Introduction

The global crisis of people affected by humanitarian crises caused by conflict, violence, climate change, and natural disasters has reached unprecedented levels. By the end of 2023, the number of individuals worldwide displaced by humanitarian crises exceeded 117.3 million, the highest level seen since World War II [1]. The United Nations (UN) defines humanitarian crisis as “an event or series of events that represents a critical threat to the health, safety, security, or well-being of a community or other large group of people usually over a wider area” [2]. Recent examples of humanitarian crises caused by conflict and violence that led to mass displacement include the Israel-Gaza conflict, the Russia-Ukraine war, the Syrian civil war, the Haiti political crisis, the military and ethnic conflict in Myanmar, and the South Sudan civil war [1]. Climate change and natural disasters such as the droughts in Ethiopia and Afghanistan and the large-scale floods in Nigeria and Ecuador may also lead to temporary or permanent displacement [3, 4]. Globally, weather-related and geophysical disasters internally displaced 26.4 million people in 2023 [4]. A situation analysis of the Israel-Gaza conflict describes the breadth of public health issues caused by the conflict and forced displacement, including acute malnutrition, increased non-communicable disease burden due to interrupted treatment, poor birth outcomes, and mental health problems [5]. Ultimately, humanitarian crises have multiple adverse health impacts that significantly increase mortality and morbidity rates in the affected regions [6, 7].

Poverty, powerlessness, political and economic instability, disruptions to the health systems, and loss of social networks are common aspects of humanitarian crises and can contribute to a person’s vulnerability to human immunodeficiency virus (HIV) infection [8,9,10,11]. In conflict- and disaster-affected settings, health-seeking behaviors such as testing for HIV among those at risk for infection and maintaining adherence to antiretroviral therapy (ART) among persons with HIV (PWH) may assume lower priority than securing more urgent and basic needs such as food and shelter [12, 13]. Limited access to money and other resources may lead individuals affected by humanitarian crises to exchange sex for money or nonmonetary items as a means of survival [14, 15]. The use of alcohol and/or drugs as coping strategies may further increase the risk of HIV infection [16]. Nevertheless, evidence on the direction and scale to which humanitarian crises affect the epidemiology of HIV/AIDS can be difficult to elucidate and is often context-specific [10, 17,18,19]. For instance, the prevalence of HIV might fluctuate before, during, and after times of conflict, with prevalence data remaining stable during conflict [18, 20, 21]. Prior research suggests the flow of humanitarian aid, forced displacement, and limited population mobility for the duration of the conflict may slow the spread of HIV [18, 20]. On the other hand, factors such as exposure to conflict-related traumatic events (e.g., sexual violence) and low HIV knowledge can increase HIV risk at the individual level during all stages of conflict [8, 18].

Individuals affected by humanitarian crises may present with a range of risk factors for HIV acquisition and transmission. Yet, there remains limited evidence on how these factors interact with the challenges of living in a conflict or disaster setting and influence the risk of HIV acquisition. Previous approaches to identifying individual risk factors of HIV have shifted towards characterizing multiple levels of HIV risk and risk contexts. One example of such approaches is the social-ecological model, which posits the complex interaction between an individual at risk for HIV (e.g., a person who injects drugs) and their social network, community, public policy, and the stage of HIV epidemic in determining their ability to protect themselves and others from acquiring HIV [22]. Researchers have also examined modifiable and non-modifiable determinants of HIV infection in the context of mobility, to understand the routes through which labor migration might exacerbate the spread of HIV and to inform potential HIV prevention interventions [23]. A study exploring how climate change may influence the HIV/AIDS epidemic identified four contextual pathways including food insecurity, erosion of public health infrastructure, human migration, and the prevalence of other infectious diseases [15]. These frameworks can serve as important guides for assessing factors affecting vulnerability to HIV in diverse settings and communities [22, 23], facilitating implementation of evidence-based HIV prevention programs [24], and informing effective strategies to advance actionable health research in humanitarian settings [25, 26].

Given the increasing number of people affected by humanitarian crises worldwide [1], the goal of reducing new HIV infections by 90% by 2030 [27], and the complexity of addressing HIV in humanitarian settings [10, 19], we conducted a scoping review [28, 29] to explore the evidence and identify key concepts concerning HIV acquisition in humanitarian settings. Reviews undertaken on this topic have focused on the associations between mental health and HIV acquisition and disease progression in conflict settings [30] and on HIV, reproductive, and sexual health services in emergency settings for specific populations such as adolescent girls and young women [31] and sex workers [32]. To our knowledge, there is no rigorous literature review on factors associated with HIV acquisition in the context of humanitarian crises. This scoping review, therefore, is guided by the following research question: “What is the global evidence on factors associated with HIV acquisition among individuals affected by humanitarian crises?” The review seeks to summarize research on HIV risk factors among people affected by humanitarian crises and identify recommendations for future research and practice.

Methods

We followed the Joanna Briggs Institute (JBI) methodology for scoping reviews [33] and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) reporting guidelines [34]. The review protocol was previously published [35]. A completed PRISMA-ScR checklist is included as Supplementary Table 1. Our multidisciplinary research team included individuals who have taken part in conducting HIV-related research and published peer-reviewed literature in the HIV/AIDS, humanitarian, and health sciences fields.

Eligibility Criteria

We included articles that met the following criteria: 1) involving participants affected by humanitarian crises (see below definition of terms); 2) including empirical data on HIV serostatus or risk factors of HIV acquisition as a main or substantial focus of the article; 3) were conducted in the context of humanitarian crises including natural and human-caused disasters [2]; and 4) were published in English between January 1990 through March 2022. The search was limited to 1990 onwards which reflects the period when the demographic data on refugees and displaced persons first became available in many host countries [36]. We excluded studies that 1) did not include humanitarian settings or participants; 2) did not report original data (e.g., reviews or opinion pieces); 3) did not address HIV (e.g., focused on sexually transmitted infections (STI) excluding HIV); and 4) studies without full-text available. All quantitative, qualitative, and mixed methods designs were included. Modeling studies reporting empirical data were included.

Definition of Terms

We recognized the use of different terms in existing literature to describe the concepts and contexts concerning HIV acquisition and humanitarian crises. Informed by relevant review papers and guidance documents published by organizations providing humanitarian assistance, the definitions of the key concepts used in this scoping review are detailed as follows.

We defined factors associated with HIV acquisition as a broad range of determinants that may increase or decrease an individual’s risk of HIV acquisition, and contribute to HIV prevalence or incidence [37]. Examples of these determinants include biological risk factors (e.g., other STI) [37], behavioral risk factors (e.g., alcohol or substance use) [38], protective behaviors (e.g., condom use, male circumcision) [37], and structural variables (e.g., access to health care) [39]. People affected by humanitarian crises include refugees, asylum seekers, internally displaced persons (IDP), and other individuals in need of protection and/or assistance based on humanitarian grounds as described by the United Nations High Commissioner for Refugees (UNHCR) [1, 40]. Humanitarian crises as defined by the UN [2] include i) natural disasters or environmental events that cause considerable morbidity and mortality, such as droughts, floods, storms, and earthquakes [41], and ii) complex emergencies associated with fragile states or areas of conflict, in which a total or considerable breakdown of authority has occurred and requires a large-scale or international response, and often includes large numbers of affected and displaced civilian populations [42, 43].

