Introduction

After the first reported case in Namibia in 1986, HIV rapidly became a generalized epidemic in the country, with prevalence rising from about 4% in 1992 to about 19% in 2010 [1]. Accordingly, HIV has been estimated to account for over 50% [2] of all deaths in Namibia in 2002. More recently, this number has dropped since then to 23% [3]. This is most likely due to the high coverage of antiretroviral therapy with about 90% of people in need being treated [4], nevertheless, the disease is still one of the major causes of illness and death in the country. Although HIV prevalence rates in Namibia are known to differ between the capital and regional settings, substantial variation between single regions, from 4 to 36%, also occur [1]. Such differences are likely to influence or to be influenced by HIV related knowledge, attitude, and behavior. In order to design and implement a successful HIV prevention program, it is therefore essential to attain a proper understanding of regional differences in perception and behavioral patterns. Despite years of HIV campaigning in Namibia however, such data is limited.

The public sector is of special interest for HIV prevention as it is the largest employer in Namibia with approximately 85.000 employees [5]. Their scope of influence is even more substantial when one considers that public sectors prevention activities include utilizing social networks and largely affects the relatives of its employees as well. Furthermore, the public sector is a good cross-sectional representation of the entire country as it is composed of numerous different social groups and income categories, from the unskilled worker to the upper management.

There is no valid data on HIV prevalence in the public sector. However, a study of the Office of the Prime Minister (OPM) extrapolated available data from other sources to the public sector employees to assess the impact of HIV and AIDS. The study estimated a prevalence of about 15% [6]. It is therefore reasonable that one component of the “National Strategic Framework for HIV/AIDS response in Namibia” (NSF) is to establish HIV Workplace Programs (WPP) [3] in the public sector. The programs commonly focus on HIV prevention and the provision of, and referral to, HIV testing, care and support services.

The present study was conducted in the public sector in Namibia in the context of the implementation of WPPs. It aimed at providing information on HIV related knowledge, attitudes, and behavior of ministerial employees. The study’s objective was to identify gaps and needs with regard to HIV, and to provide baseline data for future monitoring and evaluation of workplace programs. In addition, as most HIV campaigning takes place in the capital, we aimed to compare the capital Windhoek and four rural regions (Hardap, Erongo, Oshana, and Caprivi).

Methods

This cross-sectional study was part of a larger survey including the public and private sector in Namibia. The present manuscript reports the public sector results only. The study was conducted between January and March 2011. The study assessed the current status of knowledge, attitudes and behavior (KAB) regarding HIV and AIDS among public employees. Results are compared between the capital (Windhoek) and four different rural regions (Hardap, Erongo, Oshana, and Caprivi).

Results

Discussion

Although Namibia is a predominantly rural country, HIV prevention campaigning is centered in the capital city, Windhoek. This is likely largely due to logistical reasons, distance and availability of HIV prevention workers. This is true for the general population as well as for employees of ministries. Regional imbalances in HIV prevention generated some concerns with regard to possible differences in HIV related knowledge, attitudes and behavior between the employees in the capital and outside the capital and also with respect to the effectiveness of the campaigns. This was one reason why in the present study employees from the capital were calculated separately and compared to the regions. We found that differences in knowledge, attitudes and behavior between the capital and the four rural regions were small with the exception of Caprivi, which proved to be an outlier.

Knowledge regarding principal HIV transmission routes was found to be high in our study population. This corresponds to the findings of the 2006 Namibian Demographic Health Survey (DHS) and a study by the Nawa Life Trust reporting high levels of knowledge regarding HIV in Namibia [10,12]. An exception was knowledge of mother-to-child transmission of HIV. As the probability of transmitting the virus from mother to child ranges between 15% and 45% [13–15] without preventive interventions, this suggest a high need for education, especially in high prevalence regions. Furthermore, this study revealed an uncertainty and misconceptions in parts of the workforce regarding transmission routes such as transmission of HIV by sharing food, supernatural means and mosquito bites.

