AIDS Care, 2015 Vol. 27, No. 8, 1020–1024, http://dx.doi.org/10.1080/09540121.2015.1018862
Pregnant women’s experiences of male partner involvement in the context of prevention of mother-to-child transmission in Khayelitsha, South Africa Kirsty Brittaina*, Janet Giddyb, Landon Myerc, Diane Cooperd, Jane Harriesd and Kathryn Stinsona,e a
School of Public Health & Family Medicine, University of Cape Town, Cape Town, South Africa; bWestern Cape Department of Health, Khayelitsha SubStructure, Cape Town, South Africa; cDivision of Epidemiology & Biostatistics, School of Public Health & Family Medicine, University of Cape Town, Cape Town, South Africa; dWomen’s Health Research Unit, School of Public Health & Family Medicine, University of Cape Town, Cape Town, South Africa; eMédecins Sans Frontières, Khayelitsha, Cape Town, South Africa (Received 19 September 2014; accepted 10 February 2015) Male partner involvement (MPI) has been identified as a priority intervention in programmes for the prevention of mother-to-child transmission (PMTCT) of HIV, but rates of MPI remain low worldwide. This study used a quantitative survey (n = 170) and two focus group discussions (FGDs) with 16 HIV-positive pregnant women attending a public sector antenatal care service in Khayelitsha, South Africa, to examine the determinants of high levels of involvement and generate a broader understanding of women’s experiences of MPI during pregnancy. Among survey participants, 74% had disclosed their status to their partner, and most reported high levels of communication around HIV testing and preventing partner transmission, as well as high levels of MPI. High MPI was significantly more likely among women who were cohabiting with their partner; who had reportedly disclosed their HIV status to their partner; and who reported higher levels of HIV-related communication with their partner. FGD participants discussed a range of ways in which partners can be supportive during pregnancy, not limited to male attendance of antenatal care. MPI appears to be a feasible intervention in this context, and MPI interventions should aim to encourage male partner attendance of antenatal care as well as greater involvement in pregnancy more generally. Interventions that target communication are needed to facilitate HIV-related communication and disclosure within couples. MPI should remain a priority intervention in PMTCT programmes, and increased efforts should be made to promote MPI in PMTCT. Keywords: male partner involvement; HIV-related communication; PMTCT; pregnancy; South Africa
Introduction Much progress has been made in prevention of motherto-child transmission (PMTCT) programmes over the past decade. In South Africa (SA), the mother-to-child transmission rate is estimated to be 2.7% at six weeks of age (Goga, Dinh, Jackson, for the SAPMTCTE Study Group, 2012). However, challenges remain in PMTCT (Barron et al., 2013) and novel approaches to improving programme outcomes are needed. Male partner involvement (MPI) has been proposed as a priority intervention, and has been shown to significantly improve antiretroviral (ARV) adherence during pregnancy in SA (Peltzer, Sikwane, & Majaja, 2011). Although MPI could strengthen multiple aspects of PMTCT, rates of MPI remain low worldwide (World Health Organization [WHO], 2012) and MPI, where present, is largely limited to financial support (Byamugisha, Tumwine, Semiyaga, & Tylleskär, 2010; Nkuoh, Meyer, Tih, & Nkfusai, 2010). Documented barriers to MPI include poor communication within couples (Ditekemena et al., 2012; Kalembo, Yukai, Zgambo, & Jun, 2012; Reece, Hollub, Nangami, & Lane, 2010), less stable *Corresponding author. Email: [email protected] © 2015 Taylor & Francis
relationship status (Larsson et al., 2010; Morfaw et al., 2013) and traditional gender roles (Nkuoh et al., 2010; Reece et al., 2010; Theuring et al., 2009), including the belief that antenatal care is purely a women’s responsibility. Little is known about the role of MPI in the context of PMTCT in SA and other parts of the region. Although MPI has been the focus of much recent research, there is a lack of consensus about its definition (Ditekemena et al., 2012; Montgomery, van der Straten, & Torjesen, 2011), and the majority of research has focused on quantitative documentation of male attendance at antenatal clinics and male HIV testing. Given documented difficulties in attending antenatal appointments (Nkuoh et al., 2010; Reece et al., 2010), a broader view of MPI is needed. This article reports formative research that explored MPI among HIV-positive pregnant women using a quantitative survey and two focus group discussions (FGDs), with a view to examining the determinants of high levels of involvement and generating a broader understanding of women’s experiences of MPI, respectively.
