Patterns of condom use: Perspectives of men in KwaZulu-Natal, South Africa

In South Africa, the level of HIV infection is reaching crisis proportions. Besides sexual abstinence and mutually monogamous sexual relations, use of the condom has an important role to play in curbing the further spread of the HIV/AIDS. The aim of this article is to explore the extent and determinants of condom use inside and outside the primary sexual relationships. The research objectives are addressed using both quantitative and qualitative methodologies. The study found that knowledge about condoms, as a method of dual protection against the risk of pregnancy and disease, is high. Despite this, condoms are not a popular method of protection, especially within the primary sexual relationship. Consistent with numerous other studies, this study found widespread disapproval of condom use. Condoms are more likely to be used in non-primary than primary sexual relationships. In both relationships, condom use corresponds closely with level of education.

1. Introduction

At the end of December 2003, 37,8 million people throughout the world were living with HIV (UNAIDS, 2004). Globally, there were 2,9 million deaths from AIDS during 2003 alone, and 4,8 million adults and children became infected with HIV. Since the beginning of the epidemic, 21,8 million people are estimated to have died: 17,5 million adults (roughly 9 million women) and 4,3 million children. Almost 70 per cent of the HIV-positive people in the world live in sub-Saharan Africa. The primary route of HIV transmission in sub-Saharan Africa is through heterosexual intercourse. This is in direct contrast to other parts of the world, where a considerable proportion of HIV infection is transmitted through other routes, including intravenous drug use as well as homosexual intercourse (UNAIDS, 2004). South Africa is one of the countries hardest hit by the AIDS epidemic. The results of antenatal surveys conducted on blood samples from women attending antenatal clinics show that HIV prevalence has risen sharply, from less than 1 per cent in 1990 to almost 27 per cent in 2001 (Department of Health, 2003). Apart from sexual abstinence and mutually monogamous relationships between uninfected partners, the use of condoms is currently the most effective strategy for preventing the transmission of HIV and other sexually transmitted infections (Cates, 2001). Laboratory and epidemiological studies have shown the latex condom to be a highly effective barrier against HIV and other sexually transmitted infections, if used correctly and consistently (Cates, 2001). When the condom is used correctly, it is also a very effective means of preventing pregnancy (Trussell, 1998). The male condom is also the only widely used method that offers protection against the dual risks of unwanted pregnancy and disease (WHO, 2000).

In the mid-1980s, when AIDS first appeared as a major public health problem, relatively small proportions of sexually active individuals in most countries were using condoms (Mehryar, 1995). Recent surveys suggest that condom use is increasing more rapidly in non-regular than in regular sexual relationships. In the gold-mining areas of South Africa, where men work and live for months away from home, about two of every three men used condoms during sex outside marriage (Meekers, 2000). In the 1994 Zimbabwe Demographic and Health Survey (DHS), married men were found to be five times more likely to report use of condoms during the preceding four weeks with a non-spouse (60 per cent) than with a spouse (12 per cent). Married women reported using condoms 38 per cent of the time with non-spouses and 7 per cent with husbands (Zimbabwe Central Statistical Office, 1994). Condom use is also much higher among the unmarried than the married (Uganda DHS, 1996). According to the 1995 Uganda Demographic and Health Survey, sexually active unmarried men and women were currently using condoms much more (over 12 times more) than married respondents, perhaps to protect themselves from sexually transmitted infections (Uganda DHS, 1996).

The use of condoms in long-term, on-going relationships may have different implications for prevention from their use in short-term relationships. Several studies have found that pregnancy prevention was the main reason for condom use in long-term relationships, while prevention of sexually transmitted infections, including HIV, was the main reason for condom use in short-term relationships (Anderson et al., 1999; Adetunji & Meekers, 2001). Using data from the 1996 National Household Survey on Drug Abuse in the United States, Anderson et al. (1999) found that condom use is highly influenced by the length and intensity of relationship. The longer the relationship lasts, the greater is the likelihood that condom use will be discontinued. People in established relationships may prefer methods that are more effective at preventing pregnancy. A number of studies suggest that familiarity with a partner is accompanied by a decreased perception of risk, especially among men (Temin et al., 1999). This may, to some extent, explain the failure to use condoms in long-term, on-going relationships.