Data Sources and Search Strategy

The study team (KK, DH, HP, KN) identified initial keywords for the search strategy from previous HIV- and humanitarian-related research [37, 44,45,46] and refined them collaboratively. The search strategy was developed by a public health librarian (KN) and peer-reviewed by a second librarian not participating in the study. The search strategy in the current scoping review included terms for participant (e.g., refugee, asylum seeker, internally displaced person), concept (e.g., HIV infection or transmission, HIV incidence or prevalence, risk factor, sexual behavior), and context (e.g., disaster, war, humanitarian). Full search terms (including controlled vocabulary) tailored to individual databases are listed in Supplementary Tables 2, 3, 4, and 5.

Studies were identified through searches on several electronic databases including MEDLINE, Embase, Global Health (all accessed via Ovid), and Scopus. We identified additional studies by forward citation chaining using the “cited by” function on Google Scholar for relevant review papers. Additionally, we searched for relevant grey literature by reviewing websites of organizations involved in humanitarian assistance and development work (e.g., UN agencies), relevant non-government organizations (e.g., Elrha), research networks (e.g., Refugee Research Network), and the International AIDS Society’s conference abstract databases and abstract books (2006–2021) (Supplementary Table 6). For unpublished studies (e.g., conference abstracts, repositories), we contacted study authors to request additional data. A search on Google Scholar was conducted until 10 consecutive results unrelated to the topic were displayed in the search results. Searches were conducted between October 2021 through March 2022.

Study Selection

Following the search, all identified citations were uploaded to Covidence (Veritas Health Innovation, Melbourne, Australia), a web-based platform to manage and streamline systematic review process including deduplication, citation screening, full-text review, and data extraction. Two reviewers (DH and HP) independently screened titles and abstracts and excluded articles that did not meet the inclusion criteria for this review. A full-text assessment of the remaining articles was conducted by three reviewers (SA, DH, HP). Disagreements during the screening and full-text assessment processes were resolved by consensus, with additional input from a team member (KK).

Data Extraction

Four reviewers (SA, DH, HP, DG) extracted data into Covidence using a data extraction tool that included author, publication date, type of publication, citation, sample size, type of participants, inclusion and exclusion criteria, participant characteristics (gender, age), country/location of study, context of humanitarian crisis, study aims, study design, outcomes (i.e., HIV prevalence or incidence, modifiable factors and non-modifiable factors for HIV acquisition), and implications for future practice and research reported in the study. Two reviewers (SA and DH) verified the completed extraction forms and subsequently exported them from Covidence into a Microsoft Excel spreadsheet for analysis.

Data Synthesis

Two authors (SA and DH) categorized modifiable and non-modifiable factors of HIV acquisition through discussion and informed by the work of Weine and Kashuba on HIV risk among labor migrants [23]. We generated a list of non-modifiable factors that included age, gender, ethnicity, location (e.g., rural or urban), and place of birth or origin. Modifiable factors, defined for the purpose of our analysis as any factors that are not biologically or socially determined and can potentially be modified through policy or intervention [47, 48], were categorized into five broad categories: i) policy and structural (e.g., access to HIV prevention and care, mobility status); ii) sociocultural (e.g., religious beliefs, marital status); iii) health and mental health (e.g., substance use, STI other than HIV, post-traumatic stress disorder (PTSD), depression); iv) sexual practice (e.g., exchanging sex for money or nonmonetary items, multiple sex partners); and v) exposure to traumatic events related to humanitarian crisis (e.g., sexual violence, abduction or kidnapping). Unless specified by the study authors as refugees, asylum seekers, IDP, or other specific groups (e.g., military personnel), we used the term persons affected by conflict or persons affected by natural disasters to describe the study populations. We presented descriptive characteristics of the included studies, a narrative summary relevant to the research question, and implications for future programs and research reported in the studies. We created a world map with MapChart to visualize the geographical locations of the included studies and generated a figure with Microsoft PowerPoint (Microsoft, Redmond, WA, USA) to illustrate the range of factors affecting HIV acquisition in humanitarian settings.

Results

The search yielded 18,595 potentially relevant citations (18,477 from databases, 118 from grey literature). After removing 10,404 duplicates, we screened titles and abstracts of 8,191 citations. We excluded 8,007 citations and retained 184 articles for a full-text review (Fig. 1). These 184 were screened at the full-text level; 135 were excluded, mostly (102) due to lack of relevance to the research question. A total of 49 articles were included in the final analysis (Fig. 1).

Fig. 1
figure 1

PRISMA-ScR flow diagram

Full size image

Study Characteristics

Table 1 presents information of the included studies by author name and year, study location (country), study aims, population, design and methodology, context of humanitarian crises, and non-modifiable and modifiable factors. All studies were published in peer-reviewed journals between 1991 and 2022, with nearly three-fourths published from 2010 onward (n = 35, 71%). Studies were primarily conducted in complex emergencies (n = 40, 81.6%) [20, 49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87], complex emergencies and natural disasters (n = 4, 8.2%) [88,89,90,91], natural disasters (n = 3, 6.1%) [92,93,94], and not specified (n = 2, 4.1%) [95, 96]. In terms of geographical settings, most were single-country studies (n = 43, 87.8%) (Fig. 2). Not shown in the figure were four studies involving multiple countries in Sub-Saharan Africa (SSA) (n = 4, 8.2%) and two global analyses of 69 and 177 countries, respectively (n = 2, 4.1%). Among the single-country studies (n = 43), the majority were conducted in SSA (n = 31, 72.1%) specifically in Uganda (n = 13). Other single-country studies were conducted in Asia (n = 4, 9.3%), North America (n = 3, 7%), Europe (n = 2, 4.7%), South/Latin America (n = 2, 4.7%), and the Middle East (n = 1, 2.3%).

Table 1 Characteristics of included studies (n = 49)
Full size table
Fig. 2
figure 2

Countries of included studies (n = 43, single-country study), excluding 4 studies involving multiple Sub-Saharan African countries and 2 global analyses of 69 and 177 countries. The geographical locations of the 43 studies were Afghanistan (1), Angola (1), Cameroon (1), Canada (2), Colombia (1), Democratic Republic of Congo (4), El Salvador (1), Ethiopia (1), Germany (1), Guinea-Bissau (1), Lebanon (1), Lesotho (1), Malawi (1), Mozambique (1), Namibia (1), Nigeria (1), Pakistan (2), Rwanda (2), South Africa (1), South Sudan (1), Sudan (1), Thailand (1), Uganda (13), the United Kingdom (1), and the United States (1)

Full size image

The main study objectives, as reported by the authors, were to determine the prevalence or incidence of HIV and risk behavior correlates of HIV infection (n = 27, 55.1%) [49, 51, 54,55,56,57,58, 60, 63, 68,69,70,71,72, 74,75,76,77, 79, 81, 84, 86, 87, 91, 92, 94, 96], the prevalence or incidence and correlates of HIV and other infectious diseases (n = 12, 24.5%) [50, 53, 61, 62, 64,65,66, 73, 80, 83, 88, 95], the potential impact of complex emergencies and natural disasters on HIV prevalence or incidence through mathematical modeling (n = 8, 16.3%) [20, 52, 78, 82, 85, 89, 90, 93], and the prevalence and correlates of mental health problems and HIV (n = 2, 4.1%) [59, 67]. Among the 49 included studies, 43 (87.8%) were quantitative, 4 (8.2%) were qualitative, and 2 (4.1%) were mixed methods studies. Most were cross-sectional in design (n = 37, 75.5%), and the remaining were modeling studies (n = 8, 16.3%), prospective cohort studies (n = 2, 4.1%), and case–control studies (n = 2, 4.1%). Of the 49 studies, only one (2%) was an intervention study in the form of an HIV screening program.

