5. Discussion
5.2 General risk and sexual behavior
5.2.3 Non-commercial sex partners and numbers of sexual partners
5.2.4 Commercial sex partners and non-regular sex partners
5.3 HIV testing and exposure to interventions
5.4 Relationships between knowledge, sexual behavior and perception of risk
Results are discussed in relation to other local literature; qualitative results were used to strengthen and clarify the quantitative findings. The qualitative data show clearly that the behavioral surveillance survey was successful in opening and developing a dialogue on HIV/AIDS amongst the study respondents.
5.1 Knowledge about HIV/AIDS
5.1.1 Awareness
Awareness of the existence of HIV/AIDS was high (>90%) amongst all target groups, irrespective of gender and region. This finding supported the results of the Ethiopia Demographic and Health Survey (CSA 2001), which showed high levels of awareness of HIV/AIDS amongst the general population (96 and 85% for males and females, respectively). Moreover, qualitative data showed that awareness of HIV/AIDS was high even in rural communities. In comparison with previous studies of rural communities in Ethiopia, this was an encouraging finding; for example, according to Ismail et al. (1995) awareness of HIV/AIDS was only 74% in a rural community in the Gondar region.
5.1.2 Preventive methods
A previous study in the major cities of Ethiopia indicated that about 94% of males and 84% of females (15-49 years old) had knowledge of the preventive methods (Mehret et al. 1996). In this study, about 67% of FSWs, over 45% of all youth and over 56% of other groups combined could mention the three major methods of preventing HIV infection. Amongst the FSWs, knowledge about preventive methods was highest in Dire Dawa and lowest in Gambella town. Knowledge of preventive methods was higher in the in-school youth (ISY) than in the out-of-school youth (OSY). However, only one of every two youths could name all three preventive methods. Uniformed services respondents had better knowledge of preventive methods than the other groups studied; over 73% of ground force and air force respondents were able to mention the three major methods correctly.
Compared with other target groups, the rural populations had less knowledge of the preventive methods. Less than 25% of the pastoralists and farmers knew the three preventive methods. The lowest level of preventive knowledge (1.8%) was found amongst the female farmers of Butajira. In general, findings amongst the rural population were of a similar magnitude to the 28% reported by CSA (2001). However, it is important to note that the CSA (2001) and Mehret (1996) assessed knowledge of preventive methods by determining the proportion of individuals who knew two or more of the three preventive methods. In contrast, the BSS assessed knowledge of preventive methods by assessing the proportion of respondents who could correctly identify all three preventive methods. Nevertheless, irrespective of the differences in methodology, knowledge about the preventive methods was very low amongst the rural populations, particularly amongst the female farmers. As the great majority of Ethiopia’s population lives in rural areas, the very low knowledge levels observed amongst the rural population groups studied needs serious consideration.
Most respondents knew that male condoms existed. In contrast, many were unaware that female condoms existed, presumably because female condoms were not available on the market and because female condoms were not actively promoted. Respondents who knew about male condoms also knew where to obtain them. For most of the respondents, it took less than 30 minutes by foot to obtain a condom. Data from focus group discussions (FGDs) and individual in-depth interviews (IDIs) supported the quantitative findings relating to access to condoms; FGD and IDI participants commented that condoms were available and could be obtained at a reasonable price. These results contrasted with the findings of an earlier study by Mehret et al. (1996), which reported that availability of condoms was low outside of Addis Ababa.
5.1.3 Misconceptions about HIV/AIDS
About 50% of OSY and a slightly over 50% of ISY were able to identify the misconceptions correctly. The OSY in the Harari, SNNPR and Somali regions had the least misconceptions about HIV/AIDS; in contrast, youth in Addis Ababa and the Afar region had the most misconceptions.
Misconceptions (such as ‘mosquito bites can transmit HIV infection’) were widespread amongst FSWs, particularly in Addis Ababa, and the Amhara and Oromia regions. Despite high levels of HIV/AIDS interventions in Addis Ababa, FSWs working in the city had many misconceptions about HIV/AIDS.
Pastoralists and farmers had more misconceptions than any other target group. Both quantitative and qualitative data confirmed this finding. Misconceptions amongst farmers included the beliefs that raw eggs laid by a chicken that had swallowed a used condom and raw meat could transmit HIV infection.
Different sets of misconceptions existed within each target group. The qualitative study revealed additional misconceptions; for example, that ‘the lubricant in condoms carries the virus.’
Despite high levels of awareness of the existence of HIV/AIDS and relatively good knowledge of the preventive methods, the presence of widespread misconceptions should be of concern to all individuals and organizations involved in the fight against HIV/AIDS. Some misconceptions related to stigma whilst others raised serious questions about the extent of people’s knowledge of HIV/AIDS, in particular the depth of the understanding they require to protect themselves.