A review of the literature suggests that a number of factors may influence condom use. To date, however, limited data are available concerning the factors influencing condom use in both primary and non-primary sexual relationships. This study draws on data from focus group discussions, a household survey and in-depth interviews conducted in KwaZulu-Natal to determine the extent and determinants of condom use among men. The focus of the study is on men, because they are more likely than women to report having sexual relationships outside the primary sexual relationship (Adetunji, 2000; Adetunji & Meekers, 2001). Identifying the determinants of condom use is crucial for developing appropriate and effective prevention programmes.

2. The study setting

Located on the country's eastern seaboard, KwaZulu-Natal is one of the most populous of South Africa's nine provinces, with a population of almost nine million people. KwaZulu-Natal has one of the highest rates of HIV infection in South Africa. According to the 2002 antenatal HIV survey, 37 per cent of women attending state antenatal clinics in the province were HIV-positive (Department of Health, 2003). Several socio-economic and political factors make KwaZulu-Natal particularly vulnerable to HIV infection. These include rapid urbanisation, a highly mobile work force, an international port and high levels of poverty and unemployment (Whiteside & Sunter, 2000).

The study was conducted in an urban and a rural site in KwaZulu-Natal. Zulu-speaking people of low socio-economic status predominate in both sites. The rural site is situated approximately 80 kilometres south-west of Durban. Many parts of the district are not readily accessible by public transport and a large proportion of the households do not have access to electricity or piped water (Mbhele, 1998). The urban site is situated approximately 15 km from the city of Durban and is similar to many other urban African townships in South Africa. Homes, roads and schools are in need of repair; electricity and water supplies are constantly interrupted (LeClerc-Madlala, 1997).

3. Methodology

A combination of qualitative and quantitative methods was used. The qualitative data were derived from focus group discussions and in-depth interviews. There were 12 focus group discussions: six in the rural area and six in the urban area. The aim of the focus group discussions was to obtain attitudes to condoms and other related issues. The choice of this method was linked to its ability to generate a greater depth of understanding. Moreover, the results of the focus group discussions were used in designing the survey questionnaire. Also, 40 in-depth interviews were conducted. The aim of the in-depth interview was to gain more insights into the needs and perceptions of users and non-users of condom users. More specifically, the purpose of the in-depth interviews was to increase clarity and the depth of response. The in-depth interviews were also used to assess the validity of data collected in the focus group discussions and the survey. Both the focus group discussions and in-depth interviews were conducted in the local language.

The quantitative data were derived from a household survey. A household questionnaire was used to identify all eligible men and women in the household. In order to be eligible for the interview men and women had to be aged 20–49 years and had to be a resident of the household. Those who reported that they had ever had sexual relationships were considered eligible to complete the full questionnaire. Areas of research focus include condom use and acceptability, attitudes to risk-taking and the perception of risk, and how these differ as between the risk of pregnancy and the risk of HIV infection. Respondents were also asked to provide information about all their sexual partners outside the primary sexual relationship in the last three years. Detailed information was obtained only for sexual relationships that occurred within the last three years, and was restricted to a maximum of three partners. For our purpose, analysis is restricted to interviews with 511 men. Specially trained field staff of the same sex as each respondent conducted all the interviews. Of the total sample, 29 per cent were married and 19 per cent cohabiting. As might be expected, older men were more likely to be married or living with their partner. The mean age of this sub-sample was 35. Information was also obtained on 496 partnerships outside the marital and cohabiting relationship. Not surprisingly, men who were married or cohabiting were less likely to report having multiple sexual partners. Almost 6 per cent of men who were married or cohabiting reported having two or more sexual partners, compared with 56 per cent of men who were neither married nor cohabiting.

4. Data analysis

The quantitative data obtained from the survey were analysed using the Statistical Program for the Social Sciences (SPSS). For the purpose of this paper, the primary sexual relationship refers to the marital or cohabiting union and the non-primary sexual relationship refers to all sexual relationships outside the marital or cohabiting union. A partnership file was constructed to capture all sexual relationship outside the primary sexual relationship. The chi-square test was used to determine the statistical association between condom use and the explanatory variables. The relationships between these factors and condom use were also explored using a standard multivariate logistic regression model.

The qualitative data obtained from the focus group discussions and in-depth interviews were tape-recorded and detailed notes were taken. The tapes were transcribed and translated into English by the moderator and note-taker. All the transcripts were read and reread and organised according to particular themes and assigned initial codes. The computer package ethnograph was used to assist in analysis of the qualitative data. The transcripts are used to illustrate particular findings from the focus group discussions and in-depth interviews.