Evidence of Non-modifiable Factors for HIV Acquisition in Humanitarian Settings

Of the 49 included studies, 36 (73.5%) reported information on the effects of five non-modifiable factors on HIV prevalence, incidence, or serostatus in the context of humanitarian crises, namely age (n = 28), gender (n = 24), location (n = 16), place of birth or origin (n = 8), and ethnicity (n = 4). A list of factors and corresponding frequencies are presented in Table 2.

Table 2 Non-modifiable and modifiable factors for HIV acquisition in humanitarian crises assessed in the included studies (n = 49)
Full size table

Age

The effect of age on HIV acquisition was referenced in 28 studies [49,50,51, 53, 55,56,57, 60,61,62,63,64,65,66, 68,69,70,71,72,73, 76, 77, 80, 84, 88, 90, 94, 95] with varying results. Twenty-two studies reported HIV prevalence and correlates in different age groups [50, 51, 53, 55,56,57, 60,61,62, 64,65,66, 68, 69, 71, 72, 76, 77, 80, 84, 90, 94]; 5 examined factors related to HIV serostatus [49, 70, 73, 88, 95]; and 1 measured HIV incidence rate [63]. Several studies reported that the prevalence of HIV was higher among older individuals in humanitarian settings [50, 55, 57, 60, 66, 69, 76, 80, 90]. In a large sample of refugees and asylum seekers in Germany (n = 15,317), increased age was a significant predictor of HIV infection (Odds Ratio (OR) 1.09 [CI not reported]; p < 0.001) [50]. Similarly, among 970 male and female refugee participants in Cameroon, age > 40 years was identified as a risk factor for HIV infection (OR 3.3 [CI 1.3 to 8.9]; p = 0.01) [66]. In a survey study of 605 conflict-affected women aged 13–49 in Northeastern Uganda, increased age was associated with reporting HIV-positive serostatus in two of the three districts included in the study [70]. Conversely, younger age was associated with higher prevalence of HIV in several studies [56, 62, 68, 72, 94]. For example, in a study examining HIV prevalence and associated factors in conflict-affected sex workers in Uganda, newly diagnosed women were younger compared to women with a previous HIV diagnosis (median age: 24 (Inter-Quartile Range (IQR): 20–26) vs. 25 (IQR: 23–28), p = 0.07) [56]. In a Pakistani study of IDP, the HIV positivity rate was higher among individuals aged 18–30 years (0.7%) compared to individuals aged 31–40 years (0.3%) [73]. In a cross-sectional study of HIV prevalence and correlates among displaced pregnant women in Mozambique [53], participants who were HIV-positive reported a slightly younger mean age compared to those without HIV (24.3 vs. 25.5); however, the difference was not statistically significant.

Gender

The role of gender in HIV acquisition in humanitarian settings were examined in 24 studies [50, 51, 61,62,63,64, 66,67,68,69, 71,72,73,74,75,76,77, 79, 82, 88, 90, 93,94,95]. Nineteen studies reported on HIV prevalence and correlates [50, 51, 61, 62, 64, 66, 68, 69, 71, 72, 74,75,76,77, 79, 82, 90, 93, 94], 4 studies on determinants of HIV serostatus [67, 73, 88, 95], and 1 study on HIV incidence [63]. The prevalence of HIV appeared to be higher among females than males, as observed in multiple studies conducted among conflict-affected individuals and refugees in Uganda [68, 74, 76, 79], the Democratic Republic of Congo (DRC) [72], Sudan [62], and Lesotho [94]. Being female aged 15–19 years living in drought-affected areas in Lesotho was associated with a higher HIV prevalence compared to their male counterparts in rural areas without drought (OR 2.77 [CI 1.19 to 6.47]; p = 0.02), although this association was not significant for females of older age groups in the study (20–59 years) [94]. These findings were in line with research showing that female refugees and asylum seekers in the UK and Germany had increased odds of reporting HIV infection [50, 88]. In a large modeling study of 65 countries, being impacted by drought significantly increased the proportion of HIV burden among women relative to men [93]. In contrast, studies in Pakistan involving IDP [73] and refugees [64], respectively, indicated a higher prevalence of HIV among males compared to females. Similarly, in Nigeria, adult females affected by conflicts had a 43% lower likelihood of testing HIV-positive than adult males [51].

Location

Sixteen studies examined the impact of location on HIV acquisition by considering factors such as urban or rural settings [51, 86, 90, 94], specific districts [55, 62, 64, 67, 68, 70,71,72, 74, 76, 77, 80], conflict or non-conflict areas [51], and residing inside or outside protected IDP camps [55]. Fourteen studies focused on HIV prevalence and correlates [51, 55, 62, 64, 68, 71, 72, 74, 76, 77, 80, 86, 90, 94] and 2 studies on associated factors of HIV-positive serostatus [67, 70]. A study conducted in Nigeria found that the prevalence of HIV was significantly higher in conflict than non-conflict areas (1.75% vs. 1.0%; p < 0.01) [51]. Regarding correlates of HIV serostatus, rural dwellers in Malawi were found to be less likely to test HIV-positive than urban dwellers (OR 0.72 [CI 0.53 to 0.98]; p < 0.05) [51]. The same study also found that individuals living in conflict zones were nearly three times as likely to test positive for HIV (OR 2.80 [CI 2.08 to 3.60]; p-value not reported). In a nationally representative sample in Lesotho, females aged 15–19 years old living in a rural area affected by severe drought had almost threefold higher odds of HIV infection (OR 2.77 [CI 1.19 to 6.47]; p = 0.02) compared to their counterparts in rural areas without drought [94]. In Northern Uganda, a study among conflict-affected women revealed that residing in protected camps for IDP was associated with a higher risk of HIV infection among women who had higher education (Adjusted OR (AOR) 2.29 [CI 1.30 to 4.04]; p = 0.004) and women whose occupations (AOR 3.62 [CI 1.32 to 9.91]; p = 0.01) or partners’ occupations (AOR 2.38 [CI 1.60 to 3.53]; p < 0.001) were considered modern (i.e., not an agricultural worker or a housewife) [55]. According to the authors, both higher education and non-traditional occupation were likely to be associated with increased mobility that may heighten the exposure to HIV among women living inside the camps [55].

Place of Birth or Origin

Eight studies examined the effect of place of birth or origin on HIV acquisition [50, 57, 60, 74, 82, 84, 93]; with 7 studies identifying HIV prevalence and correlates [50, 57, 60, 74, 82, 84, 88, 93] and 1 study reporting factors associated with HIV-positive serostatus [88]. Two infectious disease testing studies conducted in large cohorts of refugees and asylum seekers in Germany and the United Kingdom found that individuals of SSA origins represented the highest number of HIV cases, reflecting the higher HIV prevalence in the region [50, 88]. In Germany, HIV prevalence was mainly observed among refugees and asylum seekers of Nigerian origin (14.7% of 15,137 individuals) [50]. In the United Kingdom, refugees with a SSA nationality had significantly higher odds of being HIV positive (AOR 51.72 [CI 20.67 to 129.39]; p < 0.001) compared to those with different nationalities [88]. In a single-site study among men who have sex with men (MSM) in Lebanon, individuals who were Lebanese-born were more likely to test HIV-positive than MSM born in in Syria and other countries (AOR 0.14 [CI 0.04 to 0.51]; p < 0.05) [60]. Similarly in a Ugandan HIV testing study [74], refugees were less likely to be diagnosed with HIV compared to Ugandan nationals (AOR 0.27 [CI 0.21 to 0.34]; p < 0.001) that may indicate a lower HIV prevalence in the refugees’ countries of origin compared to Uganda.

