Developing effective HIV education in Central Asia: a cross-sectional study among Kazakh medical students

Introduction: The HIV epidemic continues to expand in Central Asia, bridging from injecting drug users to the general population, largely through heterosexual transmission. This study describes medical students’ attitudes and behaviours regarding HIV and sexual practices in response to recently introduced HIV curriculum. Methodology: Medical students in Almaty, Kazakhstan, were invited to participate in a cross-sectional survey between April and November, 2013. The survey assessed participant’s risk behaviours (e.g., injecting drug use, condom use), concerns related to HIV infection risk (e.g., medical care exposures, personal behaviours), and ability to identify HIV-positive individuals without asking their status. Bivariate analyses (Chi-square test or Fisher’s exact test) and multivariate analysis (log-binomial regression) were conducted. Results: Of 647 medical students, the majority were female (82.2%) and single (92.8%) with a median age of 21-22 years. Sexual activity was reported by 176 (27.2%); males were substantially more likely to report sexual activity (74.8% males vs. 16.9% females, p <.0001). The majority of sexually active medical students (65.6% and 68.3% males and females, respectively) believe they can determine if their partner is HIV-positive without asking or seeking a laboratory test. This proportion is much higher (92.6%) among married or cohabitating women. No association between perceived ability to ascertain HIV status and condom use was identified. Conclusions: The medical students in our study had a relatively naïve perception of HIV risk, and their knowledge of risk did not translate to appropriate precautionary behaviours. Much work remains if the epidemic in Central Asia is to be contained.


Introduction
The HIV epidemic continues to expand in eastern Europe and central Asia. Having gained a foothold in the region through transmission among people who inject drugs (PWID) the epidemic has now bridged to the general population, largely through heterosexual sexual transmission. Kazakhstan is a central Asian country of 17 million citizens bordered by two countries with substantial HIV epidemics. Within the region, the estimated adult prevalence of HIV infection is approximately 1.0% or higher [1]. In Kazakhstan, the estimated prevalence of 0.3% is believed to be an underestimate [2], although the prevalence is thought to be lower than in neighbouring countries. Thus, there is growing concern among Kazakh public health professionals that the incidence and prevalence of HIV may dramatically increase unless substantial efforts are exerted.
Clearly, prevention programs targeting young people are critical to slow the HIV epidemic. Yet according to a 2011 review, education is limited as Kazakh schools do not comply with UNESCO recommendations [3,4]. Limited studies of students have been conducted in Kazakhstan. A knowledge, attitudes and behaviours survey of students in the city of Semey found about 75% of males and 25% of females reported sexual activity, and consistent condom use was infrequent [5]. About half the students identified health care professionals as resources they would seek out for counsel.
The purpose of this study was to understand the risk behaviours and perspectives of medical students in Almaty, the largest city in Kazakhstan. Medical students will become the educators, counsellors and treating physicians in the near future. A new HIV education program was developed at the Asfendiyarov Kazakh National Medical University (KazNMU) in 2011. Students are now required to take a one-day course on HIV clinical care during their fourth academic year (first clinical year of training). While the university does administer exams that assess students' basic HIV knowledge, these exams do not cover more subjective information, such as their perspectives on transmission. Such perceptions can impact personal risk behaviours and interactions with patients, including counselling and education.
This study assesses the attitudes and behaviours of medical students. Our primary interest was to determine the current status of students' thinking about their personal HIV exposure risk with the goal of improving HIV education. To measure these perceptions, we focused on understanding which factors were associated with the students' perception of their partner's HIV risk.

Methodology
This cross-sectional study was conducted at the Almaty AIDS Centre (Almaty, Kazakhstan) where training sessions on infection control and HIV prevention are provided for medical students. Fourthand fifth-year medical students were approached to enrol in the study while they were attending courses during the study period (April -November 2013). A researcher, neither on faculty nor involved in medical education, introduced the study to students in the classroom setting. The students were asked to complete a survey.
The study was approved by the KazNMU Research Ethics Committee. To protect students, the survey was anonymous. No names or other identifiers were included on the survey. A written informed consent was provided to potential participants; however, no signed informed consent form was required. A waiver for signed informed consent was granted because the consent itself would be the only identifier linking results to the participant. To minimize peer pressure for students related to participation, students were told that if they refused to participate they could leave the survey blank or write anywhere on the form "do not use" and it would be discarded.
The survey included questions about the participant's risk behaviours, including their injection drug use history and their use of condoms with regular and other partners. Students were also asked about their concerns related to HIV infection risk due to exposures from providing medical care or personal risk behaviours. Additionally, they were asked about their ability to identify HIV-positive individuals without asking their status. Finally, information on demographic and educational characteristics were collected, including age, gender, marital status, living situation and socioeconomic status.
Bivariate analyses were conducted to provide a basic description of the student participants. Frequencies and percentages were computed for categorical variables and means/medians with standard deviations were computed for continuous variables.
Bivariate analyses were conducted using Chi-square tests, or for sparse data, Fisher's exact test.

