Applying the exclusion criteria for this secondary analysis as described above resulted in a study sample of n = 686 participants. About 3.3% (n = 22) subjects had more than three missing values on the seven eligibility criteria and were subsequently excluded from the sample for the bivariate and multiple regression analyses (n = 664).
Graph 3 presents the weighted percentages of people with SSA migrant background in Antwerp who met the PrEP eligibility criteria (see the corresponding table A in appendix). In total, about 30.6% of them met at least one of the adapted criteria and thus were eligible to PrEP. If we exclude the three specific criteria (7.1, 7.2 and 7.3) only 17.7% of the sample were eligible for PrEP. The most frequent criterion was ‘having condomless sex with someone of SSA origin and being unaware of his/her HIV status’ (16.3%), while a minority met the criterion ‘diagnosed on a STI less than 6 months ago’ (1.7%).
Graph 3
Percentages of SSA migrants meeting the PrEP eligibility criteria
![](https://myfiles.space/user_files/58892_7798ecd9a40b82f9/58892_custom_files/img1600921582.png)
Among all HIV negative SSA migrants in our sample, 19,9% met only one criterion for PrEP eligibility, 7,4% met two criteria and 2,2% met three of them. The minority of SSA migrants meeting three or more criteria (2,3%; n = 23 out of n = 684) were a selective and vulnerable group: almost half had no health insurance, thus being probably undocumented, most of them lived less than two years in Belgium, were not in a relationship and experienced financial difficulties.
The bivariate results (see Table 1) show that those who were eligible to use PrEP were significantly more likely to be male, MSM, single (versus in a relation and cohabiting) and reported having no health insurance than those who were ineligible. They were also more likely ever having experienced forced sex.
Table 1
Factors associated with eligibility to PrEP in bivariate and multivariate analysis (weighted data)
| Total | % | Un- adjusted | 95%-CI | Adjusted | 95%-CI |
Variable (N total) | N | % | eligible | OR | | | | OR | | | |
Age | 664 | | | | | | | | | | |
Between 18–30 years old (ref.) | 239 | 36.05 | 30.33 | 1.00 | | | | 1.00 | | | |
Between 31–40 years old | 248 | 37.42 | 30.04 | 0.99 | 0.67 | 1.46 | | 1.33 | 0.83 | 2.15 | |
Older than 41 years old | 176 | 26.53 | 31.07 | 1.09 | 0.72 | 1.66 | | 1.70 | 0.99 | 2.93 | |
Gender | 664 | | | | | | | | | | |
Women (ref.) | 423 | 63.78 | 25.51 | 1.00 | | | | 1.00 | | | |
Men | 240 | 36.22 | 32.95 | 1.38 | 0.97 | 1.97 | * | 1.65 | 1.09 | 2.50 | * |
MSM | 664 | | | | | | | | | | |
No MSM (ref.) | 651 | 98.07 | 29.76 | 1.00 | | | | 1.00 | | | |
MSM | 13 | 1.93 | 61.54 | 3.62 | 1.16 | 11.25 | * | 2.58 | 0.74 | 8.97 | |
Relation status | 664 | | | | | | | | | | |
Not in a relationship (ref.) | 259 | 39.01 | 34.35 | 1.00 | | | | 1.00 | | | |
In a relation and cohabiting | 284 | 42.83 | 24.14 | 0.60 | 0.41 | 0.87 | ** | 0.56 | 0.35 | 0.88 | * |
In a relation and not cohabiting | 120 | 18.16 | 36.59 | 1.08 | 0.69 | 1.70 | | 1.05 | 0.62 | 1.76 | |
Country of Origin (region) | 664 | | | | | | | | | | |
Western Africa (ref.) | 444 | 66.97 | 28.89 | 1.00 | | | | 1.00 | | | |
Central Africa | 180 | 27.14 | 30.98 | 1.08 | 0.74 | 1.57 | | 1.55 | 1.00 | 2.40 | * |
Southern or Eastern Africa | 39 | 5.89 | 45.00 | 1.89 | 0.98 | 3.68 | a | 2.12 | 0.99 | 4.54 | a |
Migration duration | 656 | | | | | | | | | | |
Living in Belgium for more than 10 years (or born) (ref.) | 237 | 35.7 | 31.10 | 1.00 | | | | 1.00 | | | |
Not living in Belgium | 37 | 5.65 | 21.60 | 0.63 | 0.28 | 1.43 | | 0.61 | 0.23 | 1.62 | |
Living in Belgium since < 2 years | 149 | 22.52 | 32.90 | 1.08 | 0.70 | 1.68 | | 1.01 | 0.57 | 1.76 | |
Living in Belgium for 2–10 years | 232 | 34.95 | 29.30 | 0.92 | 0.62 | 1.36 | | 0.85 | 0.54 | 1.35 | |
Legal statusa | 663 | | | | | | | | | | |
Documented (ref.) | 518 | 78.12 | 28.00 | 1.00 | | | | 1.00 | | | |
Probably undocumented | 145 | 21.79 | 40.00 | 1.72 | 1.17 | 2.52 | ** | 1.83 | 1.12 | 2.99 | * |
Education | 645 | | | | | | | | | | |