Ethnicity

The relation between ethnicity and HIV acquisition was examined in 2 studies on HIV prevalence and correlates [57, 80], 1 study on HIV incidence [63], and another study on risk factors for HIV-positive serostatus [89]. Two studies in Uganda considered ethnicity as a correlate of HIV infection and incidence among conflict-affected female sex workers (FSW) and persons, respectively [57, 63]. One study suggested that FSW of the Acholi ethnic group might have a lower likelihood of HIV infection compared to other tribes although it was not statistically significant (OR 0.51 [CI 0.25 to 1.08]; p = 0.08) [57], and the other study did not report an association [63]. In a study examining the impact of armed conflict across the 177 member states of the World Health Organization (WHO), the authors found that ethnic heterogeneity (i.e., racial, linguistic, and religious characteristics) directly increased longer-term HIV morbidity and mortality, suggesting inequitable access to HIV prevention and treatment due to discrimination and political power among minorities [89].

Evidence of Modifiable Factors for HIV Acquisition in Humanitarian Settings

Of the 49 studies included in the analysis, 47 (95.9%) presented information on the effects of modifiable factors on HIV prevalence, incidence, or serostatus in the context of humanitarian crises (Table 2). These factors were further classified into five categories including policy and structural (18 factors), sociocultural (9 factors), health and mental health (11 factors), sexual practice (16 factors), and exposure to humanitarian crisis-related traumatic events (6 factors).

Policy and Structural Factors

Thirty-five studies examined 18 policy and structural factors affecting HIV acquisition in humanitarian settings, including education level [53, 55,56,57, 63, 68, 74, 76, 77, 80, 90, 91, 93, 94], employment and occupation status [55, 58, 64, 66, 81, 90, 91], income level [20, 49, 51, 56, 57, 60, 94], food and/or water insecurity [76, 77, 90, 92, 94], mobility status [53, 56, 57, 62, 65, 67, 68, 71, 74, 79, 84, 95, 96], history of travel (e.g., out of country, outside of camp) [64, 72, 73, 80, 94], access to health care (e.g., access to health facilities, distance to clinic) [63, 67, 68, 74, 79, 91, 92], access to HIV prevention and care [56,57,58, 80, 91, 92], access to HIV information and education [91, 92], HIV knowledge level [54, 80], housing status [56, 60], stigma [54, 91, 92], country-level health expenditures [89, 93], number of trained health workers [93], conflict time period (i.e., pre-conflict, during conflict, post conflict) [20], conflict intensity (i.e., < 25 deaths, 25–1000 deaths, > 1000 deaths) [20], conflict type (i.e., extra-systemic, interstate, internal, international war) [20], and criminal justice involvement [57].

At the individual-level, policy and structural factors were found to have effects on HIV prevalence or incidence, or various degrees of association with HIV serostatus in humanitarian settings. Two Ugandan studies examining the impact of war on HIV prevalence in the same samples of 400 FSW reported that around one-third of participants (33.8%, n = 135) were HIV positive; of whom 45 (33.3%) women were newly diagnosed or previously undiagnosed with HIV [56, 57]. The researchers noted that FSW diagnosed with HIV infection were significantly more likely to have been incarcerated (33.3% vs. 23%, OR: 1.67 [CI: 1.06 to 2.64); p = 0.03) [57], and that those with new HIV diagnosis had higher odds of reporting recent unstable housing (AOR: 2.48 [CI: 1.05 to 5.85]; p < 0.05) [56]. With respect to education, 8 out of 14 studies presented evidence on the significant effect of school attendance or completion on HIV prevalence, incidence, or serostatus. A study of 12,877 drought-affected individuals in Lesotho reported that attending secondary or higher education was protective against HIV infection among male and female youths aged 15–24 years (AOR 0.66 [CI 0.47 to 0.92]; p = 0.02) and 25–29 years (AOR 0.54 [CI 0.46 to 0.62]; p < 0.001) [94]. Similarly, among 384 conflict-affected youths living in transit camps in Northern Uganda where the HIV prevalence rate was 12.8%, HIV-positive individuals were significantly less likely than HIV-negative participants to currently be in school (12.2% vs. 26.9%, p = 0.03) [76]. Moreover, a modeling study of the effects of famine on HIV prevalence in Malawi suggested that a woman who worked as a farmer, was under 25 years of age, and had less than a primary education, had a 9.9% probability of being HIV positive [90]. In terms of income, four studies measured income by individual level (i.e., monthly [49, 60] and weekly [56, 57]), two studies by household level [51, 94], and one study by country level as defined by the World Bank [20]. At the individual level, among 292 MSM including Syrian refugees in Lebanon with an estimated HIV prevalence ranging from 13% to 14.9%, earning higher income was associated with HIV-positive status in bivariate analysis (OR 2.33, p = 0.03) [60]. This association did not remain significant in the adjusted regression model [60]. Among 928 conflict-affected Rwandan women of whom 75.9% were HIV positive, earning between 10,000 and 35,000 Rwandan Francs monthly was significantly associated with HIV-positive status compared to lower or higher income (81.5% vs 72.9% vs 75.2%, respectively) [49]. Lastly, in a modeling study estimating the effect of conflict periods on country-level HIV incidence in 36 SSA countries, low-income/lower-middle income nations had significantly positive associations with increased HIV incidence compared to upper-middle/upper income countries [20]. The estimates suggested an increase of 8.0 infections per 1000 people in HIV incidence on average per year in low-income countries (95% CI 5.19, 10.86), and an increase of 7.3 new infections per 1000 people per year in lower-middle income countries (95% CI 4.96, 9.56) [20].

Mobility status, an indicator to describe human movement due to livelihood, displacement, voluntary migration, or other factors [97], was measured in different ways in 13 studies. For example, three studies categorized mobility status by temporality (i.e., permanent, transient, displaced) [67, 68, 79] while others considered mobility according to legal status (e.g., IDP, refugee, local resident) [53, 62, 65, 71, 74, 84, 96]. Findings from a Canadian study of 15,269 refugee women and 126,506 non-refugee women estimated a higher HIV prevalence among refugees who initially sought asylum in a transition country before coming to Canada (i.e., secondary migration) (1.47%) than refugees with primary migration who directly arrived in Canada (0.82%) [84]. The study also reported that refugees with secondary migration were 68% more likely to have HIV compared to refugees with primary migration (AOR 1.68 [CI 1.04 to 2.71]; p < 0.05) [84]. In two studies conducted in conflict-affected areas in the DRC, HIV prevalence was significantly higher among IDP when compared to the local population [65, 71]. A study of adult females living in an IDP camp and host communities near the Congo River found that while HIV prevalence among the river populations was similar to the national estimates at 3.1% (95% CI 2.1, 4.1), prevalence was higher in the IDP population at 7.6% (95% CI 4.1, 11) [65]. In other studies involving conflict-affected individuals in Mozambique [53], Uganda [56, 57, 67, 68, 79], and Sudan [62], no effect was found between mobility status and HIV prevalence, incidence, or serostatus. Relevant to mobility status, one modeling study of multi-country data investigated the potential impact of conflict on HIV incidence rate by considering conflict intensity (i.e., < 25 deaths, 25–1000 deaths, > 1000 deaths), conflict time period (i.e., no conflict, pre-conflict, during conflict, post-conflict), and conflict type (i.e., extra-systemic, interstate, internal, international war) [20]. The study reported that HIV incidence was significantly higher in the 5 years prior to conflict with an increase of 2.1 per 1000 infections per year, with no significant increase during and immediately following periods of conflict [20]. The authors surmised that their findings support prior evidence on stabilized HIV infection rates during times of conflict that were likely due to limited access to HIV testing and reporting [20].

