Articles were included in this review if they were Cross-sectional studies reporting findings on the seroprevalence of HIV among Key populations and conducted in Africa over the past ten years.

Articles were excluded if the research was not peer-reviewed or unpublished. Review papers, case series, report cases, editorials, commentary, perspective, news, and opinion papers were not eligible for this study.

We performed a systematic literature search using Medline via PubMed, Google Scholar, and Cochrane Databases. The search was undertaken on February 20, 2024. We searched Medline and Cochrane using medical subject headings and subheadings (Mesh) to index articles. We combined the following terms: HIV and the names of the relevant key population groups and then applied a geographical search filter to retrieve results for African countries. In Google Scholar, we looked for the words (HIV AND Africa with at least one term related to key populations) anywhere in the article. We have put time restrictions on retrieving only publications from the past ten years. We used the Zotero 5.0 software (Center for History and New Media, Fairfax, Virginia) to detect and remove duplicates.

Search strategy on PubMed (Medline) and Cochrane library

((((((((((((“Men who have sex with men”) OR (MSM)) OR (“Trans and gender diverse people”)) OR (Sex workers)) OR (“People who inject drugs”)) OR (PWID)) OR (“People in prison”)) OR (Prisoners)) OR (Gender-diverse people)) OR (((((KEY POP) OR (KEY POPULATION)) OR (KEY POPULATIONS)) OR (KEY-POPULATIONS)) OR (KEY-POPULATION))) OR (Sex-workers)) AND ((“HIV” [Majr]) OR (((((HUMAN IMMUNODEFICIENCY VIRUS [Title]) OR (Acquired Immune Deficiency Syndrome Virus [Title])) OR (Acquired Immunodeficiency Syndrome Virus [Title])) OR (AIDS Virus [Title])) OR (AIDS Viruses [Title])))) AND (“Africa” [Mesh]).

Search strategy on google scholar

(HIV And Africa With At Least One Of The Words Men Who Have Sex With Men OR MSM OR Trans And Gender Diverse People OR Sex Workers OR People Who Inject Drugs OR PWID OR People In Prison OR Prisoners OR Gender-diverse People OR Key Pop OR Key Population OR Key Populations OR Key-Populations OR Key-Population OR Sex-Workers).

Titles and abstracts of records were screened for eligibility by two independent investigators. They have independently assessed the complete reports of all potentially relevant studies for inclusion using an eligibility form based on the inclusion criteria. An independent third investigator adjudicated disagreements.

Data was extracted using a pre-designed digital template for reports that meet inclusion criteria. Data elements included the overall study characteristics, study population detail, sampling approach, recruitment setting, laboratory test, and the total number of screened population and HIV-positive cases.

We used appropriate quality assessment tools to assess each article’s methodological and research quality in this review. Two team members independently rated each included article using the JBI checklist of prevalence studies [10] .

The study results were described using standard summary statistics and grouped by different types of key populations. The total number of people tested and the number of people found to be HIV positive were collected. HIV prevalence measures were calculated with their 95% confidence interval (CI) and presented in tables using Excel software version 2019 (Microsoft, Washington, USA) using the following formulas [11] . Given the methodological variability and heterogeneity among the studies included in this review, a meta-analysis was not conducted.

Proportion ( $\hat{p}$) = $\hat{p}$ = x/n

Confidence Interval = $\hat{p}+/-z\ast \sqrt{\hat{p}\left(1-\hat{p}\right)/n}$

where:

We used the most recent data point if a study reported data for multiple time points. We have rounded all prevalence to one decimal place in the text and tables.

This systematic review includes published papers reporting the seroprevalence of HIV among Key populations in Africa. After searching for available publications, we initially found 467 published articles (330 from PubMed, 21 from Cochrane Library, and 116 from Google Scholar). From this, we removed 23 duplicate records. We excluded 368 records after screening by title and abstracts, and 76 were found eligible for full-text assessment; among the total full-text screened articles for eligibility, we included only 24 studies in the review ( Figure 1 ).

Twenty-four studies met our inclusion criteria. All these studies were carried out in 14 African countries in the Maghreb in Libya, in West Africa (Mali, Burkina-Faso, Nigeria, Togo, Ghana, and Guinea Bissau), in East Africa (Ethiopia, Kenya, and Tanzania), and South Africa (Malawi, Zambia, Mozambique and South Africa). The number of publications in these countries varies from one to 4 ( Figure 2 ).