5.1.4 Comprehensive knowledge
Measures of comprehensive knowledge give a better picture of the level of understanding about HIV/AIDS, especially regarding the modes of transmission. Knowledge of preventive methods was markedly greater than comprehensive knowledge about HIV/AIDS. With the exception of male ISY (15-19 years old), less than 40% of respondents in each target group were considered to have comprehensive knowledge about HIV/AIDS. The level of comprehensive knowledge was lower in females than in males. Female farmers had the lowest level of comprehensive knowledge. This important finding should help program planners to focus on decreasing the level of misconceptions and increasing knowledge on HIV/AIDS transmission mechanisms. As demonstrated by the qualitative results, presence of widespread misconceptions affects protective behavior and thereby, weakens efforts to curtail the spread of HIV infection.
5.1.5 Stigma and discrimination
Results showed high levels of stigma relating to HIV/AIDS. In fact, the majority of the respondents showed at least one stigmatizing attitude towards people living with HIV/AIDS (PLWHA). Even amongst the youth, around 97.3% of 15- to 19-year-old OSY and around 96.4% of 20- to 24-year-old OSY expressed stigmatizing attitudes towards PLWHA. Amongst OSY, females were slightly less likely than their male counterparts to have stigmatizing attitudes towards PLWHA. Although focus group discussions and in-depth interviews revealed that level of stigma had decreased amongst the youth, stigma against PLWHA was still a major problem.
Considerable discrimination against PLWHA was reported amongst FSWs, with as few as 10% of FSWs responding positively to all questions relating to stigma and discrimination.
Proportion of respondents with stigmatizing attitudes varied between the target groups, ranging from 99.4% in male pastoralists to 55.7% in truckers.
Quantitative and qualitative results indicated that although there were individuals with accepting attitudes towards PLWHA, there were also those who showed extreme stigma and discrimination towards PLWHA (e.g. female farmers). These findings support those of Kifle (2001) who found that most respondents (>18 years old) in Dire Dawa showed at least one form of stigma towards PLWHA. Stigma and discrimination stand out as significant challenges to the prevention and control of HIV/AIDS in Ethiopia. The presence of a high level of stigma facilitates the spread of HIV/AIDS, not least by its effects on some of the prevention packages such as VCT.
5.2 General risk and sexual behavior
5.2.1 Drug and alcohol use
Drug and alcohol use were examined because of their potential roles in predisposing individuals to the practice of unprotected sex.
There are few previous studies reporting levels of drug and alcohol use in Ethiopia. In 1997, Alem showed that over 50% of the adult population of Butajira (n = 10,468) had ever used khat and about 50% were khat chewers at the time of the survey; amongst users, 17.4% used khat every day. In the same study, Alem showed that 23% of the adult population used alcohol 23% (36 and 15% for males and females, respectively). A more recent study, in the Zeway area of Ethiopia, indicated that 32% of adolescents and adults (>15 years old) currently chewed khat (Belew et al. 2000).
In this study, relatively low levels of regular drug and/or alcohol use were observed amongst most target groups. Nevertheless, there were differences between the groups and subgroups. For example, regular drug use (mainly khat) and alcohol consumption were higher amongst the OSY than the ISY. Younger respondents were less likely than older ones to have consumed alcohol in the week before the survey or to have used drugs. Qualitative data showed that drug use was higher when youths were unemployed or had completed high school but not yet found a job.
Amongst most adult populations, very low levels of regular drug and alcohol use were reported, particularly amongst the uniformed services. These results need further investigation, to determine whether there was under-reporting of drug and alcohol use. In contrast, levels of regular drug and alcohol use were very high amongst FSWs. Over 70% of FSWs drank alcohol regularly. This could be explained by the fact that most of the FSWs were working in bars and in establishments where local alcoholic drinks were sold. As clients came to these establishments, FSWs were almost always invited to have a drink.
Results indicated that the combination of khat and alcohol was a major risk factor for unprotected sex. In this respect, the customary behavior of khat chewers (revealed by the qualitative study) was of particular interest as it was common practice for khat chewers to break the effects of khat chewing by drinking alcohol (a practice often called chebbssi). Nevertheless, overall there was very little alcohol drinking following khat use, particularly amongst the rural population.
5.2.2 Premarital sex
Amongst 15- to 19-year-old youth, almost twice as many OSY (35 and 29% of males and females, respectively) as ISY (19 and 13% of males and females, respectively) reported that they were sexually active. These results agreed with the findings of several previous studies amongst in- and out-of-school adolescents (Abate 1999; Berhane et al. 2000; Eshetu et al. 1997; Fantahun and Chala 1996). Amongst the older group of OSY (20-24 years) nearly 75% of males and 60% of females were sexually active.
