This article has Open Peer Review reports available.
HIV related risk behaviours among taxi drivers and their assistants in Addis Ababa, Ethiopia: descriptive cross-sectional survey
© Lakew and Tamene; licensee BioMed Central Ltd. 2014
Received: 14 February 2013
Accepted: 21 March 2014
Published: 8 April 2014
Risk taking behaviours in relation to HIV among the mobile population is a growing public health concern in many developing countries, including Ethiopia. The aim of this study was to describe risky sexual behaviours and associated factors among male taxi drivers and assistants in Addis Ababa.
A descriptive cross-sectional survey design with multistage cluster sampling procedure was employed to select 615 individuals for interview.
Seventy six percent of the respondents were sexually active. Nearly 31% of the respondents reported casual sex and 7% of them did not use a condom with their most recent casual sex partner. More than half (58.5%) of the respondents had no condom use efficacy. Condom breakage and/or slippage during sex had been encountered by 44% of respondents with casual partners and sex during menstruation had ever occurred among 17% of respondents. Eleven percent had experienced sex with female sex workers. Thirty-three percent of the respondents were unfaithful to their spouse/steady partners. Multivariate analysis revealed that living with parents [AOR 95% CI; 2(1.14-3.60)], non-khat chewers [AOR 95% CI; 3.7(2.13-6.31)], never taken VCT [AOR 95% CI; 3.5(1.84-6.72)], middle-class monthly cash gain [AOR 95% CI; 0.5(0.25-0.98)] and more years of experience working on a taxi [AOR 95% CI; 0.17(0.60-0.47)] were statistically significant to influence lifetime abstinence. Non-khat chewers [AOR 95% CI; 0.53(0.37-0.78)], never taken VCT [AOR 95% CI; 0.54(0.36-0.88)] and higher monthly cash gain [AOR 95% CI; 2.9(1.14-7.19)] had a statistically significant association with condom use efficacy. Living with parents [AOR 95% CI; 2(1.31-3.72)], living with friends [AOR 95% CI; 6.4(3.13-12.89)] and non-khat chewers [AOR 95% CI; 2(1.34-3.53)] were risk factors found to be associated with faithfulness.
Risky sexual behaviours in this sub-population were considerable and associated factors were found to be multidimensional. Therefore, there is a need for robust intervention strategies such as tailored serial radio program targeting taxi drivers and their assistants.
HIV/AIDS emerged in the 1980s as one of the major epidemics in the history of mankind . Sub-Sahara African (SSA) countries remain most heavily affected by the epidemic. In 2011, an estimated 23.5 million [22.1–24.8 million] people were living with HIV in the region, which accounted for 69% of the global AIDS burden . Ethiopia is one of the SSA countries most affected by the epidemic. Although reports indicate that the HIV epidemic is leveling off in Ethiopia, a significant number of people live with the virus and many families are affected [3, 4]. With a population of nearly 74 million in 2007, and about 1.4 million cases, Ethiopia carries a heavy burden of disease. It is estimated that HIV is responsible for about a third of all adult deaths in the age group 15–49 years and leaves nearly three-quarters of a million orphans in the age group 0–17 years and over 30,000 newborns with the virus per year .
The HIV epidemic has penetrated almost all population groups, including the hard-to-reach rural areas. However, with the existing socio-cultural diversity of Ethiopia, the pattern and distribution of HIV in the country widely varies. Some regions are more affected than others. The epidemic is more prevalent in urban than rural communities [3, 4]. Currently, the epidemic tends to concentrate in certain population groups who have risky sexual behaviours. Female sex workers (FSWs), long distance drivers, and daily laborers have been identified with high sexual risk behaviors for the epidemics of HIV infection in Ethiopia [6, 7]. This suggests that there are key populations at risk of HIV exposure that may play a key role in the epidemics.
