Quantitative Analysis
Of the 1,092 women included in the analysis, 651 (60 %) women had discussed gel use with their partners by the week 4 visit. Of the 651 who discussed gel with their partners, 578 (89 %) said their partners always knew they were using the gel while 73 (11 %) said their partners sometimes knew they were using it. Women who had discussed gel use with their partners were younger than women who had not. The relationship between discussing gel and age was linear (OR 0.986,
p value 0.007), although differences between age groups were not statistically significant (
p value 0.058) (Table
2).
Table 2
Individual characteristics of women who discussed gel use with their partner compared to women who did not discuss gel use at week 4
| 1092 (100 %) | 441 (40 %) | 651 (60 %) | |
Age |
18–24 year olds | 309 (28 %) | 110 (36 %) | 199 (64 %) | 0.058 |
25–34 year olds | 224 (21 %) | 85 (38 %) | 139 (62 %) | |
35–44 year olds | 265 (24 %) | 111 (42 %) | 154 (58 %) | |
45+ year olds | 294 (27 %) | 135 (46 %) | 159 (54 %) | |
Mean age (SD)a
| 35.0 (11.65) | 36.2 (11.49) | 34.2 (11.70) | 0.007 |
Educational level |
Primary or lower | 535 (49 %) | 228 (43 %) | 307 (57 %) | 0.141 |
Secondary or higher | 557 (51 %) | 213 (38 %) | 344 (62 %) | |
Employment status |
Employed | 184 (17 %) | 81 (44 %) | 103 (56 %) | 0.270 |
Unemployed | 908 (83 %) | 360 (40 %) | 548 (60 %) | |
Head of household |
Partner | 472 (43 %) | 208 (44 %) | 264 (56 %) | 0.090 |
Parent/in-law | 391 (36 %) | 142 (36 %) | 249 (64 %) | |
Self | 116 (11 %) | 50 (43 %) | 66 (57 %) | |
Other | 113 (10 %) | 41 (36 %) | 72 (64 %) | |
Area of residency |
Rural | 857 (78 %) | 353 (41 %) | 504 (59 %) | 0.300 |
Peri-urban/urban | 235 (22 %) | 88 (37 %) | 147 (63 %) | |
Religion |
Zionist | 507 (46 %) | 202 (40 %) | 305 (60 %) | 0.882 |
Shembe | 238 (22 %) | 104 (39 %) | 161 (61 %) | |
Christian-mainstream | 265 (24 %) | 101 (42 %) | 137 (58 %) | |
Other | 82 (8 %) | 34 (42 %) | 48 (58 %) | |
Clinic of enrolment |
Clinic 1—township | 419 (39 %) | 191 (46 %) | 228 (54 %) | 0.019 |
Clinic 2—town | 353 (32 %) | 128 (36 %) | 225 (64 %) | |
Clinic 3—tribal authority | 320 (29 %) | 122 (38 %) | 198 (62 %) | |
Previous MDP participationb
|
No | 1040 (95.5 %) | 425 (41 %) | 615 (59 %) | 0.086 |
Yes | 49 (4.5 %) | 14 (29 %) | 35 (71 %) | |
As shown in Tables
2,
3 and
4, discussing gel use with a partner was associated at the 10 % level with age (
p value 0.058), clinic of enrolment (
p value 0.019), water source (
p value 0.026), and household ownership of cattle (
p value 0.021). Although associated at the 10 % level, relationship to the head of household (LRT
p = 0.694), previous participation in MDP studies (LRT
p = 0.270), contraceptive use (LRT
p = 0.251), sexual frequency (LRT
p = 0.099) and condom use (LRT
p = 0.177) did not contribute to the model in likelihood ratio tests so were not included in the multivariate model. Table
5 presents the output from the final multivariate model. In the multivariate analysis, older women (AOR 0.98,
p value 0.006), women enrolled at clinic 1 (AOR 1.00 v clinic 2 at AOR 1.54,
p value 0.005) and women who lived in households that owned cattle (AOR 0.72,
p value 0.021) were significantly less likely to have discussed gel use with their partners after 4 weeks in the trial.
