Study design, population and setting

We performed a cross-sectional study from September 1, 2022, to September 30, 2023. Recruitment and data collections were performed by a multidisciplinary group comprising an infectious diseases specialist, an infectious diseases medical resident, two nurses, one cultural mediator, and one volunteer healthcare worker affiliated with the Italian non-governmental organization Doctors with Africa CUAMM (University College for Aspiring Medical Missionaries). This organization has been actively engaged in assisting the agricultural migrant worker communities since 2015. STI and KAP screening surveys were delivered concurrently by face-to-face interviews.

Study sites included four community health centers for homeless people situated in the City of Bari and five resettlement camps located in the municipality of Foggia, so-called ‘ghettos,’ where agricultural migrant workers (AMW) reside. A more detailed description of sites included in the study is provided in the supplementary file.

The eligible population included all adult individuals who were beneficiaries or residents in the mentioned facilities throughout the study period. No exclusion criteria were applied for this study. Participants had the option to withdraw from the study at any time, and their decision did not impact their medical care.

Blood test result delivery was scheduled approximately 10 days after screening, and the failure to return for results rate (FTR) was assessed. Participants who tested positive for any of the infections included in the study were referred for further diagnostic evaluation and treatment to the Clinic of Infectious Diseases of the Polyclinic of Bari. Positive-screened participants received the same standard of care, in accordance with national and international guidelines [10, 11, 12, 13].

Questionnaire development and data collection

Questionnaire development was informed by literature review and administered through face-to-face interviews conducted by nurses and trained doctors. Every interaction with the subjects involved in the study was facilitated by linguistic and cultural mediators, when needed. To minimize the social desirability bias, interviewers provided clear instructions at the beginning of the questionnaire, explaining the purpose of the survey, the importance of honest responses, and reassuring participants about data confidentiality.

The questionnaire was organized into the following sub-sections:

• – Socio-demographic information (age, education, occupation, marital status, how long they were homeless, possible employment contract, residence document in Italy): 6 questions.
• – Risk factors for STI: multiple sexual partners, use of condoms, smoking and alcohol habits, chem-sex (defined as “voluntary intake of psychoactive drugs with the intention of facilitating and/or enhancing sexual activity [14]”) and medical history (including symptoms of TB and other communicable-non-communicable diseases): 6 questions.
• – Knowledge, attitude, and practices on HIV, HBV, HCV, and Syphilis: 25 questions.

All study participants underwent the following blood tests:

• – Serologies for HBV (HBsAg, anti-HBs, and anti-HBc), HCV (anti-HCV), human immunodeficiency virus (HIV), and syphilis (VDRL and TPHA).
• – The study participants, resulting positive for the screening, performed confirmation tests according to the standard of care: HBV-DNA for HBV, HCV-RNA and genotype for HCV, HIV-RNA, immune cell count, and resistance profile for HIV.

All data were entered into a secured online platform (Kobo Ttoolbox [15]), and a quality control check of the data entry was performed before data analysis.

Statistical analysis

A descriptive analysis was performed to define the distribution of the characteristics of the sample. Continuous data were summarized as median and interquartile range (IQR). Return For Test Results (RTR) was identified as the dependent variable. World regions were classified according to the United Nation Geoscheme [16]. The distribution of continuous variables was assessed with the Shapiro-Wilk normality test, and the Mann-Whitney U test was used to compare groups for continuous variables. A chi-squared test (with the Fisher’s correction if less than five cases were present in a cell) was applied for categorical variables. A logistic regression model was implemented as follows. RTR was considered as a dependent variable and each of the available factors at the baseline evaluation as independent variables (univariate analysis). The effect sizes were reported as odds ratio (OR) with 95% confidence interval (CI). All tests were 2-sided, and a p-value <0.05 was considered statistically significant. The model included a set of clinically relevant candidate predictors and variables resulted to be significant at the bivariate analysis. Statistical analyses were performed using R Statistical Software (v4.1.3; R Core Team 2021) in R Studio Version [17].

