Changes in Providers’ Self-Efficacy and Intentions to Provide Safer Conception Counseling Over 24 Months

Abstract

High rates of fertility desires, childbearing and serodiscordant partnerships among people living with HIV (PLHIV) in Uganda underscore the need to promote use of safer conception methods (SCM). Effective SCM exist but few PLHIV benefit from provider led safer conception counseling (SCC) and comprehensive national SCC guidelines are still lacking. Providers’ self-efficacy, intentions and attitudes for SCC impact provision and should inform development of services, but there are no longitudinal studies that assess these important constructs. This study reports on changes in providers’ knowledge, attitudes, motivation and confidence to provide SCC among a 24-month observational cohort of Ugandan HIV providers. Compared to baseline, providers evidenced increased awareness of SCM, perceived greater value in providing SCC, saw all SCM but sperm washing as likely to be acceptable to clients, reported consistently high interest in and peer support for providing SCC, and perceived fewer barriers at the 24-month follow-up. Providers’ intentions for providing SCC stayed consistently high for all SCM except manual self-insemination which decreased at 24-months. Self-efficacy for providing SCC increased from baseline with the greatest improvement in providers’ confidence in advising serodiscordant couples where the man is HIV-infected. Providers consistently cite the lack of established guidelines, training, and their own reluctance to broach the issue with clients as significant barriers to providing SCC. Despite providers being more interested and open to providing SCC than ever, integration of SCC into standard HIV services has not happened. Concerted efforts are needed to address remaining barriers by establishing national SCC guidelines and implementing quality provider training.

INTRODUCTION

The growing global recognition of serodisordant couples’ childbearing desires and rights has led to calls for comprehensive reproductive health services that meet clients’ needs while limiting risk of horizontal and vertical transmission (World Health Orgaanization 2017). In Uganda, about 60% of people living with HIV (PLHIV) wish to conceive [2,3] and up to 50% of PLHIV in relationships have an uninfected partner [4,5]. Up to 40% of female PLHIV become pregnant post-HIV diagnosis with about half of these pregnancies being planned or desired by at least one member of the couple [6]. Nevertheless, few PLHIV benefit from provider input on strategies to reduce risk associated with conception [6,7,8]. These high observed rates of fertility desires and childbearing underscore the need to promote safer conception methods (SCM) for serodiscordant couples, but effective interventions are lacking. A thorough understanding of providers’ knowledge, attitudes and self-efficacy for providing safer conception counseling (SCC) is key to the development of effective services. In the context of a 24-month prospective observational cohort study of patients in Uganda [9], we followed a cohort of providers to examine their knowledge, attitudes and self-efficacy for providing SCC at baseline [10] and changes in these important variables over time. Like the patients in our cohort, we anticipated that providers’ knowledge, attitudes and self-efficacy might change over time as the general awareness of SCMs and recognition of PLHIV’s childbearing desires and rights became more widespread. This is the first longitudinal study to examine changes over time among these influential stakeholders.

Effective and low-cost SCM exist but are rarely practiced by serodiscordant couples [11,12,13]. When the female is the infected partner, manual self-insemination (MSI) eliminates all risk of transmission to her uninfected male partner [14]. All serodiscordant couples can reduce their risk by limiting unprotected sex to the three-day window when the female partner is most fertile, a method known as timed unprotected intercourse (TUI). Though the absolute risk associated with TUI is still unknown, it is thought to be extremely low when combined with adequate adherence to antiretroviral therapy (ART) by the HIV-infected partner [14,15]. Use of PrEP by the uninfected partner further reduces risk associated with TUI. Sperm-washing nearly eliminates all risk of transmission from an HIV-infected male to his uninfected female partner, but it is typically too costly for couples in low-resource settings [14,16,17]. In our observational cohort study described above, almost half of serodiscordant couples reported fertility desires, but despite the negligible costs of MSI and TUI, only 2 reported using MSI and 15% reported using TUI [9].