Table 3
Socio-economic characteristics of women who discussed gel use with their partner compared to women who did not discuss gel use at week 4
| 1092 (100 %) | 441 (40 %) | 651 (60 %) | |
Water source |
Inside house/yard | 333 (30 %) | 127 (38 %) | 206 (62 %) | 0.026 |
Community source | 599 (55 %) | 234 (39 %) | 365 (61 %) | |
Free flowing | 160 (15 %) | 80 (50 %) | 80 (50 %) | |
Fuel for cooking |
Electricity | 364 (33 %) | 143 (39 %) | 221 (34 %) | 0.856 |
Gas | 88 (8 %) | 39 (44 %) | 49 (56 %) | |
Paraffin | 139 (13 %) | 57 (41 %) | 82 (59 %) | |
Wood | 501 (46 %) | 202 (40 %) | 299 (60 %) | |
Household ownership (yes) |
Cattle | 298 (27 %) | 137 (46 %) | 161 (54 %) | 0.021 |
Electricity | 542 (50 %) | 206 (38 %) | 336 (62 %) | 0.112 |
Radio | 954 (87 %) | 385 (40 %) | 569 (60 %) | 0.960 |
Television | 473 (43 %) | 182 (38 %) | 291 (62 %) | 0.262 |
Telephone | 973 (89 %) | 387 (40 %) | 586 (60 %) | 0.240 |
Fridge | 569 (52 %) | 219 (38 %) | 350 (62 %) | 0.183 |
Bicycle | 181 (17 %) | 68 (38 %) | 113 (62 %) | 0.398 |
Table 4
Sexual behaviour characteristics of women who discussed gel use with their partner compared to women who did not discuss gel use at week 4
| 1092 (100 %) | 441 (40 %) | 651 (60 %) | |
Contraceptive use at week 4 |
No | 403 (37 %) | 180 (45 %) | 223 (55 %) | 0.027 |
Yes | 689 (63 %) | 261 (38 %) | 428 (62 %) | |
Average sex in last week |
3 or less | 502 (46 %) | 217 (43 %) | 285 (57 %) | 0.066 |
4–6 | 316 (29 %) | 129 (41 %) | 187 (59 %) | |
7 or more | 274 (25 %) | 95 (35 %) | 179 (65 %) | |
Condom use in last week/4 weeks |
Always | 581 (53 %) | 221 (38 %) | 360 (62 %) | 0.092 |
Never/sometimes | 511 (47 %) | 220 (43 %) | 291 (57 %) | |
Gel use in last week/4 weeks |
Always | 1065 (98 %) | 428 (40 %) | 637 (60 %) | 0.405 |
Never/sometimes | 27 (2 %) | 13 (48 %) | 14 (52 %) | |
Gel group |
0.5 % | 388 (36 %) | 148 (38 %) | 240 (62 %) | 0.361 |
2 % PRO 2000 | 325 (30 %) | 141 (43 %) | 184 (57 %) | |
Placebo | 379 (35 %) | 152 (40 %) | 227 (60 %) | |
Impact on sexual pleasure |
Increased | 813 (74 %) | 336 (41 %) | 477 (59 %) | 0.278 |
Same/less | 279 (26 %) | 105 (38 %) | 174 (62 %) | |
Table 5
Multivariate model comparing women who discussed gel use with their partner to women who did not discuss gel use at week 4
Age (mean) | 0.98 (0.97, 0.99) | 0.006 |
Clinic of enrolment |
Clinic 1 | 1.00 | |
Clinic 2 | 1.54 (1.14, 2,07) | 0.005 |
Clinic 3 | 1.32 (0.97, 1.80) | 0.076 |
Water source |
Inside house/yard | 1.00 | |
Community source | 1.01 (0.76, 1.36) | 0.925 |
Free flowing | 0.70 (0.48, 1.03) | 0.074 |
Household ownership of cattle |
No | 1.00 | |
Yes | 0.72 (0.55, 0.95) | 0.021 |
Women who enrolled at clinic 1 in the township were significantly less likely to have discussed gel use with their partner than women who enrolled at clinic 2 in town. To explore possible reasons for differences in clinic of enrolment, we created a variable to identify clinic specific counsellors. In total 13 staff were responsible for gel adherence counselling in the three clinics during this period of observation. We created a binary variable to compare the three main counsellors at clinic 1 to the other 10 counsellors (not presented). When included in the multivariate model, there was no longer a difference between women who discussed gel use with their partners depending on whether they enrolled at clinic 2 (AOR 1.00 95 % CI 0.64, 1.56) or clinic 3 (AOR 0.81 95 % CI 0.50, 1.32) compared to clinic 1. The women counselled by the main three counsellors at clinic 1 were significantly less likely to have discussed the gel with their partners than women counselled by any of the other 10 counsellors at any clinic (AOR 0.56 95 % CI 0.36, 0.88).