AIDS Care Methods Participants The study was conducted at an antenatal service in Khayelitsha, SA, between July and November 2013. Pregnant women were eligible to participate if they were HIV-positive and had a male partner, where “partner” was defined as a primary sexual partner. Eligible participants were identified with the assistance of health care providers and support groups for HIV-positive pregnant women. All participants provided written informed consent prior to participation. The study was approved by the University of Cape Town’s Faculty of Health Sciences Human Research Ethics Committee and the local health authority.
Data collection A trained fieldworker conducted the survey and FGDs in isiXhosa using a paper-based questionnaire which was developed and piloted prior to data collection. Questions were asked regarding sociodemographic characteristics and self-reported disclosure of HIV status. Levels of HIV-related communication with partners were assessed based on reported discussion of HIV testing and preventing horizontal transmission. The number of topics discussed was summed to calculate a communication score, such that 0–1 indicated low and 2 indicated high levels of communication. Given the lack of a standardised measure (Montgomery et al., 2011), MPI was assessed using a range of activities based on previous published research (Byamugisha et al., 2010). In this study, these activities included partner involvement in: providing broad financial support for antenatal visits; knowing when visits were due; discussing what happened during visits; accompanying their partner to visits; discussing ways to prevent PMTCT; and providing support for ARV adherence. The number of activities that partners reportedly carried out was summed to calculate an MPI score, such that 0– 3 indicated low and 4–6 indicated high involvement. The Cronbach’s alpha coefficient for this scale was 0.66. Based on preliminary results of the survey, a FGD guide was developed to obtain more in-depth information on women’s experiences of MPI during pregnancy, as it was felt that group processes may help to elucidate more nuanced experiences. MPI was not defined for the purpose of the FGDs, thereby allowing participants to put forward their own views of what activities constitute MPI. The FGDs were digitally recorded.
Data analysis Variables significantly associated with MPI were identified using χ2 or Fisher exact tests for categorical variables and t tests for continuous variables. Odds ratios
1021
(OR) with 95% confidence intervals (CI) were calculated. Variables significantly associated with MPI in bivariate analysis (at p < 0.05) were included in a logistic regression model. Data were analysed using Stata 12 (StataCorp Inc, College Station, Texas, USA). The FGDs were transcribed and translated into English, and were back-translated for quality assurance. Thematic analysis was used to code the transcribed discussions according to the repeated themes which emerged.
Results Sociodemographic characteristics A total of 170 women participated in the survey, and 16 women participated in the FGDs. Detailed sociodemographic characteristics are presented in Table 1. Compared to their male partners, participants were younger and had lower levels of education and employment. Just over a quarter of participants were married (26%), and 52% of participants were cohabiting. Survey results Participants reported high levels of HIV-related communication with partners. Seventy-four per cent had reportedly disclosed their HIV status to their partner, but only 54% reported that they knew their partner’s HIV status. Most participants reported high levels of partner involvement in various pregnancy-related activities, and many (70%) reported that their partners provided support towards ARV adherence. However, only 35% reported that their partners accompanied them to the clinic for antenatal visits (Table 2). In a multiple logistic regression model (Table 3), high MPI was significantly associated with cohabitation (OR: 7.1; 95% CI: 1.8–27.7), disclosure (OR: 4.5; 95% CI: 1.3–15.6) and high levels of HIVrelated communication (OR: 6.6; 95% CI: 1.9–23.0). FGD results Although not all partners were described as supportive, there was unanimous agreement among FGD participants that pregnancy is a shared responsibility. Examples of ways in which partners could be supportive included helping at home, looking after children, enquiring about the pregnancy, providing money to go to the clinic and assisting their female partners with self-care activities when advanced pregnancy made this difficult. Partners could also provide support for ARV adherence by reminders to take ARVs; bringing their ARVs with a glass of water; and fetching ARVs from the clinic. The participants reported reluctance among male partners to attend antenatal visits, as they are prevented from entering the clinic by security guards. Despite this, women supported the idea of a male-friendly facility
1022
K. Brittain et al.
Table 1. Description of participant and partner characteristics.