The infrequent use of SCM among serodiscordant couples is unsurprising, given that services to guide couples in the use of SCM are rarely available in standard HIV care. Reproductive health and family planning services are commonly integrated in HIV clinics in Uganda and elsewhere in sub-Saharan Africa however, they focus almost exclusively on preventing unplanned pregnancies and mother-to-child-transmission (PMTCT) once pregnancy occurs. Fertility desires are not commonly discussed by clients and providers and when they are, the conversations are usually initiated by the patient and rarely involve discussion of SCM [10,18,19]. Providers may be unaware of SCM, lack the necessary skills to provide SCC and/or have reservations about raising the topic with clients [10,20,21,22,23]. In fact, providers have historically discouraged, and thereby stigmatized, childbearing among HIV-affected couples [24,25], making supportive conversations uncommon. There is growing evidence that provider stigma toward childbearing has decreased in recent years [10,20,21,22], however longitudinal studies of changes in providers’ knowledge, attitudes, motivation and confidence to provide SCC are still lacking.

Theoretically grounded studies improve the selection of variables to be assessed, measures to be used, interpretation of results and replication of study findings [26]. As such, we employed an ecological adaptation of the Information Motivation and Behavioral skills (eIMB) model of behavior change in this study [27]. Our eIMB informed model posits that the likelihood of providers offering SCC is influenced by the information, motivation, and behavioral skills they possess, as well as contextual factors in their environment. Information is necessary but not sufficient to ensure provision of SCC and includes awareness of SCM in general, technical aspects of each method, impact on risk, risk of current state of unassisted pregnancies and a thorough understanding of clients’ reasons for wanting a child. Motivation is impacted by interest in providing SCC to the full range of potential clients (i.e., concordant couples, discordant couples with HIV-infected male, discordant with HIV-infected female, clients without committed partners) and the presence of significant perceived barriers. Behavioral skills include self-efficacy for providing SCC as well as technical expertise to identify and explain specific aspects of SCM. Contextual factors in the sociocultural environment, the level of stigma toward childbearing that exists among healthcare providers, perceived acceptability of SCM among clients and peer support for providing SCC directly impact the availability of information and providers’ motivation. We did not attempt to assess the provision of SCC as widespread knowledge of SCMs, national guidelines, provider trainings and patient education tools were not available during the study.

This study reports longitudinal findings from a 24-month observational cohort study of Ugandan HIV providers. We assessed changes in knowledge, attitudes, motivation and confidence to provide SCC. As this was an observational cohort study, no intervention was provided. Findings will be used to inform the design of effective SCC services and ultimately to assist in the development of Ugandan guidelines for the provision of these services in routine HIV care.

METHOD

Study Setting

This study was conducted in collaboration with The AIDS Support Organization (TASO) sites in Kampala and Jinja, Uganda. TASO is one of the largest indigenous non-governmental organizations in Uganda providing comprehensive HIV prevention, care, and support services for HIV infected and affected Ugandans annually. TASO Kampala serves over 6700 PLHIV and TASO Jinja cares for over 8000 patients. TASO provides family planning and contraception services, but during the period of this study, no services specific to safer conception.

Participants

At baseline (May-October 2013), all medical/clinical officers and a convenience sample of nurses and counselors at the two sites were approached by the study coordinator and offered participation in the study. The time and day of the week in which potential participants were approached were varied to increase the likelihood of a diverse sample. All providers gave verbal informed consent (there were no refusals), at which time we clarified that their individual responses would not be shared with TASO. Questionnaires were administered via an interview conducted by study personnel in English or Lugandan. Only three participants chose to respond in Lugandan. Follow-up surveys were conducted at 12 and 24 months and providers received 20,000 Ush (~$6 USD) for completing each survey. The study protocol was reviewed and approved by Institutional Review Boards at Makerere University School of Biomedical Sciences and RAND Corporation, as well as the Uganda National Council for Science and Technology.

Measures

Guided by the eIMB model and drawing on our own qualitative research [21,28] and the literature, we adapted established scales and constructed original items to assess the domains described below. Most domains are reported as single items or total scores for inventories (i.e., Awareness of SCM, Barriers to Providing SCC). The internal consistency and preliminary validity of several of the adapted and original scales (i.e., Provider Stigma of Childbearing among PLHIV, Perceived Value of Providing SCC, Self-Efficacy for Provided SCC, and three Interest in Providing SCC scales) were examined and reported in detail elsewhere [29]. In short, content validity was established via expert review and face validity was explored during cognitive debriefing with volunteers who met study eligibility criteria. Construct validity was assessed via factor analysis using ordinary least squares estimation. We considered scree plots and varimax-rotated matrix of factor loadings to assign items. Internal consistency was established with Cronbach’s Alpha. Exact wording of items and response categories, as well as means (SD; range) for inventories, scales, and other items are presented in Table 1.

