intervention, comparison, outcomes and study setting) framework to determine the eligibility of the research question, as described in Additional file
2: Table S1. Additional file
2: Table S1 is provided in the supplemental documents to this protocol, as Additional file
2.
II.
Identifying relevant studies
Primary research articles, published in peer-reviewed journals; review articles; and gray literature that address the research question, will be included in this study. All study designs will be included for review. Databases that will be used to source literature include; PubMed, Web of Knowledge, Science Direct, EBSCOhost, Google Scholar, World Health Organization library database, and UNAIDS database. We will also use reference lists and existing networks such as organizations and conferences to source relevant literature. The search terms will include “HIV infection in men and sex behaviors, or cultural factors, or social networks, or sexual networks, or phylogenetic or biological factors”. The language and year of publication will be restricted to English and between 2006–2016, respectively.
One reviewer will conduct a comprehensive title screening by searching and uploading all literature search results on Endnote X7 software; all studies that do not address our research question will be excluded together with all duplicates. The final Endnote database will be shared for abstract screening; at this stage, we will use two independent reviewers to extract data in parallel, from all relevant search engines. The two reviewers will use the inclusion/exclusion criteria to identify relevant literature for inclusion and for further evaluation (s). Copies of full articles will be obtained and maintained.
Inclusion/exclusion criteria, developed based on the research question will ensure correct identification and selection of relevant studies. To be included, studies must be focusing on HIV infection, transmission, or acquisition on heterosexual males. We will include both peer-reviewed publications and gray literature of all study designs with relevant interventions. We will exclude studies on HIV infections due to mother to child transmission of HIV and studies on high-risk men (e.g., men who have sex with men, gays, bisexuals, injection drug users, male sex workers, and transgender populations). The search strategies will be piloted to test the appropriateness of the selected databases and keywords in delivering relevant information/literature. The study selection procedure will be summarized using a PRISMA chart (Additional file
3: Figure S1), modified from [
19].
We will use a standardized data extraction sheet. The sheet will include bibliographic details, study design, number of participants, intervention (s), study setting, and conclusions for the primary and the secondary outcomes of the interventions. An additional charting summary Additional file
4: Table S2 shows this in more detail [see Additional file
4].
V.
Collating, summarizing, and reporting the results
We will present a narrative account of findings from existing literature through thematic content analysis of the extracted literature, structured around the following interned outcomes: HIV transmission, HIV acquisition, and HIV infection. Results of the studies on HIV infection, acquisition, or transmission in heterosexual men outside South Africa vs South African men, on evidence of whether interaction of interventions leads to HIV infection (transmission or acquisition), will be coded by all authors independently.
We use NVivo version 10 to present emerging themes according to the relevant interventions.
The below processes will be followed:
-
Coding
-
Categorize codes into major themes
-
Build theme-related themes (cut-and-paste technique)
-
Display data
-
Identify patterns in the data and identify sub-themes
-
Summarize
Authors will interrogate the resulting themes and critically examine their relationship to the research question. Authors will also scrutinize the meanings of the findings as they relate to the overall aim of the study and address the implications for future research.