5. Results

5.1 Knowledge of and attitudes to condoms

Awareness of condoms is almost universal. When prompted and unprompted responses were combined, nearly 95 per cent of all respondents had heard of condoms. Respondents who reported knowledge of condoms were asked whether they knew where they could be obtained. The overwhelming majority of respondents who had heard of condoms (92 per cent) knew where they could be obtained. A person who has heard of condoms may nevertheless not know much about them, including how to use them correctly (Gardner, Blackburn & Upadhyay, 1999). This distinction is more clearly demonstrated by the following response made by one male participant in the focus group discussions: ‘I have never used it [condom] before because I do not know how to use it’ (rural male, FGD).

The belief that condoms can prevent HIV and pregnancy is an essential first step for their acceptance and subsequent use (Renne, 1993). Condoms are widely recognised as a method of dual protection against pregnancy and HIV infection. Almost eight of ten respondents agree that condoms protect against the risk of pregnancy and HIV infection. In general, condoms are seen as slightly more effective at preventing HIV infection than pregnancy. Nearly 89 per cent of respondents agreed that condoms are effective at preventing HIV and 82 per cent of respondents agreed that condoms are effective at preventing pregnancy.

Numerous studies have found that beliefs, both positive and negative, concerning condoms are important determinants of their use (Stiffman, Dore & Cunningham, 1994; Kapiga et al., 1995). Many respondents feel that it is acceptable for a married couple to use a condom. However, an equally large proportion feels it is not. Attitudes to condom use depend on the nature of the relationship. It is considered more acceptable for an unmarried woman to ask her partner to use a condom than it is for a married woman to ask her husband, as shown in Table 1. When asked if they considered it acceptable for a married woman to ask her husband to use a condom, 43 per cent of men found such behaviour acceptable. By contrast, 85 per cent of men said it was acceptable for an unmarried woman to ask her partner to use a condom.

Table 1: Percentage of male respondents who agree with specific statements about condoms

Statement	Per cent
Condoms are effective at preventing HIV	89
Condoms are effective at preventing pregnancy	82
It is acceptable for a married couple to use a condom	43
It is acceptable for a married woman to ask her husband to use a condom	47
It is acceptable for a woman who is not married to ask her partner to use a condom	85
Number of respondents	511

Attitudes to condoms were also explored in more detail in the focus group discussions and in-depth interviews. In the focus group discussions and in-depth interviews, men expressed a variety of fears about condoms. A common complaint is that condoms interrupt sexual activity, cause discomfort and ruin the excitement of flesh-to-flesh contact. Some men also expressed scepticism about the effectiveness of condoms as a strategy for preventing HIV transmission because, they argued, of their tendency to tear or break during sexual intercourse and they worried about the potential side effects and health risks for women.

I used it once but I was so embarrassed. I did not enjoy what I was doing. Your mind tells you that you are wearing a plastic. (Urban male, FGD)

Condoms may tear or burst inside a vagina and may lead to death. (rural male, FGD)

Condoms are more commonly associated with casual, rather than on-going, long-term relationships. In long-term relationships, many people do not consider themselves at risk because they trust their partners. However, men feel that it is important to always use condoms in short-term relationships, because of the fear of sexually transmitted infections (including HIV/AIDS).

I do not use a condom with my regular partner…I trust her, so there is no need to use a condom with her (Rural male, IDI)

The introduction of condoms in a long-term, on-going relationship may undermine the partner's trust and sometimes lead to accusations of infidelity. The perception that condoms encourage promiscuity is likely to contribute to their image as unsuitable for long-term, on-going sexual relationships. It became clear that women who request condom use are frequently viewed in a negative light.

I will not use a condom with my wife. If I have a casual relationship or if I have lots of girlfriends then I will use a condom. (urban male, FGD)

I trust my wife and she trusts me, therefore I will not use a condom with her. Condoms indicate that we do not trust each other. (rural male, FGD)

A major disadvantage of condoms is that use depends heavily on both partners' cooperation. Many of the male respondents reported that their partners had negative attitudes to the use of condoms, as is illustrated in the following statements.

Girls say that they need not eat a sweet that is still wrapped in plastic. (rural male, FGD)

Sometimes if you use a condom with your girlfriend she will think that you don't trust her. Sometimes you will want to please her so you won't use a condom. (urban male, FGD)

Some men declared their right to adopt condoms, irrespective of what their partners felt, given their responsibility to protect themselves against HIV. For them condom use is non-negotiable. They felt quite strongly that they should assert their needs and assume control over their lives. For these people, the desire to protect themselves against the risk of HIV infection was the major reason for condom use.