Table 1.

Providers’ responses to the selected survey items and scales (completers only).

	Baseline (N=35)		12 months (N=33)		24 months (N=35)		Wilcoxon p value
Survey Items	M (SD; range)	% Yes or Agree a	M (SD; range)	% Yes or Agree a	M (SD; range)	% Yes or Agree a	(0–24 mo change)
Provider Stigma of Childbearing Among PLHIV scale (1= strongly disagree to 4 = strongly agree; ∝ = .63)	2.1 (0.5; 1.4–3.2)		1.8 (0.4; 1.2–2.6)		2.1 (0.4; 1.0–2.8)		.964
Children born to an HIV+ parent face more challenges than are necessary.	2.8 (0.8; 1.0–4.0)	66%	2.4 (0.8; 1.0–4.0)	48%	2.5 (0.7; 1.0–4.0)	54%
HIV+ people often lack all that they need to bring a child into the world.	2.2 (1.0; 1.0–4.0)	31%	1.8 (0.7; 1.0–3.0)	18%	2.0 (0.8; 1.0–3.0)	31%
HIV+ people who want to have children are being selfish.	1.8 (0.9; 1.0–4.0)	23%	1.5 (0.7; 1.0–3.0)	9%	2.0 (0.9; 1.0–3.0)	37%
Helping HIV+ people have children is a distraction from more important issues that we need to address as providers.	1.2 (0.4; 1.0–2.0)	0%	1.2 (0.4; 1.0–2.0)	0%	1.3 (0.4; 1.0–2.0)	0%
Ensuring patients are always having safe protected sex is more important than helping HIV+ people to have children.	2.6 (0.8; 1.0–4.0)	54%	2.2 (0.8; 1.0–4.0)	36%	2.9 (0.7; 1.0–4.0)	74%
Awareness of SCM (0 = No, 1 = Yes, 3 = Not Sure)	4.3 (1.3; 2.0–7.0)		6.7 (1.3; 3.0–9.0)		7.3 (1.9; 0.0–9.0)		<.001
Are you aware of methods to increase the safety of conception in mixed status couples...    ...(sero-discordant) couples where one partner is HIV+ and the other is HIV-negative?		91%		97%		94%
...by having them engage in unprotected or live sex only during the few days of the month when the woman is most fertile?		69%		100%		100%
...whereby the man ejaculates into a container or condom and then the semen is injected into the woman’s vagina?		49%		67%		72%
To the best of your knowledge, have guidelines from any organization been established yet to guide providers in addressing the comprehensive reproductive needs of HIV+ individuals and couples who want to have children?		23%		27%		15%
Are you aware of technology that removes HIV from the man’s semen and thus increase the safety of conception in couples where man is HIV+ and woman negative?		63%		64%		88%
Do you know where to refer a client or couple who want to use any of the methods described above to make conception more safe?		43%		55%		79%
HIV medication that can be taken by a HIV-negative partner who wants to conceive with a HIV+ partner to reduce his/her risk of infection?		94%		88%		97%
Perceived Value of Providing SCC scale (1= strongly disagree to 4 = strongly agree; ∝ = .74)	1.9 (0.7; 1.0–3.7)		1.6 (0.5; 1.0–3.0)		1.5 (0.5; 1.0–3.0)		<.001
Providing guidance on safer conception...    ...to a female client is a waste of time as they won’t be able to get their man to agree to modify their sexual practices.	1.5 (0.8; 1.0–3.0)	17%	1.2 (0.6; 1.0–3.0)	9%	1.1 (0.3; 1.0–2.0)	0%
...to a female client is a waste of time as their man will demand live sex.	1.7 (0.9; 1.0–4.0)	20%	1.4 (0.6; 1.0–3.0)	6%	1.4 (0.7; 1.0–4.0)	9%
Clients who are counseled to have unprotected or “live” sex during a few days a month when the woman is most fertile will not want to resume using condoms afterward.	2.5 (0.9; 1.0–4.0)	57%	2.2 (0.8; 1.0–4.0)	39%	2.1 (0.8; 1.0–4.0)	24%
Perceived Acceptability of SCM to Clients (1= strongly disagree to 4 = strongly agree)
Will clients be okay with being asked to make a conception plan with a health care provider?		91%		94%		97%
Will couples be willing to collect the man’s semen [perhaps by having sex with a condom] and inject it into the woman’s vagina?		66%		78%		71%
Will couples be willing to have unprotected or “live” sex only during the few days a month when the woman is most fertile?		91%		82%		94%