Results of the BSS complemented the finding of a recent study by Taffa et al. (2002), which examined sero-prevalence amongst the youth in Addis Ababa; prevalence of HIV-1 was found to be 13-times higher amongst OSY than ISY (0.3 and 5.3%, respectively).
The proportion of sexually active individuals was higher amongst OSY in Gambella, Beneshangul-Gumuz and Tigray than amongst the youth in other regions. It was difficult to explain these differences between regions, In general, however, results indicated that level of sexual activity amongst the youth was increasing. Qualitative data indicated that reasons for this trend included the expansion of illegal video houses showing erotic films, increased use of drugs and unemployment.
5.2.3 Non-commercial sex partners and numbers of sexual partners
During the previous 12 months, greater proportions of OSY than ISY had had sexual partners. Amongst the 15- to 19-year-olds, 35.1 and 17% of OSY and ISY, respectively, reported having one or more sexual partners. Over 35% of 20- to 24-year-old OSY reported having one or more sexual partners. The proportion of youth with one or more sexual partner was higher than that observed in Dubti (Afar region) where 9% of OSY was found to have had more than one sexual partner during the previous 12 months (Assefa 2002). Moreover, a study in four regional cities (Addis Ababa, Awassa, Bahir Dar and Dire Dawa) by Mehret et al. (1996) showed that comparatively small proportions of 20- to 24-year olds (14.5 and 3.9% for males and females, respectively) had had more than one non-regular sexual partner in the previous 12 months.
Almost half of the sexually active OSY in Tigray, Beneshangul-Gumuz and Gambella had more than one partner in the last year. The percentage reporting multiple sexual partners in the last year was higher in OSY than in ISY. Amongst OSY, more males than females had more than one sexual partner in the last year.
Amongst the adult population groups (uniformed services, transport workers, rural populations and factory workers), almost a third reported having more than one sexual partner during the previous 12 months. Amongst all married adults, 33% reported having extramarital sex during the previous 12 months. Data highlight the fact that marriage in itself is not a guarantee of protection from HIV infection unless faithfulness is practiced consistently.
5.2.4 Commercial sex partners and non-regular sex partners
Commercial sex was most common amongst the uniformed services, particularly the ground forces. Amongst those who were sexually active, 76% of ground force and 21% of air force respondents reported having commercial sex partners in the previous 12 months. Moreover, about 20% of them reported having non-regular sex partners. This behavior is likely to be related to the high mobility of the ground forces and the long duration of time they spend away from their families. Amongst drivers and their assistants, around 13% had commercial sex partners and around 10% had non-regular sex partners. These results are comparable to those of Mehret et al. (1996) who studied urban populations of Ethiopia and found that 18% of males and 5% of females had non-regular sex partners in the previous year.
Only a small proportion (1%) of factory workers had commercial sex partners in the previous 12 months but about 8% had non-regular partners. The low level of commercial partners may have resulted because of interventions by the Ethio-Netherlands AIDS Research Project (ENARP) in the area.
Less than 1% of pastoralists and none of the farmers reported having had any commercial sex partners. Over 12% of pastoralists but only 0.2% of farmers (one farmer) reported having non-regular sex partners in the previous 12 months.
Qualitative data indicated that commercial sex partners were uncommon amongst the youth. However, the youth frequently had non-commercial partners and a significant proportion did not use condoms.
5.2.5 Condom use and unprotected sex
A substantial proportion of youth did not use condoms with their non-commercial partners. Consistent condom use during the previous 12 months was lower amongst the OSY than ISY and lower amongst female than male youth.
Amongst the adult population, unprotected sex was practiced despite high levels of knowledge about the preventive methods of HIV infection. Unprotected sex occurred most often amongst the uniformed services and minibus drivers. Alarmingly, about 27% of FSWs did not use condoms the last time they had sex with a non-paying partner. In contrast, only 3% of FSWs did not use condoms the last time they had sex with a paying client. The relatively low use of condoms with non-paying partners probably occurred because the FSWs trusted their non-paying partners; these men were often regular partners of the FSWs and most were considered to be the women’s lovers. Data from FGDs revealed that most street-based FSWs had lovers with whom they did not use condoms.
The qualitative data also indicated that FSWs working in peripheral areas of Ethiopia and those coming from rural areas were unable to negotiate for condom use. In a study of FSWs in small towns in northwest Ethiopia, it was reported that the rate of unprotected sex was 2- to 3-times higher in those who were illiterate than literate (Degu 2002), indicating that poorly educated FSWs were less able to negotiate for condom use. In the same study, it was noted that FSWs who used condoms less frequently, felt that they didn’t need to use condoms when having sex with farmers coming from remote rural areas.