At the community level, access to health care facilities and programs were found to affect HIV correlates of HIV-positive serostatus in humanitarian settings. Studies have examined the role of access to healthcare including distance to health facilities [74], access to HIV prevention and care [20], and access to HIV information and education [56]. A study conducted in a Ugandan refugee settlement of 7766 individuals of whom 330 (4%) tested HIV-positive revealed that residing more than one hour from the clinic was associated with increased likelihood of HIV infection (Adjusted Relative Risk 1.39 [CI 1.11 to 1.74], p = 0.003) [74]. Two studies of conflict-affected FSW in Uganda examined the association of having access to HIV education programs, such as condom demonstrations, with HIV serostatus [56, 57]. One study reported that receiving a condom demonstration was negatively associated with new HIV diagnosis among conflict-affected FSW (AOR 0.39 [CI 0.15 to 0.97]; p < 0.05) [56], but no difference was found when comparing between individuals with and without HIV serostatus in the same samples [57]. In a qualitative study in Namibia, IDP and key stakeholder participants conveyed that lack of access to family planning, condoms, and HIV testing due to flooding could increase the risk of HIV acquisition among affected individuals [92]. Similar observations were described in an interview study of Colombian IDP and stakeholders working with displaced MSM [91]. Furthermore, stigma concerning PWH, gender non-conformity, and homosexuality was perceived as a factor affecting HIV acquisition that could cause further isolation of internally displaced individuals of sexual minorities [54, 91, 92]. Studies also reported the link of food and water insecurity with HIV prevalence and correlates in humanitarian settings. In a modeling study of HIV prevalence in famine-affected Malawi, increased HIV prevalence was positively associated with hunger in rural areas, attributing to hunger-induced exchanging sex for food or money and the impact of hunger on immune function [90]. In two studies of conflict-affected young men and women in Uganda of which the overall HIV prevalence rate was estimated at 12.8%, HIV-positive individuals were significantly more likely than HIV-negative participants to indicate lacking food and/or water (57.1% vs. 37.3%, p = 0008) [76, 77].

At the country level, a modeling study analyzing the impact of drought on HIV burden in 69 countries revealed that health expenditures were not associated with women’s proportion of HIV cases; however, countries with a larger number of trained health workers were more likely to have a lower proportion of women with HIV and lower rates of HIV prevalence [93]. In another modeling study that examined the association between armed conflict and HIV morbidity and mortality in 177 countries [89], the researchers considered total HIV spending, number of refugees, asylum seekers, and IDP, and number of persons on ART as factors representing a conflict-associated vulnerability construct. The findings suggest that the conflict and HIV association was moderated by these factors, highlighting the need of HIV prevention and care among the displaced populations to reduce longer-term morbidity and mortality [89].

Sociocultural Factors

Twenty-four studies investigated 9 sociocultural factors and their influence on HIV prevalence, incidence, and HIV-positive serostatus in humanitarian settings. These factors include marital status [53, 55, 58, 62,63,64, 66,67,68, 72, 74, 76, 77, 79, 81, 86, 94, 95], number of marriages [80], religious beliefs [58, 68, 79, 93], staying in a child-headed household [68, 79], having a female household head [79], family separation due to sexual identity and HIV-positive status [91, 92], disruption of social networks [91, 92], living with someone else and not in their own house [49], and number of other household members [49]. Regarding marital status as a correlate of HIV infection, there was evidence that the risk of acquiring HIV was higher among individuals who were divorced, separated, widowed, or single (i.e., never married) [55, 62, 74, 86]. In a Sudanese study, conflict-affected women who were separated, divorced, or widowed had a higher likelihood of HIV infection compared to women in married and monogamous relationships (AOR 7.7 [CI 2.3 to 25.4]; p < 0.001) [62]. In contrast, some studies reported a higher HIV seropositivity rate among individuals who were married compared to those who were not married [64, 68, 76, 77, 94]; such as in a Ugandan cohort study where compared to individuals who never married, those who were married (AOR 4.69 [CI 3.25 to 6.76]; p < 0.001), separated or divorced (AOR 9.17 [CI 5.60 to 15.00]; p < 0.001), or widowed (AOR 20.35 [CI 10.27 to 40.34]; p < 0.001) were significantly more likely to be HIV-positive [68]. Additionally, among women living in a refugee camp in Thailand, women with HIV were more likely to report having had a greater number of marriage partners than HIV-negative women (OR 1.4 [CI 1.1 to 1.8]; p < 0.001) [80].

Two studies in Uganda of conflict-affected individuals reported higher HIV prevalence among women than men (14.6% vs. 8.5%, p < 0.001 and 17.2% vs. 10.6%, p < 0.001, respectively) [68, 79]. The authors reported that individuals who lived in a female-headed household had increased odds of HIV infection, although no association was found between staying in a child-headed household and HIV-positive serostatus [68, 79]. In two qualitative studies involving persons affected by flood in Namibia [92] and IDP in Colombia [91], participants expressed that family separation and disruption of social networks could increase the risk of HIV acquisition due to loss of social support. Lastly, regarding religious beliefs, higher percentage of Muslim population was a predictor of a lower HIV prevalence in drought-affected populations in 69 countries that may be attributable to the emphasis of conservative sexual norms, such as promoting abstinence before marriage [93]. On the other hand, female refugees who were interviewed in a South African qualitative study believed that their religious beliefs prohibited condom use and thus, may increase the women’s risk of acquiring HIV [58].

Health and Mental Health Factors

Eleven health and mental health factors impacting HIV prevalence, incidence, and serostatus were reported in 29 studies. The factors identified were current or past STI (non-HIV) diagnosis [49, 53, 56, 61,62,63, 65, 66, 68, 69, 71, 76, 77, 79, 80, 88, 91, 95], substance use (including illicit drug use, injecting drug use, needle sharing) [56, 59, 64, 66, 73, 76, 80, 83, 88, 89, 95], alcohol use [54, 56, 76, 88, 89, 92, 95], mental health conditions (e.g., PTSD, depression, suicide ideation) [49, 54, 59, 63, 67, 68, 79], blood transfusion or donor [53, 66, 71, 72, 80, 86, 88], other diseases (e.g., malaria, hepatitis A, tuberculosis) [71, 72, 92, 95], lifetime pregnancies [55, 56, 80], history of tattoos [66, 80, 88], dental procedure [66, 73], history of abortion [66], and surgery and hospitalizations [66]. Several studies conducted in humanitarian settings reported evidence of linkage between diagnosis of other STI (e.g., syphilis, herpes simplex virus (HSV)-2, genital ulcers) and HIV infection, as indicated by serological testing and/or self-report [49, 53, 61,62,63, 65, 68, 69, 76, 77, 79, 88, 91]. In a mixed methods study of conflict-affected persons in Ethiopia, FSW with HSV-2 infection were more likely to have an HIV diagnosis compared to those without (48% vs 27%, Risk Ratio (RR) 1.8 [CI 1.2 to 2.8]; p < 0.005) [61]. Other studies also reported a link between a previous STI diagnosis and the likelihood of having HIV-positive diagnosis among conflict-affected individuals [49, 53, 65, 76, 77]. The effect of alcohol and substance use on HIV acquisition in humanitarian settings was also examined in several studies. In a modeling study of 177 WHO member states, alcohol consumption, illicit drug use, and injection drug use moderated the relationship between conflict and HIV prevalence, suggesting HIV-related risk behaviors in conflict situations and in the presence of high substance use would increase the population’s risk of HIV infection [89]. Furthermore, in a Ugandan study of conflict-affected FSW, women newly diagnosed with HIV were more likely to report heavy drug or alcohol use (AOR 5.73 [CI 2.22 to 14.78]; p < 0.05) [56]. Studies have also reported mental health conditions including probable depression, PTSD, and suicide ideation to be significantly associated with HIV infection among different conflict-affected groups including military personnel in South Sudan [54] and Rwanda [59], men and women in Uganda [63, 67, 68, 79], and women in Rwanda [49]. In a study of male military personnel in conflict-affected Rwanda, HIV prevalence was higher among those who screened positive for either PTSD (3.9% vs. 2.7%) or depression (3.6% vs. 2.4%); however, the difference was not statistically significant [59]. In a Ugandan study of conflict-affected men and women, women who tested positive for HIV had 90% increased odds of probable depression (AOR 1.90 [CI 1.29 to 2.80]; p = 0.001), although no such association was found among male participants [67].

