This cross-sectional survey was conducted at a private higher education institution in Johannesburg, South Africa. Participants were invited via an online learning platform where they completed and submitted a structured questionnaire assessing sexual risk and sexual prevention behaviours.

Johannesburg is the capital of Gauteng province and is considered the largest and the wealthiest city in South Africa. With a population estimated at 5.6 million, it is the most populous city in the country, with 66% growth rate in population expected over the next 30 years. Racial profiles indicate that 76.4% of the population is black African, 5.6% is mixed race, 12.3% is white or of European descent and 4.9% is of Indian or Asian descent. Approximately, 7% of the population is illiterate, 3.4% have only a primary education, 41% have completed secondary education and 6% have a tertiary qualification.¹⁸

Random sampling was used to recruit 845 students enrolled at a private higher education institution in Johannesburg, South Africa. Random numbers were generated using a computer program to select participants from the sampling frame – enrolment records. A global email was sent to all potential participants inviting them to participate in the study. Of those invited to participate in the study, 542 responded – a response rate of 64%. Participants completed an online questionnaire via their online learning platform, which contained study information and instructions for accessing and completing the questionnaire.

An online structured questionnaire measuring knowledge, attitudes and behaviours as well as demographics (including age, gender and relationship status) was completed by the participants. The questionnaire was developed as part of a larger study using existing literature and consisted of 93 questions of which 51 questions were on knowledge about HIV and AIDS transmission and prevention, attitudes towards HIV and AIDS, including treatment and prevention methods (six questions), and sexual behaviours (36 questions).

Data were cleaned and checked for errors before coding and analysing using STATA 14.0 (StataCorp, College Station, TX, USA). To evaluate knowledge and behaviours, respondents were required to provide mostly ‘yes’ or ‘no’ responses. For the knowledge score, a score of 1 was assigned for a correct answer and 0 for a wrong answer. For the attitude questions, a rank was assigned using the Relative Importance Index to obtain an overall rank for each attitude item. For knowledge, attitudes and practice questions (knowledge and awareness of HIV & AIDS, prevention and control of HIV, students’ attitudes towards condom use and people living with HIV, risky sexual behaviours), the frequency of responses in each category was determined. A chi-square test was used to evaluate the variation in knowledge, attitude and behaviour between male and female students. For all tests, p < 0.05 was considered statistically significant.

Ethical approval to conduct the study was obtained from Monash University Human Research Ethics Committee (MUHREC) (approval number CF15/1095 – 2015000518). No identifying information was obtained from students when they completed the questionnaire. A unique identifier was generated when the questionnaire was submitted. All the data were de-identified for analysis. Data were stored in password-protected files and will be retained for up to 5 years after the study.

Data were collected from 542 students, 374 (69.0%) of whom were female students. The participants had a median age of 19 years (interquartile range [IQR] = 16–30 years), and their average knowledge score of HIV and AIDS and sexually transmitted infections (STIs) was 0.78 (standard deviation [SD] = 0.17); 397 (73.2%) students were black Africans, 67 (12.4%) students were whites, 44 (8.1%) students were of Indian/Asian descent and 29 (5.4%) students were of mixed race; 427 (80.6%) students were single and 88 (16.6%) students were in a stable relationship. Most of the students were in the Higher Certificate, Higher Education Studies’ stream (n = 357, 71.1%), and 145 (28.9%) students were undergraduates (Table 1).

There were high levels of awareness and knowledge about biomedical methods of HIV prevention amongst the sample group. More female (77.1%) than male students (64.1%, p = 0.003) said that they had heard about medication that HIV-positive pregnant women could take to reduce the risk of infecting the baby with HIV, and more female (59.9%) than male students (47.6%, p = 0.015) had heard about medication that could help to reduce the risk of HIV infection if a woman had been raped. A higher proportion of male (95.8%) versus female students (85.7%, p = 0.001) said that they were able to obtain a condom. There was no significant difference between the proportion of male (97.0%) and female students (94.1%) about where to get condoms (Table 2).