Research in the 1990s indicated that about 40% of taxi drivers were found to be at risk of HIV in Gondar, Northwest Ethiopia . However, mini-bus-taxi drivers were not specifically counted in Ethiopia’s sentinel surveillance system [7, 8]. Recent data about HIV risk behaviours and risk factors among taxi drivers is limited. A study done in 2005 by Steve Tarevella revealed that taxi drivers are vulnerable to HIV infection because they travel throughout the city during their work, encounter different people, receive money, and sometimes find themselves pursued by women who hope to have money spent on them by the drivers . Use of khat, alcohol, and other stimulants are also common practices that motivate drivers to have unsafe sex .
In this regard, understanding risky sexual behaviours and the factors that drive or influence the epidemic among taxi drivers and their assistants is an important issue. Therefore, this study is aimed to describe risky sexual behaviorus and associated factors that result in HIV infection among male taxi drivers and their assistants in Addis Ababa.
A descriptive cross-sectional survey was conducted during the period of April 8–17, 2006.
For the determination of sample size the following attributes were considered: (1) The anticipated proportion of transport workers (inter-city bus drivers) that did not use a condom (~24%) during their most recent sexual encounters with casual partners ; (2) 95% confidence level (3) 5% worst acceptable result (precision) and (4) design effect of two. The sample size, including a 12% adjustment for the non-response was estimated at about 660 individuals.
A multistage cluster sampling technique was employed to select the required study subjects. In the first stage, all available clusters (taxi areas or stations) within Addis Ababa limits were identified based on the usual taxi density. Each cluster contains more than two taxi stations or areas. A total of 12 major taxi stations/areas were selected as a sampling unit. In the second stage, eligible study subjects were randomly interviewed from 12 selected clusters (taxi stations/areas). Taxi drivers and their assistants at the age of 15 and above and who worked at least six months duration on taxi services were eligible for an interview.
Data collection instrument
The data were collected using a structured questionnaire. The questionnaire was prepared in English and then translated to Amharic, which is a local language. The Amharic version of the questionnaire was pretested and the necessary modifications were made before the actual survey. Data collectors were second and third year students at Addis Ababa University with prior experience in data collection. Training was given to the data collectors on how to conduct the interview. Interviews were often completed during off-peak times when the taxi drivers and their assistants were less busy (i.e. mostly after 9:00 am in the morning and before 5:00 pm in the afternoon).
Data management and analysis
The data were entered into a computer system using SPSS version 15.0 and analyzed using STATA version 11. Appropriate statistical methods ranging from simple descriptive statistics to multivariate logistic regression analysis were employed to describe risky sexual behaviours and associated socio-demographic and psychosocial factors. Bivariate analysis was used to identify the relative importance of each predictor to the dependent variables by controlling the effects of other variables. Those variables which were significant on bi-variate analysis at (P-value < 0.05) were entered to multivariable logistic regression analysis to examine the effect of each independent variable to the dependent variables of the study .
If respondents engaged in any form of sexual activities, it was coded as 1, otherwise lifetime abstinence was coded as 0. Faithfulness, defined as respondents that never had sex with anyone other than their wives or steady partners was coded as 1 and otherwise 0. Condom use efficacy was a binary composite score outcome response that derived ability from questions of condom use during sex with casual sex partners, and were the dependent variables. Potential socio-economic, demographic and psychosocial factors including education (cannot read and write, grades 1–4, 5–8, 9–12, 12+), type of taxi drivers (mini-bus-taxi drivers, assistants, Lada taxi drivers), alcohol use (no, yes), age (15–19, 20–24, 25+), khat chewing habit (no, yes), usual taxi working areas (central areas of the city, peripheral areas), taxi ownership status (own taxi, employed), living arrangement (alone, with parents/relatives, with girlfriend/peers), monthly cash gain from taxi driving in birr (210–435,436-630,631-880,881-3640), and years of experience driving a taxi (<1 year, 1-5 years, 5+ years) were considered as independent variables. Most of the dependent and independent variables were adopted from previous behavioural surveillance surveys [6, 7] conducted in Ethiopia and research conducted elsewhere [8, 11]. Lada taxi drivers are those who give transportation services both during day and night times with 4 people at a time whereas mini-bust taxi drivers are mostly working during day time with 12 people at a time.