Table 2. HIV-related communication, disclosure and MPI.
Variables
Variables
Survey participants (n = 170) Median age [Inter-quartile range (IQR)] 29 Race Black/African 170 Language English 1 isiXhosa 168 Sesotho 1 Education Primary school 5 Some secondary education 110 Matric/tertiary education 55 Employment Employed 75 Partners’ median age (IQR) 33 Partner’s education Primary school 5 Some secondary education 60 Matric/tertiary education 90 Unknown 15 Partner’s employment Employed 142 Marital status Married 44 Cohabitation status Cohabiting 89 Number of weeks pregnant at time of interview Second trimester 42 Third trimester 128 Gravidity First pregnancy 29 Decision making regarding pregnancy Unplanned pregnancy 115 FGD participants (n = 16) Median age (IQR) 29.5 Marital status Married 4 Cohabitation status Cohabiting 8 Number of weeks pregnant at time of interview Second trimester 3 Third trimester 13 Gravidity First pregnancy 4
n (%) (26–33) (100) (1) (99) (1) (3) (65) (32) (44) (29–37) (3) (35) (53) (9) (84) (26) (52) (25) (75) (17) (68) (26–35.5) (25) (50) (19) (81) (25)
and suggested that partners be encouraged to attend at least one visit, because they wanted them to be informed about pregnancy.
Discussion This research explored MPI during pregnancy as a potentially useful mechanism to facilitate PMTCT.
HIV-related communication Has discussed HIV testing with partner Has discussed preventing horizontal transmission with partner Disclosure to partner Participant has not disclosed Participant has disclosed Knowledge of partner’s HIV status Participant does not know partner’s status Participant knows partner’s status Items in MPI score (n = 169) Provides broad financial support for participant’s antenatal visits Knows when participant’s antenatal visits are Discusses with participant what happens during antenatal visits Reported partner discussion of PMTCT Provides support for participant’s ARV adherence Accompanies participant to the clinic for antenatal visits MPI score (n = 169)a 0 (no items with a positive response) 1 2 3 4 5 6 (all items with a positive response) Median MPI score (IQR)
n (%) 136 (80) 149 (88)
44 (26) 126 (74) 79 (46) 91 (54) 144 (85) 161 (95) 162 (96) 151 (89) 119 (70) 59 (35)
2 (1) 3 (2) 7 (4) 13 (8) 25 (15) 74 (44) 45 (27) 5 (4–6)
a
Calculated by summing MPI activities detailed above.
High levels of HIV-related communication and disclosure were reported, and participants reported MPI in a range of activities. Higher levels of MPI were significantly more likely among participants who were cohabiting; who had disclosed their HIV status; and who reported higher levels of HIV-related communication with their partner. FGD participants discussed a range of pregnancy-related activities, not limited to male attendance at antenatal visits, as being indicative of MPI. Reported disclosure to male partners in this study was high compared to other South African and African settings (Crankshaw et al., 2014; Makin et al., 2008; Roxby et al., 2013). The high reported levels of HIVrelated communication are encouraging and suggest the normalisation of HIV (Zuch & Lurie, 2012). Consistent with previous published research, we observed higher levels of MPI among participants who reported higher levels of communication with partners (Ditekemena et al., 2012; Kalembo et al., 2012; Reece et al., 2010) and stable relationship status (Ditekemena et al., 2012; Larsson et al., 2010; Morfaw et al., 2013). The high reported levels of MPI overall indicate that gender norms
AIDS Care
1023
Table 3. Factors associated with MPI. Unadjusted association Variables Gravidity First pregnancy Has previously been pregnant Marital status Not married Married Cohabitation status Not cohabiting Cohabiting Knowledge of partner’s HIV status Does not know partner’s status Knows partner’s status Disclosure of HIV status Has not disclosed to partner Has disclosed to partner HIV-related communication Low High
Adjusted association
OR
95% CI
P value
Reference 3.7
(1.2–10.4)
0.009
Reference 4.7
(1.1–43.0)
0.026
Reference 10.9
(3.0–58.6)
Pregnant women’s experiences of male partner involvement in the context of prevention of mother-to-child transmission in Khayelitsha, South Africa Kirsty Brittaina*, Janet Giddyb, Landon Myerc, Diane Cooperd, Jane Harriesd and Kathryn Stinsona,e a