Sexual Practice Factors

Twenty-nine studies documented the impact of 16 sexual practice-related factors on HIV acquisition in humanitarian settings, namely condom use [49, 53, 54, 56, 57, 60, 62, 63, 68, 69, 76, 77, 79, 94], number of sex partners [49, 53, 54, 58, 60, 63,64,65,66, 68, 69, 79, 80, 86], exchanging sex for money or nonmonetary items [49, 53, 58, 60, 70,71,72, 76, 77, 79, 80, 91, 92, 94], buying or purchasing sex [61, 68, 69, 72, 80, 86], age of sexual debut [63, 65, 76, 77, 79, 81], coerced sexual debut [63, 67, 75, 76, 79], first sexual partner ≥ 10 years older [63, 75, 76, 79], circumcision status [53, 68, 79], perceived ability to protect oneself from HIV/STI [76, 77, 79], knowing sexual partner’s HIV status [76, 77, 79], MSM [64, 95], practice of dry sex (i.e., sexual intercourse when the vagina is dry upon penetration) [76, 77], having sex while intoxicated or high on drugs [60], intergenerational sex (i.e., sexual relationships with a partner ≥ 10 years older) [94], female contraceptive use [93], and partner infidelity [58].

Condom use measurements were reported in various dimensions across studies; for example, by considering a specific timeframe (e.g., last sex, past 6 months, ever), type of sexual partner (e.g., regular or non-regular partner, last 3 partners, sexual debut), and consistency (e.g., inconsistent condom use coded as a dichotomous yes or no response). Some studies, such as two Ugandan studies assessing the correlates of HIV infection among conflict-affected young people, employed multiple condom use measures including ever use of condom (yes or no), condom use during last sex (yes or no), and condom use consistency (always vs. sometimes or never) [76, 77]. None of the condom use measures were significantly associated with HIV infection in bivariate analysis; except among young men who were HIV-negative and more likely to report using a condom during last sex compared to men who tested HIV-positive (37.8% vs. 16.7%, p = 0.05) [77]. The association was not statistically significant in a multivariable model [77]. Among Rwandan conflict-affected women, those indicating not using condom at least half the time (~ 50%) in the last 6 months were more likely to report an HIV diagnosis than women reporting more frequent condom use (66.1% vs 33.9%, p < 0.001) [49]. The authors also found that HIV-positive serostatus was an independent factor of condom use at least half the time in the last 6 months among sexually active women in the study (p < 0.001) [49]. A study among refugees in Cameroon found that individuals who reported multiple sex partners had higher odds of HIV infection than their peers who did not (OR 4.3 [CI 1.5 to 11.8]; p = 0.01) [66]. A study among 2008 conflict-affected persons in Uganda examined several correlates of HIV infection [79]. Among conflict-affected women, sexual practice-related factors associated with HIV infection included age younger than 18 at sexual debut (AOR 2.00 [CI 0.91 to 4.38); p = 0.08), first sexual partner ≥ 10 years older (AOR: 1.69 [CI 1.07 to 2.67]; p = 0.03), reporting 2 (AOR: 2.54 [CI 1.23 to 5.23]; p = 0.02) or 3 + (AOR: 4.65 [CI 2.65 to 8.18]; p < 0.001) sexual partners in the past year, inconsistent condom use with last 3 partners in the past 12 months (AOR: 0.40 [CI 0.29 to 0.57]; p < 0.001), and exchanging sex for money or other items in the past year (AOR: 5.51 [1.76 to 17.31]; p = 0.01) [79]. Among conflict-affected men, not using a condom at sexual debut was significantly associated with HIV infection as a sexual practice factor (AOR: 1.92 [CI 1.30 to 2.83]; p = 0.006) [79]. Use of female contraceptives, such as oral contraceptives, was found to be significantly protective against HIV acquisition in a modeling study of 65 countries experiencing drought [93]. In a qualitative study of female refugees in South Africa, participants noted factors that could increase their susceptibility for HIV infection including partner infidelity and low perceived ability to protect oneself from HIV/STI [58]. In Lesotho, reporting intergenerational sex with a partner ≥ 10 years older was found to be significantly associated with HIV infection among drought-affected persons aged 15–24 years (AOR 1.54 [CI 0.97 to 2.44]; p < 0.001) and 25–29 years (AOR 1.37 [CI 1.10 to 1.71]; p = 0.007), respectively [94]. In a study involving conflict-affected MSM in Lebanon, reporting multiple sex partners (p = 0.02), condomless sex (p = 0.008), exchanging sex for money or other items (p = 0.01), and having sex while intoxicated or high (p = 0.02) in the past year were identified as factors associated with HIV infection [60]. Notably, a study among conflict-affected female adults in Uganda did not find exchanging sex for money or other items to be associated with HIV positive serostatus [70]. In two Ugandan studies involving the same sample of 384 conflict-affected youths aged 15–29, the practice of dry sex was a factor independently associated with HIV infection (AOR 2.31 [CI 1.04 to 5.13]; p < 0.05) [76] and specifically among young women (AOR 7.62 [CI 1.56 to 16.95]; p < 0.05) [77]. Further, both studies found that HIV-positive participants were significantly less likely than those who were HIV-negative to know their partner’s HIV status (AOR 0.50 [CI 0.22 to 1.12]; p < 0.05) [76] and when analyzed by gender, i.e., young men (AOR 0.40 [CI 0.08 to 1.41]; p < 0.05) and young women (AOR 0.57 [CI 0.19 to 1.77]; p < 0.05) [77].

Exposure to Humanitarian Crisis-Related Traumatic Events

Nineteen studies reported 6 factors concerning humanitarian crisis-related traumatic events that affect HIV acquisition such as experience of rape, sexual violence, or other sexual trauma during conflict or displacement [49, 52, 53, 65, 67, 68, 76, 77, 79, 82, 85, 87, 91, 92], experience of abduction or kidnapping [57, 63, 68, 75,76,77, 79], experience of traumatic events [63, 67, 68, 79], night commuting (i.e., traveling at night to sleep elsewhere than family home for security and privacy reasons) [76, 77], experience of torture [88], and military recruitment [78]. Several modeling studies examined the impact of widespread rape or sexual violence in conflict-affected SSA countries on HIV prevalence and/or incidence [52, 82, 85]; such as a modeling study of six countries with a recent period of armed conflict that estimates increased HIV incidence by 10% after accounting for a high prevalence of rape, the effect of genital injury on HIV transmission, the perpetrator’s HIV serostatus, and underlying HIV incidence [85]. Findings from these studies suggest that while widespread rape did not seem to have significant impact on HIV prevalence [52, 82, 85], it might contribute to increased HIV incidence at a community level when the underlying HIV incidence was low and where systematic or mass rape had occurred during conflict [82, 85]. Additionally, reporting experience of rape or sexual violence was significantly associated with HIV infection among the IDP in the DRC [65], young people living in post-conflict transition camps in Uganda [76, 77], and conflict-affected persons in Uganda [79]. Experience of rape was also described in qualitative studies involving persons affected by flood in Namibia [92], displaced gay, bisexual, and transgender persons in Colombia [91], and conflict-affected women in the DRC [87], citing an increased risk of acquiring HIV during flooding and armed conflict when unstable situations occurred resulting in rape and violence.