Ethical clearance was obtained from Addis Ababa University, Institute of Development Research and Population Studies Research Centre. Participation in the interview was entirely based on informed consent. No personal identifiers of the respondents were recorded.
Background characteristics of taxi drivers and assistants in Addis Ababa, 2006
Categories of taxi drivers
Small taxi (‘Lada’) drivers
Can’t read and write
Place of birth
In Addis Ababa
Out of Addis Ababa
Experience on taxi
5 + years
Risk behaviours among the respondents
Level of risky sexual behaviours among taxi drivers and assistants in Addis Ababa, 2006
(n = 615)
Percent (95% CI)
Abstinence in lifetime
Faithfulness with spouse/steady partner
Casual sex in the last one year
Condom use with casual sex at last sex
Ever had sex in lifetime with casual partners while menstruating
Ever encountered condom breakage or slippage in lifetime with casual partners
Had sex with FSWs in the last one year
Factors associated with HIV related risky sexual behaviours
Factors associated with experienced in lifetime abstinence as HIV risk preventive behaviour in taxi drivers and assistants, 2006
Crude OR 95% CI
Adj OR 95% CI
Categories of taxi drivers
Monthly cash gain
Experience on taxi
5 + years
Factors associated with condom use efficacy as HIV risk preventive behavior in taxi drivers and assistants, 2006
Crude OR 95% CI
Adj OR 95% CI
Monthly cash gain
Factors associated with faithfulness as HIV risk preventive behaviour in taxi drivers and assistants, 2006
Crude OR 95% CI
Adj OR 95% CI
This study explored the risky sexual behaviours and associated factors among taxi drivers and their assistants. A number of risky sexual behaviours including casual sex, condom slippage, sex during menstruation, number of partners and age at first sex were identified in this sub-population group. This study’s findings were evidenced by a study conducted in four urban SSA that revealed the risk of HIV infection increased in casual sex and number of lifetime sexual partners . A significant proportion of this study’s participants had sex with female sex workers (11.2%), which is higher than (7.9%) a study done among rickshaw pullers in Bangladesh . This difference could be due to regional contexts as well as methodological and time variations between the present and previous studies. A higher proportion of the study participants had sex with casual partners while menstruating. This finding is consistent with a study done in the United States that associated with sexual intercourse during menstruation, regular and frequent sexual intercourse, and large number of sex partners as components of the transmission dynamics of STD including HIV . Condom breakage or slippage in lifetime sexual practices of taxi drivers and assistants was also remarkably high. Compared with other target groups, casual sex practice in this study group was a two-fold increase from 15% in a cohort study among factory workers in Ethiopia . However, with the same target population, the report of casual sex practices (33.3%) is similar to other developing countries . Early sexual debut is another important individual-level risk factor identified in Zimbabwe for HIV infection among women . Similarly, this study identified a large proportion of taxi drivers and assistants that did not delay the first sexual debut in their lifetime. The median age at first sex, in this study group, was similar with the study in Tanzania  that identified as age of first sex as one of the risk behaviours for HIV infection among youngsters. Furthermore, there was a wide variety of lifetime sexual partners reported from taxi drivers and assistants, which significantly increases their risk of contracting HIV. This finding was a two-fold increase from the results of lifetime sexual partners (1–40) reported in a former study conducted among 480 naval (military) personnel in Lagos, Nigeria  in which the difference could be explained by variations in time, target population and methodological approaches between the former and present studies.
Despite having such risky sexual behaviours, this sub-population was not included in the national behavioural surveillance study in Ethiopia  or by other researchers. Even if various governmental and non-governmental organizations have made efforts to achieve behaviuour change through mass-media exposure, community sensitization and outreach programs in promoting safe sexual practices to the general population, the adoption of preventive behaviours among male taxi drivers is relatively low. Many are still engaging in unprotected sexual activities. Institutions working for the prevention and control of HIV/AIDS had limited information about taxi drivers and their assistants and consequently failed to consider this high-risk group as an important sub-population for intervention in the country.