Will couples be willing to have the man’s sperm washed to remove HIV with the use of technology, and then inserted into the woman’s vagina if cost was not a factor?		82%		79%		71%
Will HIV+ partners would be willing to start HIV medication early if they knew it would reduce their risk of transmitting the virus to a partner.		97%		94%		100%
Will HIV negative partners of HIV+ patients would be willing to take HIV medication every day during the months in which they were trying to conceive in order to reduce their risk of infection.		83%		85%		94%
Barriers to Providing SCC (1= Not a barrier at all, 2 = somewhat a barrier, 3 = definitely a barrier)	2.2 (0.5; 1.2–2.8)		2.3 (0.4; 1.3–2.8)		2.1 (0.4; 1.2–2.7)		.006
How much of a barrier is...   ...poor access to male members of couples who want to have a child.		94%		97%		82%
...lack of HIV disclosure within couples who want to have a child.		94%		100%		76%
...no established guidelines or recommendations for how to provide such counseling.		86%		85%		91%
...not having any educational tools to use in counseling clients.		86%		85%		79%
...poor access to ARVs that can be taken by uninfected partners during periods of unprotected sex when trying to conceive?		86%		94%		85%
...lack of training for how to provide such counseling.		80%		91%		88%
...client reluctance to discuss childbearing needs.		83%		85%		85%
...lack of resources and support from the clinic administration for such counseling.		66%		64%		35%
...poor access to ART for patients who want to have a child but their CD4 is not low enough to quality for ART.		74%		61%		71%
...not having enough time to talk further with clients.		66%		85%		71%
...my personal reluctance to discus with client their desires to have children.		66%		76%		76%
Peer Support for Providing SCC (1= strongly disagree to 4 = strongly agree)
People I know and respect think I should...   ...talk to HIV patients about their desires to have children?		94%		97%		97%
...discuss the availability and use of methods to increase the safety of conception with HIV clients who have a desire to have children.		94%		100%		100%
Interest in Providing SCC scales b (1= low interest to 10= high interest or 1= strongly disagree to 4 = strongly agree; ∝ = .79)	∝=.81		∝=.66		∝=.64
Interest in Providing SCC to Serodiscordant Couples scale (∝ = .91)	8.7 (1.8; 3.0–10.0)		8.6 (1.6; 5.0–10.0)		8.2 (1.1; 5.5–10.0)		.059
How interested are you in providing guidance...    ...on how to conceive safely to a couple where the woman is HIV+ and the man is not?	8.7 (1.9; 3.0–10.0)		8.6 (1.6; 5.0–10.0)		8.1 (1.3; 5.0–10.0)
...on how to conceive safely to a couple where the man is HIV+ and the woman is not?	8.7 (1.9; 3.0–10.0)		8.7 (1.7; 5.0–10.0)		8.3 (1.1; 6.0–10.0)
Interest in Providing SCC Regarding Specific SCM scale c (∝ =.72; ∝ for first three items = .61)	7.6 (1.7; 3.5–9.5)		7.9 (1.3; 4.7–10.0)		7.9 (1.0; 5.3–9.5)		.920
How interested are you in providing guidance to mixed status couples...    ...about the use of unprotected or “live” sex only during the few days a month when the woman is most fertile?	7.0 (3.0; 1.0–10.0)		7.3 (2.0; 2.0–10.0)		7.4 (1.7; 3.0–10.0)
...(where the woman is HIV+) about how to collect the man’s semen and inject it into the woman’s vagina?	7.6 (3.0; 1.0–10.0)		8.3 (2.3; 3.0–10.0)		6.8 (2.0; 1.0–10.0)
If ARVs were approved for such use in Uganda, how interested would you be in providing guidance to uninfected partners of your HIV+ patients about taking ARVs daily during the months they attempt conception via unprotected sex?	9.2 (1.9; 1.0–10.0)		9.7 (0.7; 8.0–10.0)		8.3 (1.6; 5.0–10.0)
Most clients will not follow the advice we give regarding how to increase the safety of conception.	1.9 (0.8; 1.0–3.0)	26%	2.0 (0.8; 1.0–4.0)	21%	1.5 (0.7; 1.0–4.0)	3%
Most uninfected partners will not take HIV medications daily during the conception period.	2.1 (0.9; 1.0–4.0)	29%	2.2 (0.8; 1.0–4.0)	27%	1.9 (0.7; 1.0–4.0)	15%