According to a study on HIV incidence and correlates in Uganda where the overall HIV incidence rate was 10.1 (95% CI 7.2, 13.8) per 1000 person years, individuals who had experienced abduction during conflict had a significantly higher odds of HIV infection, with a 3.2 times greater likelihood compared to those who did not experience abduction (Hazard Ratio (HR) 3.23 [CI 1.64 to 6.34]; p < 0.001) [63]. In another Ugandan study among 384 conflict-affected persons, a significantly larger proportion of former abductees compared to non-abductees reported sexual practices that could increase their risk of HIV infection such as experiencing non-consensual sexual debut (30.5% vs. 11.2%, p < 0.001), having a first sexual relationship with a partner ≥ 10 years older (26.3% vs.12.6%, p < 0.001), having a greater median number of sex partners in the previous six months (2 vs. 1, p = 0.02), and ever use of condom (63.6% vs. 50.5%, p = 0.02) [75]. In several studies, conflict-affected individuals in Uganda who had undergone ≥ 12 traumatic events related to war were more than twice as likely to have HIV diagnosis as those who did not report traumatic events [63, 68, 79].

Implications for Future Programs

In order to effectively address HIV acquisition in humanitarian settings, evidence from ten studies highlighted the importance of systematic screenings, prevention, and treatment for HIV and other infectious diseases among people affected by complex emergencies or experiencing mobility, and recognizing the differences between the groups (e.g., refugees, asylum seekers, IDP, and migrants) and their specific health needs and experiences [50, 53, 62, 64, 66, 74, 80, 84, 88, 95]. An HIV screening intervention study in a refugee settlement in Uganda found that Ugandan nationals were more likely to be HIV-positive than refugees [74]. Similar findings were observed in an epidemiologic study of HIV and hepatitis B and C infections among refugees in Cameroon [66] and an HIV prevalence study of Lebanese and Syrian-born MSM in Lebanon [60]. Studies conducted in countries that have resettlement programs for refugees such as Germany, Canada, the US, and the UK highlight that prevalence of HIV varies among refugees based on individual-level risk factors and prevalence in the country of origin [50, 84, 88, 95]. Authors of these studies emphasized the importance of tailoring HIV screening and prevention programs for different groups and needs, and training providers to create culturally relevant healthcare approaches, particularly when addressing behaviors considered as illegal or stigmatized in the country of origin such as illicit drug use, sex work, or same-sex sexual relationships [88, 95].

Six studies in the present review examined the impact of widespread rape and sexual violence in conflict settings on HIV prevalence and incidence [20, 49, 52, 82, 85, 87]. Evidence from a modeling study involving seven countries in SSA showed that widespread sexual violence might not have a direct population-level impact on HIV prevalence in conflict-affected settings [52]. Nevertheless, the authors underlined that consequences of sexual violence such as physical and psychological trauma could not be understated. Improved surveillance for sexual violence and identifying its drivers in conflict settings were recommended to understand the impact on HIV and STI epidemiology at the population- and individual-levels [52, 85]. Researchers in Rwanda endorsed interventions that address the specific needs of women with a history of sexual violence in conflict settings, including implications for long-term reproductive health and susceptibility to high-risk sexual behaviors [49]. Others suggested integrating HIV prevention services and psychosocial support interventions that provide access to emergency contraception and HIV post-exposure prophylaxis to prevent unwanted pregnancies and HIV transmission [82, 87], support the needs of young people (aged 15–29) who survived sexual violence and abduction [75,76,77], and promote women’s empowerment [49, 85].

Findings from five studies indicated associations between reporting depression and PTSD symptoms with HIV diagnosis [49, 59, 67, 68, 79], highlighting the need for evidence-based strategies to enhance mental health screening and treatment for people affected by conflicts. HIV prevention and treatment interventions also need to address war trauma experiences among this population and provide a gender-sensitive programming and trauma-informed care to foster resilience and facilitate healing at the individual, family, and community levels [67, 68, 79]. Strategies to screen and treat alcohol and other substance use are also needed to reduce HIV acquisition risk, in particular among military communities [54, 59] and people who inject drugs [83] in conflict settings.

Informed by findings from a qualitative study involving PWH, service providers, and other stakeholders, an HIV and flooding framework was developed to demonstrate health-related vulnerabilities in the context of flooding in Namibia [92]. The framework represents the interlinkages between the livelihood loss caused by flooding (e.g., destruction of agriculture, lack of shelter), the emergence of flood-related diseases (e.g., malaria, cholera, malnutrition), the limited accessibility of health services including HIV prevention and treatment (e.g., condoms, ART), and the increased risk of HIV acquisition due to contributing factors (e.g., violence, alcohol use, stigma, displacement). The researchers endorsed adapting and applying the framework in other disaster conditions to assess capacity and vulnerability in emergency response plans and potentially reduce HIV risk acquisition and disease progression [92].

Implications for Future Research

Recommendations for future research were proposed in several studies [20, 51, 52, 54,55,56,57, 59, 60, 65, 70, 71, 76, 78, 82, 84, 87, 93,94,95,96]. Regarding methodology, studies using structural equation modeling could refine the existing theoretical models of HIV acquisition in humanitarian settings and identify the role of multiple risk factors for HIV morbidity and mortality [89]. Modeling studies could also be useful to facilitate rigorous hypothesis testing to examine the relationship between humanitarian crises and HIV cross-nationally, adjusting for economic, social, political and cultural factors [20, 93]. The use of qualitative and mixed methods studies was recommended to provide insights into the lived experiences and perceptions related to HIV acquisition in humanitarian settings [57, 76, 87, 93]. A modeling study investigating the impact of drought in multiple SSA countries recommended employing qualitative research to further understand the relationship between drought and HIV vulnerability among women by considering multiplicity of experiences according to poverty, race, religion, and disability [93]. Longitudinal and mixed methods studies could also be useful to understand the potential causal pathways between criminal justice system involvement with HIV risks, and to inform the development of HIV and sexual reproductive health interventions that promote health and human rights [57]. Additionally, studies recommended standardization of definitions (e.g., transactional sex work), measurements of risk behaviors, and surveillance methodologies to ensure precision and reproducibility [52, 70].

Nine studies suggested further research on the impact of complex emergencies on HIV prevalence [54, 55, 65, 71, 76, 78, 82, 84, 96], specifically by identifying individual- and community-level factors associated with HIV acquisition [54, 76, 82, 84, 96], engaging displaced populations and host communities [55, 71], and determining national-level HIV prevalence [71, 78] both during and after the conflict period [65]. Lastly, two studies recommended exploring the effects of complex emergencies on HIV care cascade including linkage to care, ART adherence, resistance, and toxicity [56, 95].