The minimum sample size was 142 participants in a study conducted in Burkina Faso [12] , while the largest sample size was 132,383 in Nigeria [13] . 4 studies focused exclusively on the seroprevalence of men who have sex with men [12] [14] - [16] , five on female sexual workers [17] - [21] , nine on people who inject drug [22] - [30] and two on prisoners [31] [32] . One study combined men who have sex with men and gender-diverse people [33] , 1 included female sexual workers [34] , and two studies included people who inject drugs, men who have sex with men, and female sexual workers [13] [35] . Fourteen studies focused on a population aged 18 and over [12] [14] - [16] [19] [21] - [23] [26] [28] - [30] [33] [35] , one on those aged 16 and over [20] , five on those aged 15 and over [13] [18] [25] [27] [34] , one on those aged 14 to 24 [17] and one on people of all ages [34] . In two studies, the age of the participants was not stated in Table 1 .

Table 1. General characteristics of the included studies (n = 24).

*NS: not stated, **RDT: rapid diagnostic test.

The sampling frame was inappropriate for meeting the target population in one study [30] . Study participants were adequately sampled in all included studies. The sample size was insufficient in 6 studies [17] [20] [22] [30] [35] , and the study topics and setting were unclear in 1 study [34] . Data analysis was not performed with sufficient coverage of the sample identified in one study [34] , and valid methods were not used to identify diseases in one study [16] . HIV seroprevalence was measured in a standard and reliable manner for all study participants. Appropriate statistical analysis was not calibrated in two studies [30] [34] , was not established in one study [18] , and was not applicable in three studies [17] [20] [21] . The response rate was unclear in three studies [18] [30] [34] ( Figure 3 and Figure 4 ).

Questions:

Q1: Was the sample frame appropriate to address the target population?

Q2: Were study participants sampled properly?

Q3: Was the sample size adequate?

Q4: Were the study subjects and the setting described in detail?

Q5: Was the data analysis conducted with sufficient coverage of the identified sample?

Q6: Were valid methods used for the identification of the condition?

Q7: Was the condition measured in a standard, reliable way for all participants?

Q8: Was there an appropriate statistical analysis?

Q9: Was the response rate adequate, and if not, was the low response rate managed appropriately?

Among the 24 studies included, HIV seroprevalence among men who have sex with men was reported in 7 studies, including two in Burkina Faso, two in Mali, one in Nigeria, one in South Africa, and one in Togo. Reported seroprevalence ranged from 3.6% to 42.9% in studies conducted in Burkina Faso and South Africa, respectively ( Table 2 ).

Among the 24 studies included, HIV seroprevalence among female sex workers was reported in 8 studies, including two in Ethiopia, one in Guinee Bissau, one in Kenya, one in Malawi, one in Nigeria, one in Togo, and one in South Africa. Studies conducted in Nigeria and Malawi reported seroprevalence ranged from 5.6% to 69.0% ( Table 3 ).

Among the 24 studies included, HIV seroprevalence among people who use injection drugs was reported in 11 studies, including one in Ethiopia, three in Kenya, one in Libya, one in Mozambique, one in Nigeria, one in South Africa, and three in Tanzania. Studies conducted in Nigeria and Libya reported seroprevalence ranged from 3.3% to 89.6% ( Table 4 ).

Three studies have reported HIV seroprevalence among prisoners of wish: one in Ethiopia (1.9%), one in Ghana (2.4%), and one in Zambia (25.1%) ( Table 5 ).

Only one study conducted in Mali has reported HIV seroprevalence of 24.8% among gender-diverse people ( Table 6 ).

