Overview of included studies
Fig. 1 shows the procedure of study inclusion. We identified 4856 articles using the specified search criteria (PubMed: n=732; EMBASE: n=1612; Scopus: n=2277; PsycINFO: n=235). 940 duplicated papers were removed. Based on the inclusion criteria, 3867 papers were excluded. 49 papers remained for full text review, and 13 papers met inclusion criteria.
Table 2 provides details regarding the 13 studies included in the review. Eight took place in Africa (three in Zimbabwe[18, 30, 31], two in South African[19, 32], two in Ethiopia[33, 34] and one in Tanzania[35]), while the remaining five were performed in Asia (three in Thailand[17, 36, 37] and two in China[38, 39]). Two were RCTs[30, 36] and eleven were cohort studies[17-19, 31-35, 37-39]. All studies reported a statistically significant effect on one or more outcomes, which were reported as follows: eight reported HIV risk behaviors[17, 33-39], four reported HIV/AIDS knowledge[17, 36, 37, 39], four reported HCT uptake[18, 19, 31, 32], one reported HIV public stigma[19], one reported HIV/AIDS attitude[17] and one reported HIV incidence[30]. Risk behaviors included presence of multiple sex partners, commercial sex, recreational drugs or alcohol use before sex, and condom use, etc. According to the UNAIDS[40], the definition of HIV public stigma was a process of devaluation of people either living with or associated with HIV infection, such as, those who blamed foreigners/migrant workers/prostitutes for spreading HIV/AIDS.
The types and definitions of HIV intervention among factory workers in LMICs are summarized in Table 3. Further details on the risk of bias are reported in Additional file 1: Table S2. Of the cohort studies, five were assessed as strong quality[17, 33, 34, 38, 39], four as moderate quality[19, 31, 35, 37], and two as weak quality[18, 32]. Of the RCTs, two were assessed as weak quality[30, 36].
Efficacy of different intervention methods among factory workers
Educational intervention
Three studies focused on educational intervention[17, 36, 38]. Two studies indicated that educational intervention could improve condom use (condomless sex in the last 12 months decreased from 6.9% at baseline to 3.8% at month 12, P < 0.001; use of condom during sexual intercourse in the last 3 months increased from 41% at baseline to 70% at month 3, P < 0.05)[17, 38]. The other two studies showed that educational intervention could improve HIV/AIDS knowledge[17, 36]. For example, the proportion of workers who knew that antibiotics did not prevent HIV increased from 46.9% to 56.3% (P = 0.03) and that mother-to-child could spread HIV increased from 82.6% to 93.4% (P < 0.05). One study showed that educational intervention could reduce the proportion of workers with extra-partners (from 16% to 5%, c2=5.32, P = 0.021)[36]. One study showed that educational intervention could reduce the proportion of workers who used recreational drugs (from 2.6% to 0.7%, P < 0.01) or alcohol (from 17.0% to 6.3%, P < 0.01) before sex[17]. In addition, educational intervention changed HIV attitudes. For example, the proportion of workers who perceived that if they had HIV/AIDS they would not be able to live in society decreased from 46.6% to 30.6% (P < 0.05)[17].
Community intervention
One study conducted community intervention[31]. The study indicated that workers having high risk behaviors were more likely to take HCT, and proactive provision of HCT could increase the detection rate of HIV (relative risk [RR]: 1.87, 95% confidence interval [CI]: 1.01 to 3.61) and sexually transmitted diseases (STDs) (RR: 3.47, 95%CI: 2.51 to 4.89). Moreover, HIV seroconversion was higher among subjects who obtained their test results at the first follow-up visit compared to those who did not (19.5% vs. 16.7%, respectively, P = 0.01).
Combination of lottery intervention and community intervention
Two studies focused on lottery intervention combined with community intervention, both of which analyzed the changes in HCT uptake before and after the intervention[19, 32]. Moreover, these studies had demonstrated that lottery intervention could improve HCT uptake (from 30% to 85% (P < 0.001)[32] and from 27.3% to 53.6% (P < 0.001)[19]). In addition, lottery intervention could also reduce HIV public stigma. For example, the proportion of subjects who thought that foreigners/migrant workers/prostitutes were to blame for spreading HIV/AIDS decreased from 22.2% to 9.6% (P < 0.05)[19].
Combination of educational intervention and community intervention
Four studies conducted educational intervention combined with community intervention[33-35, 39]. Three studies demonstrated that educational intervention combined with community intervention reduced the proportion of workers with casual sex (from 12.0% to 6.1%, P = 0.03[33]; from 17.5% to 3.5%, P < 0.001[34]; from 8.8% to 4.6%, P < 0.01[35]). Two studies showed a decrease in the proportion of workers having sex with sex workers[33, 34], but only one report[34] had statistically significant result (from 11.2% to 0.75%, P < 0.001[34]). One study reported the proportion of workers with more than one sex partner and indicated that the proportion decreased from 17.6% to 10.2% (P < 0.05) for having two sex partners and from 4.7% to 2.0% (P < 0.05) for having three or more sex partners[35]. In addition, the combination of these two interventions increased condom use (from 7.6% to 27.3%, P = 0.002)[35], reduced premarital sex (10.9% in intervention group, 31.3% in control group, P < 0.001)[39], and improved HIV knowledge (P < 0.05)[39] and the awareness rate of location providing free health educational counselling (from 3.5% to 6.7%, P < 0.001)[39].
Combination of peer education and community intervention
Two studies focused on peer education combined with community intervention[18, 30]. One studies indicated that peer education reduced incident HIV infection rate (1.51 vs. 2.52 per 100 persons-years, P < 0.05)[30]. Another study concluded that peer education rendered more workers to take their partners to HCT (odds ratio [OR] = 1.37, 95% CI: 1.04-1.79), but statistical significance was not found for individuals to take up HCT (OR = 1.05, 95% CI: 0.92-1.20)[18]. In addition, workers with STDs (OR = 2.78, 95%CI: 2.25-3.43), commercial sex (OR = 1.27, 95%CI:1.09-1.49) and multiple sex partners (OR = 1.31, 95%CI: 1.14-1.50) in the last 6 months were more likely to take up HCT[18].
Combination of policy intervention and educational intervention
One study conducted policy intervention combined with educational intervention[37].
This study indicated that combination of these intervention increased HIV/AIDS knowledge (t = 2.84, P = 0.005), perceived condom accessibility (OR = 2.80, 95% CI: 1.13-6.86, P < 0.05), and condom use with regular partners (OR = 1.25, 95% CI: 1.01-1.54, P < 0.05) at the last sex.