5.2 Predictors of condom use

A large majority of the respondents reported that they were not using condoms. Of the remainder only a few reported using consistent condom use in the primary sexual relationship, as shown in Table 2. The use of condoms was more consistent in sexual relationship outside the primary sexual relationship. Condom use is higher in non-primary than primary sexual relationships. Nearly 14 per cent of men reported using condoms with their primary sexual relationship while nearly 51 per cent of men reported using the condom with non-primary sexual relationship. Although consistent use offers the best protection against HIV infection, the low number of those reporting ‘always’ using condoms prompted the combining of the categories ‘always’ and ‘occasionally’.

Table 2: Percentage distribution of frequency of condom use with primary and non-primary partners

Frequency of condom use	Primary sexual relationship (per cent)	Non-primary sexual relationship (per cent)
Always	1,9	37,9
Occasionally	12,5	12,8
Beginning	6,3	4,3
Never	79,3	45,0
Number of respondents	248	496

Table 3 shows the relationship between condom use within the primary sexual relationship and selected characteristics of men. Condom use is considerably higher among the better educated than the less educated, and among those living in the urban area than those living in rural area. In primary sexual relationships, the younger age group record a higher level of condom use than the older age group. Generally, men under 35 years are more likely than other men to use condoms with their primary sexual partner. The bivariate analysis also shows that the use of condoms declines as the number of children increases. Condom use is higher among those who desire more children than those who do not, perhaps because the latter are more likely to be using a highly effective method of contraception. The risk of HIV depends, among other things, on the number of sexual partners. Men who engage in sexual relations with more than one partner are obviously at higher risk than those who have a long-term partner within a mutually monogamous relationship. Men with multiple sexual partners are more likely than those who are monogamous to report using a condom; but this is not significant for men within the primary sexual relationship, perhaps because few married or cohabiting men report having other sexual partners. In the multivariate model, the strengths of these associations were attenuated through their directions remained the same. Only two statistically significant effects remained: level of education and place of residence. The odds of using a condom within the primary sexual relationships was significantly higher among the urban, more educated than the rural, less educated. After adjusting for other variables, age was no longer a significant predictor of condom use.

Table 3: The use of condoms within the primary sexual relationship, by selected background characteristics

Background characteristics	Number	Per cent	Odds ratios and 95 per cent confidence intervals
Age (years)
Less than 35	81	21,3*	1,00
35 or more	167	10,8	0,50 (0,21–1,21)
Marital status
Married	149	11,4	1,00
Cohabiting	99	19,2	1,39 (0,53–3,59)
Place of residence
Rural	156	8,3**	1,00
Urban	92	25,0	2,77 (1,16–6,61)
Level of education
Less than secondary	87	2,3**	1,00
Secondary or more	161	21,1	9,36 (2,00–43,87)
Desire for more children with partner
Yes	122	17,2	1,00
No	84	11,9	0,50 (0,17–1,45)
Unsure	42	11,9	0,47 (0,14–1,54)
Number of living children with partner
0–1	94	16,0	1,00
2–4	129	13,2	1,29 (0,50–3,30)
5 or more	24	12,5	2,19 (0,40–12,01)
Two or more sexual partners
Yes	44	22,7	2.09 (0.79–5.50)
No	203	12,8	1.00
Number	248	14,4
Note: *Significant at 5 per cent. **Significant at 1 per cent.

A good measure of risky sexual behaviour is the extent to which those who engage in sexual relations outside the primary sexual relationship use condoms (Adetunji, 2000). Table 4 shows the relationship between condom use outside the primary sexual relationship and selected characteristics of men. Condom use is closely associated with level of education and place of residence. Moreover, use of condoms outside the primary sexual relationship declines with increasing age. Men who are not married and not cohabiting are significantly more likely to use condoms than men who are married or cohabiting. Also, men who have children with their partner are significantly more likely to use condoms than those who do not. The relationship between condom use and desire for children is not significant. The adjusted results indicate that level of education and number of living children are significant predictors of condom use. None of the other variables remain significant, although age is close to achieving a significant effect at the five per cent confidence level.