Results
A total of 785 medical students agreed to participate in the study, of whom 647 (82.4%) provided sufficient information to be included in these analyses. (Table 1) Similar to the general student body, the majority of students were females (82.2%) who were 21-22 years old. Most students were single (92.8%), but females were twice as likely as males to be married (5.9% compared to 3.3%, respectively). Female students were also more likely to live in dormitories (29.2%) compared to male students (12.2%); conversely, male students were more likely to live on their own or with friends (51.3%) compared to female students (29.2%) (p < 0.001). Aside from these gender differences, female and male participants were equally likely to be from urban vs. rural areas and had a similar, more affluent socio-economic background compared to the average Kazakh family, reflecting the relatively small middle class in Kazakh society. Sexual activity was reported by 176 (27.2%) students with males much more likely to report sexual activity compared to females (74.8% vs. 16.9%, p < 0.0001). Female students living in the dormitory and from rural areas were among the least likely to report being sexually active. While not statistically significant, increasing socioeconomic status was associated with increased reports of sexual activity. The majority of sexually active medical students believe they can determine if their partner is HIVpositive without asking the question directly or seeking a laboratory test. For males and females who are neither married nor cohabitating, 65.6% and 68.3%, respectively, believe they usually can spot HIV-positive individuals. (Table 2) For married or cohabitating women, 92.6% believe they know their partner's HIV status without asking. No association between perceived ability to ascertain HIV status and condom use was identified. In general, among non-cohabitating couples, about half of males and a quarter of females reported using condoms all the time with no significant difference based on perceived ability to ascertain the HIV status of partners. Multivariate regression analysis did not identify any significant predictors of a student's reported ability to assess his/her partner's HIV status without asking.

Discussion
This study found that a critical piece of HIV education is lacking among medical students in Kazakhstan, and likely other regional countries as well. The majority of medical students surveyed perceived that they can determine if a partner is HIV positive without inquiry. This perspective identified an important lack of understanding of the typically asymptomatic presentation of HIV. This misunderstanding impacts personal protective behaviours, and potentially how the new doctor delivers medical care in the future. This naïve perspective may arise from a misunderstanding about symptomatic AIDS presentation (e.g., cutaneous lesions due to Kaposi's sarcoma) and HIV infection. Additionally, because the HIV epidemic historically occurred primarily among PWID and other high risk groups, some medical students may think HIV-positive individuals should necessarily have physical signs of risk behaviours.
The surveyed students had one day of HIV/AIDS training in medical school and typically no formal education prior. The previous curriculum focused on microbiology and immunology. In 2011, HIV education was introduced into the medical curriculum. During years 1 and 2 the students take virology and immunology classes at the Department of Microbiology and Virology. In year 6, Ob-Gyn interns learn about the prevention of mother-to-child transmission and Dentistry interns study HIV prevention specific to dentistry.
The misperceptions may have been precipitated by a novel change to the HIV curriculum: introducing persons with HIV into the classroom to discuss the disease. The medical university is in the process of moving from didactic lectures to an active learning environment. Programs such as these can provide an opportunity for faculty to see how active teaching occurs and, when done well, strengthens the understanding of core concepts. However, due to the lingering stigma of HIV, only PWID and their partners working for non-governmental organizations located at the Almaty AIDS Centre volunteer to participate in these educational endeavours; professionals and students prize confidentiality and do not volunteer. Active learning, such as interviewing patients, can have a profound impact on knowledge acquisition. In light of our results, we realize that effective HIV education requires more than one day of training, and more than one infected demographic. This is particularly true given that no systematic HIV education exists in secondary schools. Students need to better understand risk behaviours that lead to HIV and other diseases transmitted sexually or through nosocomial prevention breaks.
Experiential education may be particularly useful in teaching students about HIV [6]. Students need to understand that HIV-positive individuals may be young and appear healthy. Today's clinician needs to have a full understanding that there is no "typical" HIVpositive patient as the epidemic continues to bridge to the general population. The primary mode of transmission of HIV infection is now heterosexual sex [7]. Students reported infrequent condom use, consistent with student reports in other countries [8][9][10]. Condom use was not associated with concern about risk of HIV infection. Substantial research on key facilitators of behaviours has suggested that fear is not sufficient for behaviour change. Rather, fear of infection needs to be taught in conjunction with culturally appropriate educational messages and access to condoms that are perceived to be effective [11,12].
As with all studies, some caveats exist. We asked about knowledge of risk factors for HIV status in the context of a personal relationship rather than the more didactic fact list often used by teaching institutions. As such, we may have elicited an artificially high (but realistic) proportion of students who believe they can predict HIV status based on personal experience. While this is less than precise, the student's opinion is informed by personal knowledge of important environmental and experiential characteristics of the individual which might not otherwise be available to a clinician without such a relationship.
The sample was from one medical university. However, the findings are consistent with results from medical students surveyed in another Kazakh city and a national survey of young people [13], suggesting that the findings may be generalizable to other university students. We suggest that medical students may have minimally better knowledge and behaviours than students in other disciplines and thus a slight bias may exist. We suspect that women reported sexual activity based on the still relatively conservative norms, particularly for students from rural areas. However, the focus of the study was largely among those students who reported sexual activity. While academic metrics ascertain general HIV knowledge, we focused on attitudes and behaviours, and the survey results reported here suggest knowledge did not translate well to either. Finally, although we did not fully assess HIV knowledge on the survey to make correlations at the individual student level, studies which have assessed both directly also find a lack of correlation between knowledge and behaviour [14,15].

Conclusions
This preliminary study takes a somewhat novel research approach in that we assessed medical student perceptions of their partner's HIV status to develop a real-world perspective on risk behaviours among the most educated Kazakhs. As such, this study provides a window on the best case scenario for the state of HIV knowledge and awareness in order to inform public health interventions focusing on the next generation of clinicians. The medical students in our study had a relatively naïve perception of HIV risk, and their knowledge of risk did not translate to appropriate precautionary behaviours. Much work remains if the epidemic in Central Asia is to be contained.