5.2.6 History of STIs and treatment seeking behavior
In general, prevalence of STI symptoms (genital discharge or ulcers/sores) was low. Amongst the youth, STI symptoms were reported (in the previous 12 months) by only 4, 5 and 2% of the younger OSY, older OSY and ISY, respectively. Similarly, the prevalence was only 5.8% amongst the adult group respondents. Nevertheless, the level of STIs was higher than that reported previously in Ethiopia (Mehret et al. 1996; CSA 2001); for example, CSA (2001) reported a prevalence of 1.4%. In part, this difference may have occurred because the CSA survey included much larger groups of the rural population.
5.3 HIV testing and exposure to interventions
5.3.1 HIV testing
Very few of the study participants reported ever having taken an HIV test. With the exception of the uniformed services, amongst whom 20% had been tested, in other groups less than 11% had ever been tested for HIV. Qualitative data revealed that the majority of people had not taken HIV tests because of high costs, fear of stigma and the lack of VCT services.
Although direct comparison between the BSS and previous studies is difficult, the quantitative and qualitative results of this study indicated that levels of HIV testing were higher than the 1% (of the population) reported previously by the CSA (2000b). Despite this possible trend towards an increase in HIV testing, the practice of VCT was still very low. Nevertheless, the majority of respondents (over 76%) reported that they would be willing to undergo VCT in the future. In general, VCT services were exceedingly inadequate considering the scale of the HIV problem in Ethiopia.
5.3.2 Media interventions
In order of increasing importance, the major sources of information on HIV/AIDS were printed media, television and radio. These data agree with a recent report from Addis Ababa, which revealed that the most common sources of information about HIV/AIDS were radio and television followed by anti-AIDS clubs (Cherie 2002). With the exception of farmers and pastoralists, coverage of HIV media messages was high amongst the various population groups included in this study. The majority of respondents commented that messages regarding HIV/AIDS were clear.
Amongst FSWs, condom use was associated with exposure to one or more of the types of media in the previous 12 months. Knowledge about the preventive methods increased with exposure to media information sources. In adult groups (the uniformed services, factory workers, drivers and rural respondents), knowledge about preventive methods, and condom use with commercial partners were associated with exposure to the media. Qualitative results indicated consistently that radio and television were the major sources of information on HIV/AIDS; moreover, these information sources were often considered very useful. These findings were encouraging as they confirmed the important role of the media in increasing knowledge about the preventive methods of HIV and in influencing positive behavior, such as condom use.
5.4 Relationships between knowledge, sexual behavior and perception of risk
Amongst the youth, condom use was positively associated with knowledge about the three preventive methods. Most of the youth who were aware that condom use protects against HIV transmission had used a condom the last time they had sex in the previous 12 months. As shown by Zemenfes (1996), knowledge that condoms protect was more likely to be associated with the use of condoms than with other preventive behaviors such as abstinence or faithfulness. Most youth respondents perceived themselves to be at no or low risk of HIV infection; this was even true amongst the subgroup of youth who had had unprotected sex in the previous 12 months.
The majority (85%) of FSWs perceived themselves to be at no or low risk of HIV infection because they used condoms. Qualitative data indicated that none of the FSWs perceived themselves to be amongst the high-risk groups. However, the quantitative survey showed that nearly 67% of FSWs who practiced unprotected sex at their last commercial sex encounter and about 74% of those who used condoms inconsistently with paying clients during the previous 30 days perceived themselves to be at moderate risk for HIV infection because of their risky behavior.
The majority (80-95%) of other adult population groups perceived their risk of HIV infection to be nil or low. Relatively higher perception of risk was observed amongst some uniformed service personnel and minibus drivers (16.5 and 11%, respectively); these individuals perceived themselves to be at moderate or high risk of HIV infection.
In all target groups, the majority of respondents perceived themselves to be at low risk for HIV infection. Even amongst those who did not use condoms, particularly with non-commercial partners, individuals perceived themselves to be at low risk because they trusted their partners. Interestingly, qualitative data revealed that the FSWs perceived themselves to be at low risk for HIV infection because they used condoms consistently with commercial partners. Moreover, FGDs and IDIs, revealed multiple factors underlying high risk behavior including suspicion that condoms carried HIV, unemployment and loss of hope, and the combination of drug and alcohol use.
5.5 Limitations to the study
1. The segmentation method of selection of OSY was applied based on the number of households reported by the 1994 census (CSA 1998). Since the household numbering was not updated there were problems in identifying the selected households.
2. The sensitive nature of the questions may have affected the level of honesty with which respondents gave their answers.
3. The timing of the survey coincided with the completion of classes and final exams for 12th graders (ISY) and this made the mapping difficult and time consuming.