School of Public Health & Family Medicine, University of Cape Town, Cape Town, South Africa; bWestern Cape Department of Health, Khayelitsha SubStructure, Cape Town, South Africa; cDivision of Epidemiology & Biostatistics, School of Public Health & Family Medicine, University of Cape Town, Cape Town, South Africa; dWomen’s Health Research Unit, School of Public Health & Family Medicine, University of Cape Town, Cape Town, South Africa; eMédecins Sans Frontières, Khayelitsha, Cape Town, South Africa (Received 19 September 2014; accepted 10 February 2015) Male partner involvement (MPI) has been identified as a priority intervention in programmes for the prevention of mother-to-child transmission (PMTCT) of HIV, but rates of MPI remain low worldwide. This study used a quantitative survey (n = 170) and two focus group discussions (FGDs) with 16 HIV-positive pregnant women attending a public sector antenatal care service in Khayelitsha, South Africa, to examine the determinants of high levels of involvement and generate a broader understanding of women’s experiences of MPI during pregnancy. Among survey participants, 74% had disclosed their status to their partner, and most reported high levels of communication around HIV testing and preventing partner transmission, as well as high levels of MPI. High MPI was significantly more likely among women who were cohabiting with their partner; who had reportedly disclosed their HIV status to their partner; and who reported higher levels of HIV-related communication with their partner. FGD participants discussed a range of ways in which partners can be supportive during pregnancy, not limited to male attendance of antenatal care. MPI appears to be a feasible intervention in this context, and MPI interventions should aim to encourage male partner attendance of antenatal care as well as greater involvement in pregnancy more generally. Interventions that target communication are needed to facilitate HIV-related communication and disclosure within couples. MPI should remain a priority intervention in PMTCT programmes, and increased efforts should be made to promote MPI in PMTCT. Keywords: male partner involvement; HIV-related communication; PMTCT; pregnancy; South Africa
Introduction Much progress has been made in prevention of motherto-child transmission (PMTCT) programmes over the past decade. In South Africa (SA), the mother-to-child transmission rate is estimated to be 2.7% at six weeks of age (Goga, Dinh, Jackson, for the SAPMTCTE Study Group, 2012). However, challenges remain in PMTCT (Barron et al., 2013) and novel approaches to improving programme outcomes are needed. Male partner involvement (MPI) has been proposed as a priority intervention, and has been shown to significantly improve antiretroviral (ARV) adherence during pregnancy in SA (Peltzer, Sikwane, & Majaja, 2011). Although MPI could strengthen multiple aspects of PMTCT, rates of MPI remain low worldwide (World Health Organization [WHO], 2012) and MPI, where present, is largely limited to financial support (Byamugisha, Tumwine, Semiyaga, & Tylleskär, 2010; Nkuoh, Meyer, Tih, & Nkfusai, 2010). Documented barriers to MPI include poor communication within couples (Ditekemena et al., 2012; Kalembo, Yukai, Zgambo, & Jun, 2012; Reece, Hollub, Nangami, & Lane, 2010), less stable *Corresponding author. Email: [email protected] © 2015 Taylor & Francis
relationship status (Larsson et al., 2010; Morfaw et al., 2013) and traditional gender roles (Nkuoh et al., 2010; Reece et al., 2010; Theuring et al., 2009), including the belief that antenatal care is purely a women’s responsibility. Little is known about the role of MPI in the context of PMTCT in SA and other parts of the region. Although MPI has been the focus of much recent research, there is a lack of consensus about its definition (Ditekemena et al., 2012; Montgomery, van der Straten, & Torjesen, 2011), and the majority of research has focused on quantitative documentation of male attendance at antenatal clinics and male HIV testing. Given documented difficulties in attending antenatal appointments (Nkuoh et al., 2010; Reece et al., 2010), a broader view of MPI is needed. This article reports formative research that explored MPI among HIV-positive pregnant women using a quantitative survey and two focus group discussions (FGDs), with a view to examining the determinants of high levels of involvement and generating a broader understanding of women’s experiences of MPI, respectively.