Male circumcision is a medical intervention which significantly lowers the risk of HIV transmission, as several studies have demonstrated [16]. According to the 2006 DHS, in Namibia about 20% of all men were circumcised, but rates do vary between 1% and 57% between regions [12]. In our male study population one third indicated to have been circumcised. As could be expected, circumcision rates varied largely between the regions, with all rates exceeding national survey numbers. Among the uncircumcised more than half expressed willingness to get circumcised and only about a third rejected this completely. This is encouraging for campaigns promoting circumcision; especially as willingness to circumcise was the same in high or low HIV prevalence regions.

With a reported average HIV prevalence of about 19% in Namibia [1], appropriate personal risk assessment and protective behavior is essential to stop the on-going spread of the disease. In this regards, a considerable share of 48% self-declared HIV-negative study participants indicated to be at high or moderate risk “of getting HIV”. According to the aim of the study design, the different study regions varied significantly with regard to location, cultural context and, predominantly, HIV-prevalence; results generally reflect these preconditions. Hence, risk perception differed significantly between the regions and was in general positively associated with regional HIV prevalence; with the exception of Caprivi, where risk perception was surprisingly low taking into account that Caprivi has the highest HIV prevalence in the country [1]. This finding is consistent with other misconceptions being more prevalent in Caprivi and may be a possible explanation for the high prevalence rates of this region. Other factors, such as Caprivi’s remote location, its status as a transit zone between four countries (Angola, Zambia, Botswana, and Zimbabwe), and weak infrastructure may also play a role.

Several studies have demonstrated that general infection risk depends on multiple personal factors including age, gender, and marital status [17,18]. In this study, these three variables did not affect HIV risk self-perception. Risk self-assessment needs to be translated into behavior change in order to be a marker of HIV risk, and is expected to be influenced by knowledge on transmission routes and employees’ own exposure. Many previous studies however have demonstrated a weak association between knowledge and behavior [19–21]. It is to note our findings suggest an association between HIV related knowledge and risk behavior. One of the major indicators of perceived risk resulting in behavior change is condom use. Accordingly, this study found participants with low or no perceived risk for HIV infection to be less likely to use condoms. This finding was independent of living in a stable relationship (married or cohabitating) or not. This highlights the need for comprehensive information about HIV transmission and prevention, and is an indicator of the potential impact of information campaigns enabling a realistic and reasonable risk assessment. In this regard, it is particularly interesting to investigate what influence the actual sexual behavior has on the risk perception of participants. We found that reported concurrent sexual relationships in the last 12 months were associated with a perceived high or moderate risk of getting HIV, which suggests a more realistic risk assessment in this population group. Accordingly, average condom use at the most recent sexual encounter was relatively high. This behavior may result from informed decision making: engaging in concurrent sexual relationships changed behavior of some participants.

An important factor for successful prevention and treatment of HIV is early and reliable HIV testing. In the participating ministries universal testing has been promoted for several years. 98% of the participants of our study indicated knowing where to get tested and 87% declared to have been tested for HIV before. Although these numbers are high, more than 10% had never been tested with lower rates in men compared to women (82% vs. 91%). This result is similar to DHS rates (87% vs. 92%) [12]. A high percentage of participants would accept an offer by the ministry itself to receive an HIV test. This indicates a trust in institutional confidentiality, and individuals not afraid of possible negative consequences with regard to their job. Given this result, offering voluntary counseling and testing (VCT) in ministerial institutions seems to be an important strategy as part of the WPP.

To investigate behavior patterns and perceptions regarding HIV transmission, participants assessed their sexual behavior. Concurrent sexual relations are seen as particularly problematic for the spread of the disease, as newly infected persons are highly infectious because of rapid virus replication in the first weeks [17]. Thus, infected persons may pass on the virus rapidly before knowing their own status. In our study, 21% of participants indicated concurrent sexual relations in the last 12 months. These numbers correspond to the results of the DHS (2000: 21%, 2006: 16%) [11,12], with higher rates in young, male and single employees. Even though 77% of employees with multiple concurrent partners reported condom use (2006 DHS: 66-74%), these numbers are high, taking in consideration the high HIV rates in Namibia. The potentially dramatic consequences of such behavioral risks should therefore be continually addressed and highlighted in future public health campaigns.