It is not a good use of resources to recommend that uninfected partners take HIV medications daily during the conception period.	1.6 (0.8; 1.0–4.0)	17%	1.5 (0.8; 1.0–4.0)	15%	1.4 (0.8; 1.0–4.0)	15%
Interest in Providing SCC in the Context of Relational Factors scale (∝ = .83)	8.0 (2.4; 1.8–10.0)		8.3 (1.8; 3.0–10.0)		8.1 (1.6; 1.0–10.0)		.774
How interested are you in providing guidance...    ...to an HIV-infected woman who wants to conceive, but does not have a committed partner?	7.9 (3.0; 1.0–10.0)		8.1 (2.4; 2.0–10.0)		8.0 (1.8; 1.0–10.0)
...to an HIV-infected man who wants to conceive, but does not have a committed partner?	7.7 (2.9; 1.0–10.0)		8.1 (2.3; 2.0–10.0)		8.0 (1.8; 1.0–10.0)
...about HIV disclosure to HIV-infected client who wants a child with an HIV-negative partner, to whom they have not disclosed their HIV status?	8.7 (2.6; 1.0–10.0)		8.9 (1.5; 5.0–10.0)		8.5 (1.8; 1.0–10.0)
...to HIV-affected couples who want to conceive if they already have children?	8.0 (2.9; 1.0–10.0)		8.1 (2.7; 2.0–10.0)		8.0 (2.4; 1.0–10.0)
Self-efficacy for Providing SCC scale (1= not at all to 10= extremely; ∝ = .87)	7.4 (1.6; 4.0–9.9)		7.6 (1.4; 5.0–9.9)		8.0 (0.8; 6.9–10.0)		.022
How confident do you feel in your ability to    ...ask clients about their future childbearing goals?	8.1 (2.0; 4.0–10.0)		7.8 (1.7; 5.0–10.0)		8.1 (1.2; 5.0–10.0)
...provide safer conception guidance to a couple in which the woman is HIV-infected and the man is not?	6.9 (2.5; 3.0–10.0)		6.9 (2.1; 3.0–10.0)		7.6 (1.4; 4.0–10.0)
...provide safer conception guidance to a couple in which the man is HIV-infected and the woman is not?	6.7 (2.3; 3.0–10.0)		7.1 (2.1; 2.0–10.0)		7.8 (1.2; 5.0–10.0)
...provide guidance to an HIV-infected woman who wants to conceive, but does not have a committed partner?	6.8 (2.4; 1.0–10.0)		7.5 (2.0; 2.0–10.0)		7.8 (1.3; 5.0–10.0)
...provide guidance to an HIV-infected man who wants to have a child, but does not have a committed partner?	6.8 (2.3; 1.0–10.0)		7.4 (2.1; 1.0–10.0)		7.9 (1.4; 5.0–10.0)
...provide guidance about disclosure to HIV+ client who wants a child with HIV-negative partner, to whom they have not disclosed?	7.3 (2.2; 1.0–10.0)		7.2 (2.4; 1.0–10.0)		8.2 (1.2; 6.0–10.0)
If ART initiation was not restricted by CD4 count, how confident are you that you could provide guidance for early initiation of ART among HIV+ patients with uninfected partners who want to conceive?	8.3 (2.0; 2.0–10.0)		9.1 (1.2; 6.0–10.0)		8.3 (1.3; 6.0–10.0)
If pre-exposure prophylaxis was readily available in Uganda, how confident are you that you could provide guidance to uninfected partners of your HIV+ patients on taking ARVs daily during the months they attempted conception via unprotected sex?	8.2 (1.8; 5.0–10.0)		8.2 (1.7; 5.0–10.0)		8.2 (1.4; 4.0–10.0)
Intentions to Provide SCC (1= low intention to 10= high intention)
How much do you intend to discuss/talk with...    ...male clients any desires or plans they may have regarding having children?	8.3 (1.6; 5.0–10.0)		8.1 (1.5; 5.0–10.0)		8.4 (1.4; 5.0–10.0)
...female clients any desires or plans they may have regarding having children?	8.8 (1.4; 5.0–10.0)		8.8 (1.2; 6.0–10.0)		8.6 (1.1; 7.0–10.0)
...patients who have a desire to have children, the availability and use of methods to increase the safety of conception?	9.5 (1.2; 4.0–10.0)		9.3 (1.2; 5.0–10.0)		9.1 (1.0; 7.0–10.0)
...mixed status couples who want to have a child about the use of timed unprotected intercourse- i.e., having “live” sex only during the few days a month when the woman is most fertile?	7.0 (2.7; 1.0–10.0)		7.2 (2.0; 2.0–10.0)		7.4 (1.5; 4.0–10.0)
...mixed status [woman is HIV+] who want to have a child about how to collect the man’s semen and inject it into the woman’s vagina?	7.7 (3.2; 1.0–10.0)		7.9 (2.9; 1.0–10.0)		6.2 (1.8; 2.0–10.0)