AIDS Care Methods Participants The study was conducted at an antenatal service in Khayelitsha, SA, between July and November 2013. Pregnant women were eligible to participate if they were HIV-positive and had a male partner, where “partner” was defined as a primary sexual partner. Eligible participants were identified with the assistance of health care providers and support groups for HIV-positive pregnant women. All participants provided written informed consent prior to participation. The study was approved by the University of Cape Town’s Faculty of Health Sciences Human Research Ethics Committee and the local health authority.
Data collection A trained fieldworker conducted the survey and FGDs in isiXhosa using a paper-based questionnaire which was developed and piloted prior to data collection. Questions were asked regarding sociodemographic characteristics and self-reported disclosure of HIV status. Levels of HIV-related communication with partners were assessed based on reported discussion of HIV testing and preventing horizontal transmission. The number of topics discussed was summed to calculate a communication score, such that 0–1 indicated low and 2 indicated high levels of communication. Given the lack of a standardised measure (Montgomery et al., 2011), MPI was assessed using a range of activities based on previous published research (Byamugisha et al., 2010). In this study, these activities included partner involvement in: providing broad financial support for antenatal visits; knowing when visits were due; discussing what happened during visits; accompanying their partner to visits; discussing ways to prevent PMTCT; and providing support for ARV adherence. The number of activities that partners reportedly carried out was summed to calculate an MPI score, such that 0– 3 indicated low and 4–6 indicated high involvement. The Cronbach’s alpha coefficient for this scale was 0.66. Based on preliminary results of the survey, a FGD guide was developed to obtain more in-depth information on women’s experiences of MPI during pregnancy, as it was felt that group processes may help to elucidate more nuanced experiences. MPI was not defined for the purpose of the FGDs, thereby allowing participants to put forward their own views of what activities constitute MPI. The FGDs were digitally recorded.
Data analysis Variables significantly associated with MPI were identified using χ2 or Fisher exact tests for categorical variables and t tests for continuous variables. Odds ratios
1021
(OR) with 95% confidence intervals (CI) were calculated. Variables significantly associated with MPI in bivariate analysis (at p < 0.05) were included in a logistic regression model. Data were analysed using Stata 12 (StataCorp Inc, College Station, Texas, USA). The FGDs were transcribed and translated into English, and were back-translated for quality assurance. Thematic analysis was used to code the transcribed discussions according to the repeated themes which emerged.
Results Sociodemographic characteristics A total of 170 women participated in the survey, and 16 women participated in the FGDs. Detailed sociodemographic characteristics are presented in Table 1. Compared to their male partners, participants were younger and had lower levels of education and employment. Just over a quarter of participants were married (26%), and 52% of participants were cohabiting. Survey results Participants reported high levels of HIV-related communication with partners. Seventy-four per cent had reportedly disclosed their HIV status to their partner, but only 54% reported that they knew their partner’s HIV status. Most participants reported high levels of partner involvement in various pregnancy-related activities, and many (70%) reported that their partners provided support towards ARV adherence. However, only 35% reported that their partners accompanied them to the clinic for antenatal visits (Table 2). In a multiple logistic regression model (Table 3), high MPI was significantly associated with cohabitation (OR: 7.1; 95% CI: 1.8–27.7), disclosure (OR: 4.5; 95% CI: 1.3–15.6) and high levels of HIVrelated communication (OR: 6.6; 95% CI: 1.9–23.0). FGD results Although not all partners were described as supportive, there was unanimous agreement among FGD participants that pregnancy is a shared responsibility. Examples of ways in which partners could be supportive included helping at home, looking after children, enquiring about the pregnancy, providing money to go to the clinic and assisting their female partners with self-care activities when advanced pregnancy made this difficult. Partners could also provide support for ARV adherence by reminders to take ARVs; bringing their ARVs with a glass of water; and fetching ARVs from the clinic. The participants reported reluctance among male partners to attend antenatal visits, as they are prevented from entering the clinic by security guards. Despite this, women supported the idea of a male-friendly facility
1022
K. Brittain et al.
Table 1. Description of participant and partner characteristics.
