This study examined whether a woman’s confidence or capability to negotiate condom use with her sexual partner(s) has any consequence on the actual use of condoms. The study also sought to describe the proportion of women who feel confident to negotiate condom use and prevalence of condom use at the last sexual intercourse in Tanzania. Findings reveal a large proportion of women being confident to negotiate condom use with their partners. Generally, condom use at the last sexual intercourse was significantly dependent on characteristics of the individual. A majority of the women who did not use condoms at the last sexual intercourse reported that they trust their partners.
Overall results show a strong evidence of increased likelihood of condom use by confidence to negotiate it. This observation remained adamantly significant even after controlling for marital status, age, education, religion, household wealth status, condom knowledge and number of sexual partners in the preceding 12 months. This suggests that a woman’s confidence or capability to speak for herself about condom use with her sexual partner represents her protection against acquisition of not only HIV/AIDS and other STIs but also unintended pregnancies. On the other hand, the results imply that women who lack confidence to negotiate condom use with their sexual partners may be exposed to unprotected intercourse, thus at risk of contracting STIs including HIV/AIDS and consequently being more vulnerable to adverse sexual and reproductive health outcomes. This is consistent with findings from other studies in which it is reported that a woman’s confidence to negotiate condom use correlates with higher levels of condom use [38, 39]. The literature further shows that women who are in relationships where they have limited decision-making powers are less likely to use condoms than those with adequate control of their relationships [40, 41].
Regarding marital status, condom use was lowest among women who were married or living with partners as married and highest among women who were single, and ever married. This observation persisted in the multivariate model, with both women who were single, and ever married being almost four times more likely than women who were married to have used a condom at the last sexual intercourse. This is consistent with extant evidence [42, 43] showing that, in marriage, couples will use condoms most likely as a family planning method, not primarily for disease prevention purposes . However, where one partner suspects or know that the other partner is not faithful or is infected with STIs especially HIV, condom use may be expected, although fear of breaking the relationship coupled with religious stance that condom use is a sin, makes it difficult for such married couples to ask their partners to use condoms . One study in Malawi considered condom use in marriage as an ‘intruder’ in the domestic space , to mean that condom use in marriage threatens the relationships thus less expected for purposes other than pregnancy prevention. Condom use is denied due to claims that they connote mistrust or dearth of love intimacy [45, 46]. Therefore, low condom use in marriage was expected, given the close association between marriage and fertility and the fact that condom use is not even one of the popular fertility control methods in Tanzania . On the other hand, the increased condom use among unmarried and ever married women is likely due to perceived risk of contracting STIs including HIV and unintended pregnancies.
Age and condom use at the last sexual intercourse related inversely, with the likelihood of condom use declining rapidly and constantly with ageing and vice versa. This is consistent with the recent Tanzania Demographic and Health Survey findings . The greater use of condoms observed among the youngest (<20) women is probably because of perceived risk of contracting STIs including HIV and unintended pregnancies because most of them were unmarried. Despite being low, condom use in subsequent age categories in which most women were married was probably for birth spacing purposes primarily, not as a disease control measure. This is also supported by the fact that over three-quarters of women who did not use condoms at the last sexual intercourse said that they did so because they trust their partners and 84.4% of them were married. Additionally, the likelihood of condom use in the oldest age category (≥40) was even much slim, a reflection of aging towards menopausal where fertility control mechanisms including condom use are rarely used because of declining reproductive capacity or fecundity with age .
Regarding education attainment, women with at least a secondary education were twice as likely as those who have never been to school to have used a condom at their last sexual intercourse. The likelihood of condom use was similar between women with primary education and those who have never been to school. The evidence of condom use with secondary or higher education is likely due to an imperative role that education plays on societal transformation and also that education enhances women’s self-esteem, self-confidence, ability to make decisions and freedom of expression  concerning their sexual and reproductive inclinations. This underscores a need to promote women’s education beyond primary school as a prerequisite for change (e.g. behavioural change) in all aspects of their life. Education functions as a one powerful input upon which numerous outcomes, including informed choices of safer sex options, upshot. Education is also acknowledged in the literature as a catalyst for change in gender relations [48, 49].
Even though women who reported multiple sexual partners in the last 12 months were few, they were more likely than women who had not more than one sexual partner in the same period to have used a condom at the last sexual intercourse. Engaging in sexual relationships with multiple partners is a risky behaviour  and emphasis has always been centering on condom use at each risky intercourse to ensure protection against STIs including HIV [27, 50]. Therefore, it may be because of perceived risk of contracting STIs especially HIV that women with multiple sexual partners were more likely to use condoms. This agrees with findings from another study in Tanzania, where high-risk sexual behaviour was associated with increased condom use .
Finally, condom use at the last sexual intercourse was less likely among women who reported that people cannot reduce their chances of contracting HIV by using a condom every time they have sex compared to those who did. This observation reflects a true context, since people may not use condoms unless they believe that condoms are capable of preventing transmission of HIV or pregnancy. This may be linked to condom misconceptions or negative outlook towards condom effectiveness  and condom use which some women, depending on their culture, values and norms, may be having. Therefore, it is important that interventions that promote condom use also highlight key issues about condoms effectiveness.
Unlike other HIV prevention methods such as male circumcision where a female partner benefits indirectly, condom use protects both partners. For uninfected women, it is the strategy that they have the most immediate ability to influence. A woman could ask her uninfected male partner to get circumcised in order to reduce his likelihood of acquiring HIV (from her, or from another partner); she could ask her already-HIV-infected partner to get on antiretrovirals (ARVs) or get on ARVs herself to lessen her chances of passing the infection on; and she could negotiate for condom use - condom use being the least technically and logistically demanding of the three, and also a strategy that can be accomplished within minutes.
Condom use was self-reported with no means to validate the responses other than probing the respondent. We understand that self-reports of sexual behaviour are often invalid and unreliable as already known [52–54]. Also these findings may not be generalized to the entire population of Tanzania since three districts only were studied. Similarly, no causal inferences may be drawn because temporality cannot be established in cross-sectional studies.