Socio-demographic factors independently associated with HIV risk behaviours were types of taxi drivers, living arrangements, monthly cash gain and taxi driving experience. This study recognized that monthly cash gain has a two-way association with sexual risky behaviours: 1) it has a negative relation with lifetime abstinence and 2) a positive association with condom use efficacy. Alongside with these socio-demographic factors, a number of psychosocial and behavioural related factors were also found to predict risky sexual behaviours. For this group of population, less exposure to VCT service had an association with unfaithfulness to their sexual partners and reduced the ability to use condom during sex with casual partners. This indicates that VCT service provision has been missed as an important strategy for HIV prevention for this population group in the country. Whereas, less exposure to VCT services had association with lifetime abstinence that possibly due to most people have been initiated to have VCT during the time of marriage.
Substance use is generally believed to be one of the associated factors for sexual risk behaviours in HIV transmission. Khat is a locally produced psycho-stimulant commonly used in Ethiopia. Khat is widely consumed among youth in the city. A study assessing the magnitude of sexually transmitted infection (STI) together with self-reports of sexual risk behaviour among youths (15–24 years old) in Addis Ababa, Ethiopia, reported that increased sexual activity was significantly associated with khat consumption . Similarly, khat chewing habit among taxi drivers was also associated with unfaithfulness to sexual partners and early sexual initiations. However, khat chewers in this population group showed a higher condom use efficacy than non-chewers which might be due to prior knowledge and skills accumulated among the chewers.
The study population is extremely mobile which makes it difficult to properly classify clusters of taxi stations. Taxi working routes anywhere in the city and stations might have also potentially led to affect the homogeneity of clusters in this sup-population group. Furthermore, the study was a cross-sectional survey in its design which makes it impossible to determine causal relationships between the predictors and risk sexual behaviours. The self-reported measurements that were used might also have inherent biases and the potential for both underreporting and/or over-reporting in risky sexual behaviours. Despite these limitations, however, the study is important to give clue for further researches in similar population.
This study documented higher risk sexual behaviours of taxi working populations. The findings showed that monthly cash gain, living arrangements, categories of taxi drivers, taxi driving experiences and khat chewing habits were likely to significantly influence the practice of adopting HIV/AIDS preventive behaviours. Targeted interventions tailored to taxi drivers as mobile population are recommended to reduce substance use habits and to protect their sexual risk behaviours.
The authors would like to thank the Population Studies Research Centre and the School of Graduate Studies of Addis Ababa University for providing financial assistance. We also acknowledge Dr. Yared Mekonnen for his supervision and technical assistance in the overall research process. Our special thanks also go to the respondents and data collectors. Last but not least, we are grateful to thank Prof Susan Davies-Jones and Abigail Greenleaf for their professional language editing the document.