About 50% of participants reported having used a condom during the most recent sexual encounter. This rate is substantially higher than in the last 2006 DHS with only about 22% indicating to use condoms in general as a contraceptive method [12]. However, the DHS represents a nationwide household survey, covering all regions and settings; thus, representing a very diverse population. Additionally, data collection in the DHS was done with face-to-face interviews, potentially provoking underreporting in sensitive issues. A further exclusively urban household survey reported even higher condom use at the most recent sexual encounter (55-78%) [10]. This lends validity to our data, as our study includes only employees living mostly in an urban setting.

Limitations of this study include the methodological limits of a cross-sectional design, that can demonstrate associations, but cannot determine causality as well as the standard limitations of questionnaire and interview based surveys, including e.g. selection bias, reporting bias, translation bias. We tried to minimize potential bias by choosing a random sample of participants, anonymization, training of fieldworkers, etc. The questionnaire was quite detailed due to multiple questions the participating organizations and stakeholders wanted answered. Therefore, a potential reporting bias with less time and thoroughness regarding the last questions may have occurred. In our post-study validation we found no evidence for such an effect, however, observations during the study have compelled us to recommend shortened questionnaire for follow-up studies. Another potential reporting bias may be caused by the inclusion of illiterate participants. We tried to minimize barriers to answer sensitive questions using indirect interviews. Post-study validation did not find differences between literate and illiterate participants which could not be explained by socio-demographic factors. Nonetheless, the inclusion of illiterate participants remains challenging and requires careful preparation and training. Data for this study was collected in an environment where the workplace program (WPP) has already been implemented for years. Generalization of the results to employees of other Namibian ministries with no or different WPPs is therefore limited. The results of this study, however, might inform workplace programs in similar settings of high HIV prevalence. In an era of up scaling antiretroviral treatment with the aim to prevent new HIV infections, further research is necessary how to integrate this approach into workplace programs. Our results demonstrate the need to further investigate regional and cultural particularities, especially in Caprivi, to overcome barriers for prevention and treatment. The aspect of concurrent partnerships also needs to be better understood to appropriately adjust HIV prevention measures, as it is considered as one of the key factors for high HIV prevalence rates. Considering the limited prospects of rapid success by promotion of behavioural change, best-practice for promotion and implementation of more technical prevention strategies (condoms and VCT) have to be evaluated. This is especially true for the most at risk groups; namely young and male. It is obvious that in a complex social system assessing the impact of a WPP is difficult, as participants are always exposed to multiple information sources (media, family, etc.). We addressed this issue by including an open question asking participants what their main source of information regarding HIV and AIDS was in the survey. Most indicted electronic and printed media as their main source of information. More evidence is needed however, to determine how to construct WPPs to complement other campaigns to avoid cost-ineffective redundancy.

Conclusions

We found good HIV-related knowledge among ministerial employees in Namibia. Differences between employees in the capital and in the regions were surprisingly low, with the exception of Caprivi. Several behavioral risk factors for HIV transmission persist in Namibian public employees such as multiple concurrent partnerships. They should be continuously targeted in national campaigns and educational programs as well as workplace programs with a special focus on remote high prevalence regions.

Acknowledgments

Results presented are part of a study hosted by the Ministry of Health and Social Services (MoHSS) of Namibia, and supported by the Office of the Prime Minister (OPM) and the Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ). It is part of the HIV Research and Evaluation Agenda and stated as a study to be conducted during the period of the NSF 2010/11 and 2015/16. The study was made possible due to the support of several people. Explicitly named should be Jennifer Loewen, Carmen Perez-Samaniego, Matthew Black, Stephan Klose, Ingrid de Beer, Martin Odiit, Abner Axel Xoagub, Uaeta Muzuma, Edward Shivute, Alexia Krug von Nidda, and Karin Shiimi.