Knowledge about HIV transmission was high, with 98.1% of female students and 96.4% of male students knowing that the virus could be passed on through unprotected sex. A high proportion of male students (92.9%) and female students (90.9%) knew that a person can have HIV and pass it on to others without showing symptoms. Also, most female students (77.4%) and male students (76.8%) knew that STIs put people at greater risk of HIV infection. However, only 33.6% of female students and 39.3% of male students admitted that they knew that anal sex increased the risk of HIV infection. A low proportion of female students (16.9%) and male students (21.4%) knew that people can reduce their chance of getting HIV by using a condom every time they have sex. More female (96.2%) than male students (91.7%) were aware that AIDS could not be cured. Most of the female students (n = 307, 82.5%) and male students (n = 119, 70.8%) indicated that partners could not have sexual intercourse if both partners were HIV-positive (p = 0.002) (Table 3).

As illustrated in Table 4, a substantial number of students expressed positive attitudes towards condom use. Across age and gender groups, a significant majority of students disagreed with the statement that a woman loses a man’s respect if she asks him to use a condom (96.8% of female participants and 91.7% of male participants); that they only use condoms if their sexual partner wants to use them (95.7% female participants and 86.3% of male participants); and that condoms should only be used if having sex with a person who is not the main sexual partner (94.6% of female participants and 83.9% of male participants). About a third of the students and 41.7% of the male students said that condoms felt unnatural, and nearly a quarter of the students aged 20–32 years and 36.3% of the male students said that condoms alter climax or orgasm.

Most participants had positive attitudes towards HIV-positive people, but 38 (7%) participants were unwilling to be associated with or share living space with people living with HIV. Based on the relative importance index score, the most important attitudes towards people living with HIV, ranked in order of relative importance, were: (1) About 83.4% of students indicated that even if a family member had HIV, their relationship with them would remain good; (2) 20% of students said that sharing a house with HIV-positive people would be very difficult for them; and (3) 7.6% of students felt that people who get infected with HIV are promiscuous. Forty-one (7.6%) students also said they do not want to be associated with HIV-positive people, and 6.4% of students felt that HIV-negative people should not be allowed to socialise with HIV-positive people. Students had positive attitudes towards treatment for HIV and AIDS. Around two-thirds (63.1%) of the students agreed that HIV treatment would keep an HIV-positive person alive (ranked the most important attitude). Two hundred and seventy-four (50.8%) students agreed that HIV medication really works (with a relative importance score of 2). Most of the respondents (69% students) rejected the notion that antiretroviral (ARV) medication is poisonous (Table 5).

Condom use at last sex was higher when with a regular partner: female students (n = 147) and male students (n = 93). The difference is statistically significant (p < 0.001). Fewer female students (n = 53) and male students (n = 57) reported using condoms consistently with a non-regular partner (p = 0.049). More female students (n = 83) reported consistent condom usage (every time) with regular sex partners than male students (n = 54); however, the difference was not significant (p = 0.240) (Table 6).

There were a disproportionate number of female students, compared to male students in this study. It is possible that this might have affected the robustness of the chi-square test and therefore skewed the findings. In addition, the sample came from one private higher education institution in Johannesburg, South Africa. It would have been ideal for more institutions to be included in the study to get a better picture of the knowledge levels, attitudes and behaviours towards HIV and AIDS at private higher education institutions in the Johannesburg metropolitan area. Notwithstanding these limitations, these findings could serve as a springboard for national research that would be more representative of the wider student population in both private and public higher education institutions.

It is recommended that future research should sample students from several private and public higher education institutions with a representative number of both male and female students. This could then inform interventions that reduce HIV infection among the youth in South Africa through improved attitudes facilitated by communication at both social and parental levels.²⁴ It is further recommended that sex education in secondary schools should be introduced to close the gender gap in knowledge and prevent risky sexual behaviours.