- The World Bank: Disease and mortality in Sub-Saharan Africa. The International Bank for Reconstruction and Development. 2006, 2: 8-17. Second editionGoogle Scholar
- The Joint United Nation Program on HIV/AIDS: Global Report. Fact Sheet on Sub Saharan African countries. 2009Google Scholar
- Ethiopian Central Statistical Agency (CSA) and ICF International: Ethiopian Demographic and Health Survey. 2005, Addis Ababa, Ethiopia and Calverton, Maryland, USA: Measure DHSGoogle Scholar
- Ethiopian Central Statistical Agency (CSA) and ICF International: Ethiopian Demographic and Health Survey. 2011, Addis Ababa, Ethiopia and Calverton, Maryland, USA: Measure DHSGoogle Scholar
- MOH: AIDS in Ethiopia. Disease Prevention and Control Department. Addis Ababa. Ethiopia. 2002, 491Google Scholar
- Rahlenbeck SI, Yohannes G, Molla K, Reifen R, Assefa A: Infection with HIV, syphilis and hepatitis B in Ethiopia: a survey in blood donors. International Journal of STD & AIDS. 1997Google Scholar
- Ministry of Health: HIV/AIDS Behavioural Surveillance Survey (BSS). Round one in Ethiopia, Addis Ababa. 2002, 1:6–20Google Scholar
- Ministry of Health: HIV/AIDS Behavioural Surveillance Survey (BSS). Round two in Ethiopia, Addis Ababa. 2005, 1:6–10Google Scholar
- Steve T: Driving Away HIV: Helping Taxi Drivers protect themselves in Addis Ababa, Ethiopia. Family Health International. 2005, 1-6.Google Scholar
- Landanu S, Everitt S: A handbook statistical analysis using SPSS. Multiple regressions. 2004, London: Chapman and hall/crc press LLCGoogle Scholar
- Auvert B, Buvé A, Ferry B, Caraël M, Morison L, Lagarde E: Ecological and individual level analysis of risk factors for HIV infection in four urban populations in sub-Saharan Africa with different levels of HIV infection. AIDS. 2001, 15 (Suppl 4): S15-S30.View ArticlePubMedGoogle Scholar
- Hafiz M, Ehsanul H, Masako O, Saman Z, Shahrzad M, Masahiro K: HIV-related risk behaviours and the correlates among rickshaw pullers of Kamrangirchar, Dhaka, Bangladesh: a cross-sectional study using probability sampling. BMC Public Health. 2009, 9: 80-10.1186/1471-2458-9-80.View ArticleGoogle Scholar
- Aral SO: Sexual Behavior as a Risk Factor for Sexually Transmitted Disease: Reproductive Tract Infections. 1992, New York: Plenum Press, 185-198.Google Scholar
- Mekonnen Y, Sanders E, Messele T, Wolday D, Dorigo-Zestma W, Schaap A, Mekonnen W, Meless H, Mihret W, Fontanet A, Coutinho RA, Dukers NHTM, Mekonnen Y, Sanders E, Messele T, Wolday D, Dorigo-Zestma W, Schaap A, Mekonnen W, Meless H, Mihret W, Fontanet A, Coutinho RA, Dukers NHTM: Prevalence and Incidence of, and Risk Factors for HIV-1 Infection among Factory Workers in Ethiopia. J Health Popul Nutr. 2005, 23 (4): 358-368.PubMedGoogle Scholar
- Orisatoki RO, Oguntibeju OO: HIV-related knowledge and condom use by taxi drivers in Southern St. Lucia, West Indies. Sci Res Essays. 2009, 5 (3): 304-308.Google Scholar
- Pettifor AE, van der Straten A, Dunbar MS: Early age at first sex: a risk factor for HIV infection among women in Zimbabwe. AIDS. 2004, 18: 1435-1442. 10.1097/01.aids.0000131338.61042.b8.View ArticlePubMedGoogle Scholar
- Hargreaves JR, Emma SL, Elizabeth F, Laura DH: Changes overtime in sexual behaviour among young people with different levels of educational attainment in Tanzania. J Int AIDS Soc. 2012, 15 (Suppl 1): 17363-PubMed CentralGoogle Scholar
- Nwokoji UA, Ajuwon AJ: Knowledge of AIDS and HIV risk-related sexual behavior among Nigerian naval personnel. BMC Public Health. 2004, 4: 24-10.1186/1471-2458-4-24.View ArticlePubMedPubMed CentralGoogle Scholar
- Mitike G, Mariam BH, Tsui A: Patterns of knowledge and condom use among population groups: results from the 2005 Ethiopian behavioral surveillance surveys on HIV. Ethiop J Health Dev. 2011, 25 (1): 35-45.View ArticleGoogle Scholar
- Alem A, Kebede D, Kullgren G: The prevalence and socio-demographic correlates of khat chewing in Butajira, Ethiopia. Acta Psychiatrica Scand (Suppl). 1999, 100: 84-91.View ArticleGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2458/14/330/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.