Table 4: The use of condoms outside the primary sexual relationship, by selected background characteristics

Background characteristics	Number	Per cent	Odds ratios and 95 per cent confidence intervals
Age (years)
Less than 35	416	53,4*	1,00
35 or more	78	37,2	0,57 (0,32–1,00)
Marital status
Married or cohabiting	50	36,0*	1,00
Neither	445	52,4	1,15 (0,57–2,31)
Place of residence
Rural	337	46,6	1,00
Urban	159	59,7*	1,48 (0,98–2,26)
Level of education
Less than secondary	105	33,0**	1,00
Secondary or more	390	55,4	2,23 (1,37–3,65)
Desire for more children with partner
Yes	259	55,6	1,00
No	180	46,1	0,79 (0,51–1,22)
Unsure	56	42,9	0,74 (0,40–1,37)
Number of living children with partner
0	309	42,7**	1,00
1+	186	64,0	1,51 (1,68–3,74)
Two or more sexual partners
Yes	389	50,1	1,05 (0,64–1,73)
No	107	53,3	1,00
Number	496	50,7
Note: *Significant at 5 per cent. **Significant at 1 per cent.

6. Conclusion

As the AIDS epidemic enters its third decade, it is necessary to assess its impact on behavioural change. In the absence of a cure for AIDS, the condom has a powerful role to play in the fight against AIDS. Alongside sexual abstinence and mutually monogamous relationships between uninfected partners, the use of condoms is currently the most effective strategy for preventing the transmission of HIV and other sexually transmitted infections. Obviously, the more sexual partners a person has, the greater is the likelihood of encountering a partner who is infected with and who may transmit the virus (Finer, Darroch & Singh, 1999). In general, men are more likely than women to report having more sexual partners (Orubuloye, Caldwell & Caldwell, 1992; Lansky, Thomas & Earp, 1998; Adetunji & Meekers, 2001).

Awareness of condoms as a method of dual protection against the risk of pregnancy and HIV infection is high. Knowing that condoms protect against pregnancy and HIV/AIDS is a necessary precondition for their acceptance and use. However, such knowledge is a necessary, if not sufficient condition for use, it is not very popular. Condoms are seen as interrupting sexual activity, causing discomfort and ruining the excitement of flesh-to-flesh sex, as found in countless other studies (Mehyrar, 1995; Preston-Whyte, 1999). There is widespread distrust of condoms, especially within the primary sexual relationship. In general, condoms are not seen as appropriate for use in marital relationships. Both men and women expressed strong negative attitudes to the use of condoms within marital relationships. The association of condoms with lack of trust has had a detrimental effect on condom use within marital relationships.

The focus group discussions and in-depth interviews found that women who request condom use are usually viewed in a very negative light. If the woman was to suggest condom use it is likely to lead to conflict, misunderstanding and suspicions of infidelity and violence. Some men also complained that they faced resistance from female partners who are opposed to the use of condoms. Since condoms are highly emotive it is therefore not surprising that many couples choose to take risks in order to avoid the embarrassment or punishment of having to suggest condoms with their partners. Prevention programmes have an important role to play in challenging traditional gender stereotypes and promoting new ways of looking at gender roles. However, it is worth noting that there are somecontrary signs. Some men felt that they had the right to use condoms, irrespective of what their partners felt, in order to protect themselves against the risk of HIV infection.

Consistent with other studies, condom use varies substantially between primary and non-primary sexual relationships (Preston-Whyte, 1999; Adetunji & Meekers, 2001). The level of current condom use reported by men outside the primary sexual relationship exceeds that reported by men within the primary sexual relationship. In the primary sexual relationship, the majority of respondents reported that they are not using condoms. Nearly 14 per cent of men reported using condoms within the primary sexual relationship, compared with 50.7 per cent of men outside the primary sexual relationship.

The association with lack of trust has reduced the acceptability of condoms within the primary sexual relationship. Other studies have also shown that condom use in marital relationship is low and is one of the least frequently used methods of contraception (Gardner et al., 1999; Ali, Cleland & Shah, 2001). Many respondents do not see the need for condoms with their primary sexual relationship because they feel certain that they can trust their partners. This is probably because men in stable, on-going relationships may perceive a higher risk of unwanted pregnancy than disease. In non-primary sexual relationships, men may perceive a higher risk of HIV and are therefore more motivated to protect themselves by using a condom. A study conducted by Evans and colleagues (1997) found that men with non-regular partners might compensate for their increased risky lifestyle by using condoms more frequently. Prevention programmes have an important role to play in promoting condom use in all sexual relationships.

The multivariate analysis shows that condom use is closely associated with level of education. In many parts of South Africa, urban areas are often associated with higher levels of education, better access to health facilities, family planning and other social services. Hence, it is hardly surprising that more educated, urban respondents have a greater propensity to use condoms than the less educated, rural respondents. In rural areas, prevention programmes have an important role to play in creating awareness in order to dispel some of the fears, rumours and false beliefs about condoms.