Ethical considerations

Prior to recruitment, written informed consent (prepared in 4 languages: Italian, French, English, and Arabic) was obtained from all participants. The objectives of the study, along with methods used and data confidentiality, were explained with the aid of cultural mediators, when needed. The study has been approved by Ethics Committee ‘Azienda Ospedaliero-Universitaria Consorziale Policlinico,’ protocol number DG1563. During the screening process, we actively worked to establish and enhance personalized pathways for connecting individuals to healthcare through various social and healthcare services. We raised awareness about the potential progression of HIV and liver diseases, as well as the risk of transmission to others, by directing these individuals to the Infectious Disease Clinic at the Policlinic of Bari for counseling, diagnosis, and appropriate treatment.

Characteristics of the population

From September 1, 2022, to September 30, 2023, a total of 149 participants were recruited in the study, of which 64 (42.9%) were agricultural migrant workers and 85 (57.1%) were homeless, with a median age of 35 (IQR: 27–45) years. Characteristics of the sample stratified for study subpopulation and RTR are displayed in Table 1. Males accounted for 87.5% (n = 56) of agricultural migrant workers and 75.3% (n = 64) of homeless people. The most frequently reported comorbidities were diabetes (8.1%, n = 12), hypertension (6%, n = 9), and respiratory pathologies (4%, n = 6). Among agricultural migrant workers, countries of origin were located mainly in sub-Saharan Africa (64.1%, n = 41) and North Africa (20.3%, n = 13), while 32.9% (n = 28) and 31.8% (n = 27) of homeless people were originally from Sub-Saharan Africa and Italy. Educational level was low, with 64.4% (n = 96) of the included subjects reporting only primary school or no formal education. Overall, 69.1% (n = 103) of the participants did not have access to primary healthcare, and this rate was alarmingly high among people residing in the Apulian ghettos (85.9% n = 55 vs 56.5% n = 48, p < 0.001). Apart from country of origin (p < 0.001) and access to care, the migrant and homeless subpopulation differed significantly on the following variables: reporting chem-sex (p = 0.049) and HCV and HIV knowledge (p < 0.001 and p = 0.02)—with homeless people scoring better.

Knowledge, attitudes, and practices towards sexually transmitted infections

Concerning HIV/AIDS, only 60% (n = 90) of the recruited individuals were aware that the disease can be transmitted through sexual contact. Additionally, 45.7% (n = 65) were not aware of the protection offered by condoms. A notable 34.9% (n = 52) agreed or strongly agreed that HIV/AIDS might be transmitted by sharing utensils or having a meal together. Moreover, over 75% (n = 111) of the sample did not agree with the statement ‘thanks to HIV treatment, people can lead a high-quality life and not be contagious anymore.’ Awareness of Hepatitis C was notably low, with limited knowledge about its modes of transmission, potential complications, and available treatment options in both subpopulations. Details about the distribution of responses to the KAP questionnaire are provided in Supplementary Table 1. Both HIV (p < 0.001) and HCV (p = 0.02) knowledge were significantly lower among agricultural migrant workers (Table 1).

Regarding the perception of stigma, the majority (59.1%, n = 88) of participants expressed fear that a positive HIV test would lead to social exclusion, while only 25.5% (n = 38) held similar concerns about a positive HCV test. In terms of practices, the majority of participants (85.9%, n = 128) reported inconsistent condom use, while 14.1% (n = 12) of homeless individuals stated that they had engaged in chem-sex.

Sexually transmitted infections screening

The total number of infections detected is illustrated in Figure 1. Overall, out of 149 participants screened, 3 tested positive for HIV (2%), 14 for HCV (9.4%), 21 (14.1%) for HBV, and 5 (3.4%) for syphilis. Only 14% were vaccinated for HBV. HBV serology results (Figure 2) indicate that most of the study participants have neither been vaccinated against HBV nor have they ever been exposed to the virus. Approximately one out of four people (24.8%, n = 37) were positive for at least one infection, while three people out of four (74.5%, n = 111) reported that they had never undergone prior testing for these infections.

Mono and multivariable analysis for return rate for test results

Overall, only half of the screened participants (50.3%, n = 75) returned to collect test results. As shown in Table 1, at univariate analysis, people who had access to primary healthcare were significantly associated with returning (41.3% n = 31 vs 20.3% n = 15, p < 0.009), as well as people who performed better at HCV knowledge questionnaire (p = 0.04). Alarmingly, people whose results indicated infection with HCV or with at least one STI were more likely to be lost to follow-up.

At multivariable analysis, the only factor independently associated with reduced odds of returning to collect test results was being positive to HCV (Table 2).