The review underscores considerable variability in HIV seroprevalence across different populations and regions, reflecting diverse public health challenges. HIV seroprevalence among men who have sex with men (MSM) ranges from 3.6% to 42.9% ( Table 2 ), with the highest rates observed in South Africa and Togo [12] - [16] [33] [35] . For female sex workers, prevalence ranges from 5.6% to 69.0% ( Table 3 ), peaking in Malawi [13] [18] - [21] [34] [35] . Among people who use injection drugs (PWID), seroprevalence varies from 3.3% to 89.6% ( Table 4 ), with extremely high rates in Libya and Kenya [13] [22] [24] [26] - [28] [30] [31] [35] . HIV prevalence among prisoners is generally lower, between 1.9% and 25.1% [29] [32] [34] . Data on gender-diverse individuals is limited, with one study from Mali reporting a seroprevalence of 24.8% ( Table 5 ) [33] . The available scientific literature suggests that these divergences may be explained by the complex interaction of social, cultural and political factors that influence HIV transmission and access to care, requiring region-specific public health strategies. Southern Africa has the highest HIV prevalence in the world, with countries such as South Africa, Botswana and Eswatini particularly affected, with gender inequality, labor migration and poverty being key factors [36] - [38] . In East Africa, HIV prevalence varies considerably from one country to another, with high rates in Uganda, Kenya and Tanzania, the main factors being cultural norms such as polygamy or the levirate, population mobility linked to conflict and the fragility of the healthcare system, particularly in rural areas [37] - [39] . In West and Central Africa, HIV prevalence is lower than in South and East Africa, but countries such as Nigeria, Cameroon and the DRC face significant challenges, with key factors including cultural practices such as scarification or ritual circumcision, population mobility due to trade and labor migration, and fragile health systems, particularly in rural and conflict-affected areas [40] - [42] . In North Africa, HIV prevalence is relatively low, but key populations, such as men who have sex with men and injecting drug users, are vulnerable. Factors mentioned in this region include stigma and discrimination hampering access to healthcare services. Indeed, most countries in North Africa have laws and social norms that are repressive towards key populations, hindering access to prevention and treatment services [43] .

One notable challenge is the methodological heterogeneity across studies, which limits the ability to make direct comparisons ( Figure 3 and Figure 4 ). Variability in study design, sample size, and reporting standards complicates the synthesis of results and introduces potential bias. For instance, one study had an inappropriate sampling frame, potentially impacting the generalizability of its findings [30] , while six studies had insufficient sample sizes, raising concerns about the reliability of their results [20] [22] [30] [34] [35] . Several studies also had limitations in data analysis and disease identification methods, including inappropriate statistical and invalid disease identification methods [16] [30] [34] . These methodological flaws, coupled with unclear response rates in some studies and ambiguous study settings, hinder the overall robustness of the findings. Systematic reviews should be carried out in the future, including sensitivity analyses, as soon as enough studies are available with methods for selecting study groups, comparing groups and determining exposure or results in a consistent manner.

Variability in HIV seroprevalence is likely due to differences in study design, methodologies, and reporting standards. This inconsistency makes direct comparisons challenging and affects the synthesis of results. The review indicates multiple studies with a risk of bias due to sampling frame issues and data analysis methods, influencing the overall conclusions. Another significant limitation is the lack of research on underrepresented populations such as gender-diverse individuals and prisoners. The small number of studies on these groups to date compromises our understanding of HIV prevalence among them and reduces the accuracy of the overall conclusions. Additionally, social, economic, and political factors including stigma, discrimination, legal barriers, and access to healthcare play a key role in shaping HIV transmission dynamics in these populations. For example, laws that criminalize same-sex relationships, sex work, and drug use often deter key populations from accessing prevention services or seeking treatment, thereby exacerbating their vulnerability to HIV [5] . In regions with rigid legal frameworks, the stigmatization of certain behaviors further isolates these groups from healthcare services, thereby increasing transmission risks.

Health interventions must be tailored to the seroprevalence rates observed in different populations. For example, targeted prevention and treatment strategies are crucial for high-prevalence groups like MSM in South Africa and female sex workers in Malawi. Addressing the unique needs of these populations can help reduce HIV transmission more effectively. Policymakers should address identified gaps, such as improved sampling methods and standardized reporting practices. Allocating resources based on the severity of HIV prevalence in various groups ensures the effective use of funds and support. Decriminalizing behaviors such as same-sex relationships and sex work can improve access to essential HIV prevention and treatment services. Future studies should focus on enhancing sampling methods, increasing sample sizes, and employing valid disease identification methods to improve the accuracy of seroprevalence estimates. There is also a need for more research on underrepresented populations, such as prisoners and gender-diverse individuals, to provide a comprehensive understanding of HIV prevalence and inform public health strategies.

This review includes studies with a high risk of bias that could weaken the conclusions. However, it should be noted that this study deals with a taboo subject and that stigmatizes respondents. The study aimed to be as exhaustive as possible to provide a comprehensive overview of the issue. The presentation of the analysis of the quality of the studies included enables the reader to form an opinion that goes beyond the postures of the authors.