We repeated the analysis comparing women who had talked to their partners about the gel to women who had not, any time during their participation in the trial, which was up to a maximum of 52 weeks. By the end of the trial, 84 % (914/1092) of women in this sample had discussed the gel with their partners. Based on univariate associations, the following variables were included in multivariate analysis: water source, cattle ownership, gel randomisation group and consistent gel use. By week 52, 73 % of the women in this sample were defined as consistent gel users. In this model, women who relied on the cheapest water source which was free flowing water such as from a river or stream (AOR 0.53 95 % CI 0.32, 0.89), lived in households that owned cattle (AOR 0.66 95 % CI 0.47, 0.93), and were randomised to 2 % PRO2000 (AOR 0.63 95 % CI 0.42, 0.96) or placebo gel (AOR 0.65 95 % CI 0.44, 0.98, compared to 0.5 % PRO2000) were significantly less likely to have discussed gel with their partners, and women who consistently used gel were significantly more likely than women who had not, to have discussed gel with their partners (AOR 1.51 95 % CI 1.06, 2.14).
Throughout the course of the trial, 31 women reported that it was difficult to insert the gel or inconvenient to use it. In open-ended questions, 12 women reported that the difficulty or inconvenience was due to the fact that they had not told their partners about the gel at the time of use. By week 52, there was a strong association between whether women had discussed the gel with their partners and whether women found it difficult or inconvenient to use gel as only four of these women proceeded to inform their partners (p value <0.001). This association remained in the multivariate model at week 52, without substantially changing the other associations (AOR 0.08 95 % CI 0.02, 0.27).
Qualitative Analysis
Four main themes emerged from the qualitative data: socio-cultural norms of sexual communication, expectations regarding communication about microbicides, women’s experiences of discussing microbicides with their partners, and attitudes and experiences of using microbicides without a partner’s knowledge. We present findings from each of these themes below.
Sexual Communication
All respondents in the community FGDs and most of the women in the IDIs had a shared understanding of the traditional norms regarding sexual communication. Within this traditional context, women were not supposed to talk about sex, initiate sex, or even refuse to have sex with their partner. However, there was a palpable schism in opinion about how these traditional norms informed contemporary sexual communication. Approximately half the FGD respondents and about a quarter of IDI respondents, believed that these traditional norms still dominated, while the rest believed that the advent of HIV had altered social expectations regarding sexual communication. The differences were predominately gendered and generational, with most women and younger men believing that communication norms in relationships had changed or were changing.
These shifting expectations were evident in narratives in both the FGDs and IDIs. There was a tension between different expectations regarding the role of women as both submissive and independent. These contradictory expectations appeared to be informed by traditional images of women obeying their husbands, and nationalist images of independent women, which exalt women’s empowerment and promote women’s rights.
Nonetheless, the overwhelming sentiment from the FGDs and IDIs was that in contemporary KwaZulu-Natal, both women and men must break with tradition and talk about sex in response to the HIV epidemic, as this young woman explains:
‘There should be no secrets… It’s not like the olden days. Tell him that there’s something I have found and I will be using it to protect ourselves because no-one wants to die, everybody wants to live, no-one wants to be HIV positive’ (Community FGD, 24 year old woman).
There were risks involved for women who discussed sex. For merely initiating communication about sex, women talked about the risk of abandonment, mistrust, financial marginalisation, verbal conflict or even physical abuse.