Table 2. HIV-related communication, disclosure and MPI.
Variables
Variables
Survey participants (n = 170) Median age [Inter-quartile range (IQR)] 29 Race Black/African 170 Language English 1 isiXhosa 168 Sesotho 1 Education Primary school 5 Some secondary education 110 Matric/tertiary education 55 Employment Employed 75 Partners’ median age (IQR) 33 Partner’s education Primary school 5 Some secondary education 60 Matric/tertiary education 90 Unknown 15 Partner’s employment Employed 142 Marital status Married 44 Cohabitation status Cohabiting 89 Number of weeks pregnant at time of interview Second trimester 42 Third trimester 128 Gravidity First pregnancy 29 Decision making regarding pregnancy Unplanned pregnancy 115 FGD participants (n = 16) Median age (IQR) 29.5 Marital status Married 4 Cohabitation status Cohabiting 8 Number of weeks pregnant at time of interview Second trimester 3 Third trimester 13 Gravidity First pregnancy 4
n (%) (26–33) (100) (1) (99) (1) (3) (65) (32) (44) (29–37) (3) (35) (53) (9) (84) (26) (52) (25) (75) (17) (68) (26–35.5) (25) (50) (19) (81) (25)
and suggested that partners be encouraged to attend at least one visit, because they wanted them to be informed about pregnancy.
Discussion This research explored MPI during pregnancy as a potentially useful mechanism to facilitate PMTCT.
HIV-related communication Has discussed HIV testing with partner Has discussed preventing horizontal transmission with partner Disclosure to partner Participant has not disclosed Participant has disclosed Knowledge of partner’s HIV status Participant does not know partner’s status Participant knows partner’s status Items in MPI score (n = 169) Provides broad financial support for participant’s antenatal visits Knows when participant’s antenatal visits are Discusses with participant what happens during antenatal visits Reported partner discussion of PMTCT Provides support for participant’s ARV adherence Accompanies participant to the clinic for antenatal visits MPI score (n = 169)a 0 (no items with a positive response) 1 2 3 4 5 6 (all items with a positive response) Median MPI score (IQR)
n (%) 136 (80) 149 (88)
44 (26) 126 (74) 79 (46) 91 (54) 144 (85) 161 (95) 162 (96) 151 (89) 119 (70) 59 (35)
2 (1) 3 (2) 7 (4) 13 (8) 25 (15) 74 (44) 45 (27) 5 (4–6)
a
Calculated by summing MPI activities detailed above.
High levels of HIV-related communication and disclosure were reported, and participants reported MPI in a range of activities. Higher levels of MPI were significantly more likely among participants who were cohabiting; who had disclosed their HIV status; and who reported higher levels of HIV-related communication with their partner. FGD participants discussed a range of pregnancy-related activities, not limited to male attendance at antenatal visits, as being indicative of MPI. Reported disclosure to male partners in this study was high compared to other South African and African settings (Crankshaw et al., 2014; Makin et al., 2008; Roxby et al., 2013). The high reported levels of HIVrelated communication are encouraging and suggest the normalisation of HIV (Zuch & Lurie, 2012). Consistent with previous published research, we observed higher levels of MPI among participants who reported higher levels of communication with partners (Ditekemena et al., 2012; Kalembo et al., 2012; Reece et al., 2010) and stable relationship status (Ditekemena et al., 2012; Larsson et al., 2010; Morfaw et al., 2013). The high reported levels of MPI overall indicate that gender norms
AIDS Care
1023
Table 3. Factors associated with MPI. Unadjusted association Variables Gravidity First pregnancy Has previously been pregnant Marital status Not married Married Cohabitation status Not cohabiting Cohabiting Knowledge of partner’s HIV status Does not know partner’s status Knows partner’s status Disclosure of HIV status Has not disclosed to partner Has disclosed to partner HIV-related communication Low High
Adjusted association
OR
95% CI
P value
Reference 3.7
(1.2–10.4)
0.009
Reference 4.7
(1.1–43.0)
0.026
Reference 10.9
(3.0–58.6)