Primary school or less (ref.) | 101 | 15.18 | 34.00 | 1.00 | | | | 1.00 | | | |
Completed secondary | 313 | 47.18 | 33.50 | 0.98 | 0.61 | 1.57 | | 0.99 | 0.58 | 1.67 | |
Continued education | 231 | 34.86 | 26.20 | 0.70 | 0.42 | 1.15 | | 0.56 | 0.31 | 1.00 | * |
Employment | 664 | | | | | | | | | | |
(Self)employed | 317 | 47.78 | 29.90 | | | | | | | | |
Unemployed/non-employed | 288 | 43.41 | 32.60 | 1.14 | 0.81 | 1.61 | | 0.81 | 0.52 | 1.24 | |
Full time student | 58 | 8.81 | 24.10 | 0.77 | 0.40 | 1.46 | | 0.80 | 0.37 | 1.70 | |
Financial problems | 625 | | | | | | | | | | |
No (ref.) | 228 | 34.30 | 26.80 | 1.00 | | | | 1.00 | | | |
Sometimes/most of the time | 397 | 59.86 | 33.50 | 1.39 | 0.97 | 1.99 | a | 1.23 | 0.82 | 1.84 | |
Housing | 636 | | | | | | | | | | |
Stable (ref.) | 595 | 89.69 | 30.60 | 1.00 | | | | 1.00 | | | |
Unstable | 41 | 6.13 | 40.00 | 1.54 | 0.80 | 2.94 | | 1.35 | 0.61 | 2.99 | |
Forced sex (lifetime) | 664 | | | | | | | | | | |
Never (ref.) | 620 | 93.44 | 29.40 | 1.00 | | | | 1.00 | | | |
Ever | 44 | 6.56 | 47.70 | 2.21 | 1.19 | 4.11 | ** | 2.10 | 1.01 | 4.39 | * |
*p < 0,05 **p < 0,01 ***p < 0,001; ref. = reference category; n = 664 for the logistic regression analyses |
(a) operationalized by the proxy ‘no health insurance’ |
For the multivariate logistic regression analyses, there were no large differences between the results of Model 1, 2 and 3 when adding the socio-economic and forced sex variables. Therefore, only Model 3 is presented in Table 1 (adjusted ORs): Men, those without a relationship, those without health insurance, and who ever had experienced forced sex were more likely to be eligible for PrEP than women, those in a relationship and co-habiting, with health insurance, and those who had not experienced forced sex. Being MSM was no longer significantly associated with ’being eligible to PrEP use’ when controlling for the other variables in the model. Educational level and region of origin became significantly associated to PrEP eligibility. Participants with only primary or no education and those from Central-Africa were more likely to meet one of the eligibility criteria when compared with those with vocational or university education (continued education) and people originating from Western Africa.
The results of the logistic regression analyses for the separate criteria did not change during the stepwise procedure. The third model is presented in Table B of the Appendix in addition to the unadjusted ORs.
SSA migrants without health insurance were more likely to have four or more sex partners and to report condomless sex. Without controlling for the other variables they were also more likely to report condomless transactional sex. Male SSA migrants reporting same sex behaviour (MSM) also had a higher likelihood of transactional condomless sex and of being in a concurrent relationship with condomless sex and low likelihood of condom use in the future, also after taking the other variables into account.
SSA migrants of the age-group between 31 to 40 years, from Central, Southern or Eastern Africa, or experiencing economic hardship, were more likely to take drugs and/or alcohol while having condomless sex compared to younger SSA migrants, SSA migrants from Western Africa and SSA migrants without economic hardship. Taking the other variables into account, SSA migrants older than 40 were more likely to have had a STI during the last six months, to have reported condomless sex during travelling and condomless sex with a partner of African origin without knowing his/her HIV status, but they were less likely to had four or more sexual partners the last 12 months compared to the youngest age group (below 30 years of age).
In a sensitivity analysis, the control variable study setting was included. Study setting is a categorical variable consisting of 4 categories: bar/party of African organization, church, public place (park, street, square), meeting of African organization, and other (shop, hair salon, library, asylum center). This variable had no significant effect on the eligibility to PrEP and taking this control variable info account did not change the other effects.