Expectations Regarding Communication About Microbicides
The vast majority of FGD respondents and more than three quarters of the IDI participants believed that women should discuss gel use with their partners before using it. However, this apparently simplistic statement was not without its complexity, as this quote illustrates:
‘It is important to tell (uku
tshela) your partner about the things that you do but only if you know that your partner will agree with you’ (Trial IDI, 38 year old woman) (emphasis added).
Similarly, there were a number of examples in the IDIs where women clearly drew on particular scripts of expected behaviour depending on what they wanted to do. For example, if they wanted to tell their partner about the gel they claimed that culturally they were supposed to talk about sex; alternatively if they did not want to tell their partners about the gel they claimed that culturally they were not supposed to talk about sex. Women at times used these scripts of expected behaviour interchangeably, assigning different priorities to often competing expectations of behaviour depending on the topic.
In the FGDs and IDIs there were different expectations about the form that the discussion about microbicides should take which implied different expectations regarding the decision making process and the role of the male partner. In the FGDs, the main isiZulu words used in this context were imvume, cela, xoxa, tshela and azisa. Imvume means ‘permission’ and was used in the context of women asking men for permission to use the gel with men being the ultimate decision makers about whether or not women could use gel. Cela means to ‘ask’ or ‘negotiate’, xoxa means to ‘talk’ or ‘tell someone’ about something, tshela means to ‘tell’ or ‘narrate’ or ‘give an account’ of something, and azisa means to ‘inform’ but was also used in relation to convincing someone of something. Cela, xoxa, tshela and azisa all had an inference of negotiation in the process of seeking agreement.
However, a few mainly younger women used either
tshela or
azisa in the sense of literally telling or informing the partner without any expectation of a negotiation or any requirement for permission or consent. In these rarer examples, women were viewed as the ultimate decision makers about whether or not to use gel. The following quote illustrates this perspective but is unusual in coming from a married instead of an unmarried woman:
‘I think I must discuss (ngi
phumele
obala – speak out or pronounce) so that he will know that I am using this thing (gel). This is my life not his life, I can tell (ngingam
tshela) him that there is something that I am using like this and this, I am protecting myself from the diseases because you are not faithful, I do not know the places you go, you cannot trust a person these days. I can tell (ngim
tshele) him that I am using this thing father (husband) with my life, the life is mine’ (Community FGD, 44 year old woman).
In the IDIs, women most frequently used the word tshela to describe their own discussions about the gel with their partners. In addition, woman in the IDIs sometimes used the word chaza meaning to explain. The use of this word is understandable in the IDIs, although it never emerged in the FGDs, as women in the trial were describing how they ‘explained’ the gel within the context of the clinical trial.
Throughout the IDIs, there was a sense that the whole discussion hinged on the woman ‘knowing’ her partner and being able to guess his response well enough to find the right words, use the right strategy, at the right time, as this quote demonstrates:
‘A person knows her partner and how he reacts if he is told something’ (Trial IDI, 28 year old woman).
Women’s Experiences of Discussing Microbicides
Of the 79 women interviewed at week 4, 56 had talked to their partners about the gel. There were different reasons for talking about the gel, different ways of talking about it, and discussions took place at different times in the process of introducing gel into the relationship.
Reasons for Talking About Gel Women offered two main reasons for discussing the gel with their partners. The first reason was that the couple usually discussed sex and the women felt that as the gel would be present during sex, men should be aware of it. The majority of the women interviewed were in long-term stable relationships. Many of the women described these as loving relationships in which they trusted each other and did not have secrets from each other, as stated by this woman:
‘We don’t hide things from each other, he also doesn’t hide anything from me. We usually discuss things before doing them’ (Trial IDI, 22 year old woman).
The second reason, which often overlapped with the first reason, was to avoid conflict if the partner found out about the gel. The concerns they expressed included whether the partner noticed the gel during sex; found the applicators; heard about the gel and suspected his partner was using it; if he had penile problems; or if the gel was found to have safety concerns. Women were also concerned that if their partner felt a difference during sex they may assume the woman was having sex with someone else, which is based on a commonly held myth that men can physically tell during sex if the woman has had sex beforehand.
