University of California, San Francisco Logo

University of California, San Francisco | About UCSF | Search UCSF | UCSF Medical Center

Perceptions of community HIV prevalence, own HIV infection, and condom use among teachers in KwaZulu-Natal, South Africa
Global Health Sciences Literature Digest
Published June 18, 2007
Journal Article

Chao LW, Gow J, Akintola O, Pauly M. Perceptions of community HIV prevalence, own HIV infection, and condom use among teachers in KwaZulu-Natal, South Africa.. AIDS Behav 2007 May;11(3):453-62.


To examine teachers’ subjective beliefs about the likelihood that their pupils, other teachers, and community members have HIV; the likelihood that they themselves have been or will become infected with HIV; the determinants of such beliefs; and the relationship between protective behaviors and such beliefs before and after they participated in one of two educational interventions.

Study Design

Cross-sectional study with pre- and post-intervention assessment.


Pietermaritzburg (PMB) region, KwaZulu-Natal (KZN), South Africa.


A total of 120 teachers were recruited in a non-randomized manner as part of a larger study by the KZN Department of Education, which was designed to determine the effectiveness of CD-ROM and Care and Support Workshop interventions. The teachers represented a total of 17 primary, secondary, and combined primary and secondary schools in the PMB region. All but eight of the recruited teachers participated in the interventions and were included in the analysis. Most teachers were female (77%), almost one-third had finished university (31%), and almost half resided in rural areas (48%). The only demographic difference between the two groups was that more teachers in the CD-ROM group lived in rural areas (chi-square =9.08, p<0.01).


Teachers either participated in a two-hour CD-ROM intervention or a two-day Care and Support Workshop. The CD-Rom intervention, entitled "My Year with Tony," was adapted from a CD-ROM intervention developed by the American Association of Colleges for Teachers Education and the Centers for Disease Control and Prevention to educate teachers on how to deal with various situations related to HIV/AIDS in the classroom setting. The Care and Support Workshop was conducted at a central location in the PMB region and included information on transmission and risk factors, as well as the actions that teachers should know and undertake when confronted with the many daily challenges that result from HIV infection.

Primary Outcomes

Teachers completed nearly identical pre- and post-intervention questionnaires that assessed their demographic characteristics; subjective beliefs about the HIV prevalence among their pupils, other teachers, and other community members; perceived likelihood of their own HIV infection; and knowledge, attitudes, and self-efficacy regarding reaching pupils about sex and HIV and their own condom use.


Data analysis consisted of descriptive statistics, bivariate comparisons and regression analysis. At baseline, the average teacher estimate of HIV prevalence was 36% among pupils, 48% among teachers, and 61% among the community’s general population. Male teachers gave lower HIV prevalence estimates for the community than female teachers (chi-square=5.93, p<0.05; odds ratio [OR] -1.63, p<0.01). Teachers with university degrees and teachers in rural areas gave lower estimates for pupils (OR -1.37 for both, p<0.05). Over all, 8% of teachers thought they were very likely to have HIV and another 24% thought they had about a 50% chance of being infected, while 23% felt that it was very unlikely, and 45% thought it was impossible for them to be infected. Teachers who were single gave higher risk assessments to themselves than teachers who were married/widowed/divorced (chi-square=4.11 and OR -0.43 for current infection, and chi-square=4.46 and OR 3.87 for eventual infection, p<0.05). Teachers in the CD-ROM group had higher risk perceptions than those in the Workshop group at baseline (chi-square=6.49, p<0.05). Teachers under the age of 40 and single teachers were more likely to use condoms (OR 2.6 and 3.0, respectively, p<0.05), as were teachers with perceived self-efficacy in the classroom (OR 1.20, p<0.05).

Of the 112 teachers who participated in one of the two interventions, 42 increased and 29 decreased their HIV prevalence estimates for students following participation. With regard to their perceptions about their own infection status, 75 teachers did not change, nine decreased, and 28 increased their perception following the intervention. While many teachers maintained their pre-intervention levels of risk perceptions, more teachers increased their risk perceptions than those who decreased such perceptions in the Workshop group only (p<0.01-0.05). No statistically significant differences between teachers in the two intervention groups were seen concerning post-intervention perceived risk for becoming infected with HIV in the future and HIV prevalence estimates for pupils, other teachers, and other members of the community at large.


The authors concluded that the teachers in this study made high overestimates of HIV prevalence in other people; teachers who made higher HIV prevalence estimates in other people also were more likely to believe that they themselves had already been or would eventually become infected with HIV; and teachers who made higher HIV prevalence estimates in other people also were more likely to have consistent condom use.

Quality Rating

This study was of poor quality. The authors did not thoroughly discuss how teachers were selected for this study, nor were teachers randomized to intervention groups. It is unclear to what extent participation in the intervention affected teachers' subjective beliefs. The small sample size may have limited the statistical significance of the findings and many reportedly significant findings reflected p<0.10 (not included in this digest). The study also did not include biomarkers to validate the teachers’ perceptions of HIV infection or other sexual-practice variables. Additionally, because the teachers in the Care and Support Workshop had lower perceived risk of being already infected with HIV prior to the intervention, the post-intervention differences between the two groups could have been confounded by other variables that differed between the teachers in the two intervention groups, such as locality of residence.

In Context

The teachers’ estimates of HIV prevalence in this study were far higher than the actual prevalence found in recent population based surveys with biomarkers for HIV in South Africa, which were less than 6% for children under 19 years of age for the whole country (but higher for KZN),(1) 21.8% among KZN teachers,(2)and 16.5% among KZN's general population.(1) This tendency to overestimate risk in others is consistent with studies elsewhere.(3)

Programmatic Implications

This study showed that a sample of South African teachers made considerable overestimates of HIV prevalence, and their condom-use behavior was positively associated with overestimated HIV prevalence. If the perceptions and behavior of this population reflect those of the general population, providing accurate information about HIV risks through public health educational campaigns may reduce protective behavior. This potential relationship is a concern, and should be evaluated in prospective, randomized studies involving large sample sizes of representative populations and biomarkers of HIV prevalence.


  1. Shisana O, Stoker D, Simbayi LC, Orkin M, Bezuidenhout F, Jooste SE, et al. South African national household survey of HIV/AIDS prevalence, behavioural risks and mass media impact--detailed methodology and response rate results. S Afr Med J 2004 Apr;94(4):283-8.
  2. Shisana O, Peltzer K, Zungu-Dirwayi N, Louw JS. The health of our teachers: A focus on HIV/AIDS in South African public schools, 2004/5 survey. Cape Town, South Africa: Human Sciences Research Council. 2005. (No abstract available.)
  3. Kalichman SC, Cain D. Perceptions of local HIV/AIDS prevalence and risks for HIV/AIDS and other sexually transmitted infections: preliminary study of intuitive epidemiology. Ann Behav Med 2005 Apr;29(2):100-5.