Discussion

This scoping review provides a comprehensive summary of a range of factors associated with HIV acquisition in the context of humanitarian crises caused by natural disasters and complex emergencies resulting from conflict. We identified a total of 49 relevant studies, primarily conducted in SSA, with a notable concentration in Uganda. The amount of published evidence on HIV acquisition in humanitarian settings based on geographical locations gave the general impression that such data were abundant in the African region, and less so in Asia and Europe where humanitarian crises also occurred. The fact that the HIV epidemic has had the greatest impact on SSA [98] and the majority of individuals experiencing displacement reside in SSA [1] might explain the high number of studies conducted in this region. It is important to note that the UNHCR acknowledged the rising number of refugees in European countries due to the ongoing Russia-Ukraine war that has created the fastest and one of the largest displacement crises, from 7 million displaced people at the end of 2021 to 12.4 million at the end of 2022 [1]. In the current review that is inclusive of literature published between January 1990 and March 2022, the lack of published research on HIV acquisition in the context of Russia-Ukraine war can be attributable to the complexity in conducting research amid an ongoing humanitarian crisis where the conditions are often unstable and can present risks to both researchers and research participants [99].

Across the 49 studies included, the majority employed a quantitative approach and a cross-sectional design. Around half of the studies were designed to investigate the prevalence or incidence of HIV and risk behavior correlates of HIV infection in humanitarian crises. Regarding the types of humanitarian crises, complex emergencies were the most reported setting in over 80% of the studies, followed by natural disasters, and crises that were considered both complex emergencies and natural disasters. The review identified five salient non-modifiable factors associated with HIV acquisition in humanitarian settings across 36 studies, including age, gender, location, place of birth or origin, and ethnicity. Both older and younger age groups were associated with an increased risk of HIV infection in humanitarian settings. As for gender, the prevalence of HIV appeared to be higher among females than males. Residing in some locations, such as rural areas affected by crises, was associated with HIV-positive serostatus. Regarding place of birth or origin, there was some evidence of a higher HIV prevalence among refugees and asylum seekers with SSA origins. Lastly, the four studies investigating the role of ethnicity in HIV acquisition reported varying results.

The present review also identified 60 factors linked with HIV acquisition that could potentially be modified through policy or intervention. These factors were examined in 47 out of 49 studies and categorized into five domains, including policy and structural, sociocultural, health and mental health, sexual practice, and exposure to humanitarian crisis-related traumatic events. The policy and structural factor most often examined in the included studies was education level, with evidence of being in school or completing school was associated with lower odds of HIV infection among persons affected by humanitarian crises. The sociocultural factor most often associated with HIV acquisition was marital status, suggesting that individuals who were married had a lower risk of HIV infection compared to those who were divorced, separated, or widowed. The health factor most associated with HIV infection was current or past STI diagnosis, while substance use was identified as a mental health factor predicting a higher risk of HIV infection. Sexual practice factors most often associated with HIV infection were inconsistent condom use, exchanging sex for money or nonmonetary items, and reporting a higher number of sex partners, respectively. Lastly, experience of rape or sexual violence during conflict or displacement was most frequently identified as correlate of HIV infection in the included studies.

Our analysis of the studies highlights various multilevel factors that contribute to the risk of acquiring HIV in humanitarian settings. Understanding the interlinkages of these factors can help researchers, service providers, and others to identify the needs of the population affected by humanitarian crises related to HIV prevention; develop priority research questions and methodological approaches that are ethical, appropriate, and adaptable for humanitarian crises; and adapt evidence-based HIV prevention interventions developed in non-crisis settings to implement in humanitarian settings [25]. Building on the findings of the present review and informed by relevant HIV literature [22, 30], we applied the social-ecological model to map the identified multidimensional factors associated with HIV acquisition in the context of humanitarian setting at the levels of individual, social and sexual networks, community, public policy, and the context of humanitarian crises (Fig. 3). Each dimension in the model presents a range of modifiable and non-modifiable factors that were most often investigated in the published studies to date. We also highlighted recommendations for programming and opportunities for research. The framework can guide data collection to further understanding of different factors contributing to HIV acquisition in humanitarian crises and the development of tailored interventions to reduce HIV prevalence and incidence in these settings by considering different contexts, for example, social and cultural norms that may influence sexual practices and acceptability of HIV prevention strategies.

Fig. 3
figure 3

Social-ecological model of the identified multidimensional factors associated with HIV acquisition at the levels of individual, social and sexual networks, community, public policy, and the context of humanitarian crises. This figure illustrates a range of modifiable and non-modifiable factors that were most often investigated in the published studies to date, with the adapted social-ecological model at the center. The outside boxes at the top (orange) present recommendations for future programming in line with the levels of the social-ecological model. The outside boxes at the bottom (blue) present opportunities for research including methodological approaches and topic areas (Color figure online)

Full size image

From a methodological perspective, the reviewed studies had several common limitations. We found that the sampling approaches in some studies were not clearly described, raising the possibility of selection bias. The reliance on self-reported data to document HIV risk factors might expose these studies to social desirability bias, with potential effects such as underreporting of sensitive behaviors (e.g., sexual practices) and experiences (e.g., sexual violence). With only two studies employing prospective cohort designs [63, 69], the predominance of cross-sectional designs in the included studies limits inferences about causality between risk factors and HIV infection. Our review also showed a noticeable lack of intervention studies, with only one study reporting an HIV screening program in a refugee settlement in Uganda [74]. Our findings highlight the need for researchers, practitioners, policymakers, and other stakeholders to identify and use standardized measures and screening tools of underlying factors that affect HIV vulnerability in humanitarian settings. Routine HIV surveillance efforts are critical for documenting the impact of humanitarian crises on the HIV epidemic and informing resource allocation, as well as prospective studies with a community-based participatory approach to develop meaningful HIV prevention and care interventions that are ethical, trauma-informed, and culturally relevant. Establishing HIV prevention and care plans for individuals who are most at risk for HIV and those diagnosed with HIV should consider the lived experiences of people affected by humanitarian crises and their perceived needs. Future efforts that examine the directionality of social determinants of health affecting HIV acquisition in this population are warranted. The lack of studies from South America and Asia, regions with a high HIV burden that have experienced humanitarian crises, highlights the need for more research to generate evidence on the impact of crises on HIV acquisition in these settings.

The present scoping review has several important strengths. First, our review provides a thorough synopsis of the global evidence on modifiable and non-modifiable factors associated with HIV acquisition among people affected by humanitarian crises caused by natural disasters and human-made complex emergencies. Second, we considered all research using any type of study designs and methodologies that were published in both peer-reviewed journals and grey literature. Third, our multidisciplinary review team members are trained as infectious disease epidemiologists or public health professionals or research librarians. Record screening, full-text review, and data extraction approaches were piloted for reliability and conducted by multiple reviewers who had independently completed each step and resolved disagreements through consensus. Fourth, we used a systematic, iterative, and reproducible approach to map the extent of the literature, determine knowledge gaps, and suggest implications for future practice and research. This review also has some limitations. First, non-English language articles were excluded due to time and human resource constraints. Second, while we reported on the individual study designs, we did not assess the quality of the included publications. Third, we did not characterize sources of funding or the role of funders in the included studies.

Conclusion

In conclusion, this scoping review systematically summarizes the current evidence and understanding of factors related to HIV acquisition in humanitarian settings, emphasizing the multidimensional risks and vulnerabilities faced by the affected populations. Existing literature demonstrates that non-modifiable and modifiable factors were found to be associated in varying degrees with HIV acquisition in humanitarian settings. Moving forward, investigations are warranted to further examine the directionality of the modifiable and non-modifiable factors affecting HIV acquisition and consider the multilevel barriers and facilitators to the uptake of HIV prevention strategies in the context of humanitarian crises. Intervention research aimed at promoting HIV prevention strategies should consider individual, social, community, and structural levels of risk factors and contexts and tailor to the needs of the affected populations in these settings.