Ways of Talking About Gel Discussion about the use of microbicides took place in different ways. In the majority of cases, women introduced the microbicide study to their partners and then discussed and negotiated the use of the gel. A few women explained that they were with their partners when they first heard about the microbicide study and this triggered the discussion about the potential of the woman joining the study. A few other women described how they simply told their partners about the microbicide study and men accepted their decision to use gel.
The strategies employed by women to discuss the gel with their partners largely depended on the decision-making roles in the relationships. In some examples, it was clear that the ultimate decision of whether or not to use the gel rested with the man. In these cases woman described how they had to convince, cajole and plead with their partners in order to use the gel. In the vast majority of cases, the decision-making was based on a process of negotiation with the aim of reaching joint agreement. In some cases, the decision-making process was on going and involved continuous dialogue, as this quote demonstrates:
‘The first time that I heard from my friends about the study, I sat down with him, talked to him about it, and then he allowed me. Even by the time I came back from the clinic, I as well sat down and informed him about what had been said. I also told him about the gel that there is a preventative thing that they have also given us at the clinic which is in a form of a gel and he asked how it is being used. I then told him. He then said, can I please demonstrate for him how is it being done, I then did as taught, I demonstrated for him’ (Trial IDI, 29 year old woman).
There were other cases where men only provided non-verbal cues. For example, some women talked about the fact that their male partner walked away during the conversation, in which case the woman would assume he was not particularly happy about the gel but was not going to object to her using it. In other cases, women read men’s silence and lack of objection as acceptance of the gel. Throughout most of the narratives, it was evident that women firstly decided whether they wanted to use the gel, then opened up discussions with their partners. This appeared to shift the balance of decision-making in the women’s favour.
Less than a quarter of women viewed the decision to use the gel as theirs alone. However, some were willing to state this independence of decision making even when faced with culturally loaded questions, as the exchange below shows:
-
Interviewer: ‘Does your partner allow (uyaku
vumela) you to insert gel?’
-
Participant: ‘I insert gel on my own, not because my partner allows (engi
vumela) me to insert’ (Trial IDI, 40 year old woman).
In a few cases, women negotiated their participation in the study with their partner on the basis of only partial information. For example, a few women told their partners they had joined a research study but did not tell them anything about the gel. Other women told their partners that they were using a vaginal gel, although they did not tell them that it was intended as a HIV prevention method. Instead, the gels were described as preventing other sexually transmitted infections, as treatment for vaginal problems, or to prevent cervical cancer. It appeared that by providing partial information, women felt that they could claim they had tried to explain the gel if their partners subsequently challenged them about it. The main reason for only providing partial information appeared to be to avoid having to discuss HIV, which was tied up with issues of trust and fidelity.
Timing of Talking About Gel The majority of women who discussed the gel with their partners did so after first learning about the trial at a study screening visit, but before enrolling in the trial. However, about a fifth of women who discussed gel use with partners did so after enrolling in the trial. In some of these cases it appeared that women only wanted to enter into negotiations with their partners after deciding for themselves whether they wanted to join the study and use the gel. While the women who delayed enrolment until after talking to their partner’s may reflect more traditional gendered decision-making roles, women’s decision to discuss gel with their partners only after enrolling in the trial provides another example of the decision-making being shifted to women’s advantage. About a sixth of other women initially used the gel without discussing it with their partners at all and only told them about it when they found study material or noticed a difference during sex.
Although the majority of women told their partners that they were using the gel, few referred to its use every time they had sex. There were three main ways in which gel insertion was managed before sex: (1) the majority of women inserted the gel discretely before sex without telling their partner; (2) a smaller group of women overtly told their partner they were going to insert the gel when passions were roused. Some of these women interpreted their partner’s willingness to wait for them to insert as an act of support. Only a very few women actually inserted the gel in front of their partners and only one woman talked about her partner inserting the gel for her; and (3) a small group of women described how their partners would remind them to insert the gel as a hint for sex, as this quote shows:
‘If he wants sex he just says ‘”gel”, or, “are we ticking” (ticking refers to ticking the coital diary), I know that it is time for sex, so … he has found the easy way to ask for sex’ (Trial IDI, 34 year old woman).
In this way, women described the gel as encouraging communication about sex, which often opened up opportunities for other discussions, for example about condom use. There were numerous examples of women stating that talking about sex and the gel with their husband was more difficult than with an unmarried partner, and that it was especially difficult for women to use the gel without talking to their partner if they were living together.
Use of Microbicides Without a Partner’s Knowledge
Although everyone in the FGDs agreed that ideally male partners should know about the gel before it is used, a minority of women and younger male respondents thought that women could be justified in using the gel without their partner’s knowledge in some circumstances. One example was if a woman had experience of her partner refusing specific requests previously. In these circumstances, some women believed that women should use the gel without telling their partner:
‘You do not do something without asking him, you know your partner. You firstly ask him that can I use this or can we use this. If you see that he is not allowing it, you just keep quiet and continue using it secretly’ (Community FGD, 41 year old woman).
Other examples of when use without a partner’s knowledge was justified were if he was HIV positive, had other partners, refused to use condoms, or was frequently drunk, thereby unreliable in terms of condom use. The sense in these examples was that if men failed to be what would traditionally be considered a good and reliable husband, then his partner had the right to breach traditional norms in response to his failings as a husband. In these circumstances, the respondents commented that if the woman was caught using the gel secretly, then she would have to tell her partner the truth. However, they all agreed that this could lead to accusations of infidelity and would cause conflict in the relationship.
After 4 weeks in the trial, 23 of the 79 trial participants interviewed had not discussed their use of the microbicide gel with their partner. All but one of these women had at least one subsequent interview, and by their last interview, 15 had still not discussed gel with their partner (4 at week 24 and 11 at week 52). When talking about using gel without their partner’s knowledge, women mainly referred to using it secretly (imfihlo) or hiding it (fihla) and described how it was only theirs (not their partners) as in the description of ‘kuphela ukwazi kwami’, mine alone.
Four main reasons for women using the gel without telling their partner emerged from the data. Firstly, most of the women still thought it was preferable to discuss gel with their partner and hoped to do so some time in the future but had not yet found the right time. However, of the seven women who subsequently told their partners about the microbicide gel, half only did so after the partner found the applicators or noticed the gel during sex. Secondly, some of the women, mainly young and unmarried, did not think that it was important to discuss the gel with their partners and had no intention of doing so. Thirdly, some women did not want to risk talking to their partners about the gel as they assumed that they would object. This young woman illustrates that by not living in her partner’s house (not being married or cohabitating) she felt she had more ability to decide about the gel:
‘I think he will have a problem, maybe say I should stop the gel, so I thought it is better to continue and hide it from him. He cannot control me because it is my home’ (Trial IDI, 29 year old woman).
Fourthly, about half a dozen women explained that they had not discussed the gel with their partners because they were afraid that they would take their use of gel as a sign of mistrust and respond violently:
‘He doesn’t know about the gel and I don’t want him to know because he is jealous and if he finds out he will beat me. I can’t just talk, I’m afraid of him’ (Trial IDI, 21 year old woman).
Most of the women who were still using gel without their partner’s knowledge after about a month in the study were concerned in case their partners found out about the gel before they had chance to discuss it. However, a few had decided that it was a risk they were willing to take. Using the gel without the knowledge of a stable partner was not viewed as the ideal, but importantly was viewed as possible in some circumstances.
We did not systematically ask women if they would use the gel regardless of their partner’s response. However, of the women who discussed the gel with their partners, approximately a tenth spontaneously reported they would not have used the gel if their partners had objected. An equal number spontaneously reported the opposite, that they would have used the gel even if their partners had objected.
Most women were able to describe strategies that they used to insert the gel before sex, such as going to the toilet to insert the gel if passions were roused. However, it was evident in some women’s accounts that it was more difficult to use the gel when partners were unaware of it, and that inserting additional applicators of gel between sex acts could be particularly difficult.
Only a few women in the IDIs admitted having secondary casual partners and in the main, they had not talked to these partners about the gel. There was no suggestion in the qualitative data that women who did not use condoms were more likely to use the gel without their partner’s knowledge.