^aReported percent is the combined percent of “Agree/Strongly Agree” or “somewhat/definitely” responses.

^bInterest in Providing SCC Regarding Specific SCM scale scores were computed by converting the three reverse-coded 4-point Likert items to a 10-point scale before averaging across the six items.

^cThe Interest in providing SCC for specific SCM scale included both 4- and 10-point Likert-type items; we converted responses on the 4-point items to a 10-point scale (1=1, 2=4, 3=7, 4=10) before averaging across the six items.

Provider demographics and practice characteristics.

We recorded age, sex, current position, years in practice and years worked with HIV clients.

Frequency of childbearing discussions.

We developed six items that asked providers to report on whether they had ever discussed childbearing plans with a patient (yes/no), as well as what proportion of reproductive aged female and male clients they had discussed childbearing plans with in the last 30 days, and what proportion of those consultations were initiated by the female or male patient or by the provider themselves. Providers’ report of the proportion of female and male clients with whom childbearing plans had been discussed in the past month was averaged to produce a total proportion for use in analyses.

Provider Stigma of Childbearing among PLHIV Scale and Attitudes.

We constructed five items to gage providers’ views about PLHIV having children. Positively worded items were reversed scored and a mean item score was computed with higher scores representing more negative attitudes. In addition, we asked providers four general questions about childbearing among PLHIV and to list their top three concerns about PLHIV having children.

Awareness of SCM Inventory.

We developed seven items to assess providers’ awareness of SCM. The sum of affirmative responses represented level of awareness of SCM. In addition, we asked providers to rate whether they had adequate information to provide SCC, if they needed training, and whether they wanted training, using single items and a “yes/no/not sure” response format. “Not sure” responses were coded as “no” for analyses.

Perceived Value of Providing SCC.

We developed six items to assess providers’ views of the value of providing SCC. After reverse scoring all items, a mean item score was computed with higher scores representing greater perceived value.

Perceived Acceptability of SCM to Clients.

We adapted seven items from the WHO assessment of contraceptive method preferences [30] to assess providers’ perceptions of whether clients will view specific SCM as acceptable. Five of the seven items were used descriptively, and the final two on TUI and MSI were used as individual variables in analyses. We explored the development of a scale with all seven items, but likely due to the variety of topics covered, the psychometrics were poor.

Barriers to Providing SCC Inventory.

We developed 12 items to assess barriers to providing SCC. All items were reverse scored and a mean score across items was calculated, with higher scores representing a perception of the barriers being greater. Individual item scores were used to compute the percentage of providers reporting some barriers (vs. no barrier at all).

Peer Support for Providing SCC.

We used two items to assess providers’ views about the receipt of peer support. A mean item score was computed with higher scores representing greater perceived peer support.

Interest in Providing SCC Scales.

We constructed 12 items that formed three scales; Interest in providing SCC to serodiscordant couples, Interest in providing SCC regarding specific SCM, and Interest in providing SCC in the context of relational factors. A mean item score for each scale was computed with higher scores representing greater interest.

Self-Efficacy for Providing SCC Scale.

We adapted a self-efficacy measure developed by Johnson et al. [31] to create eight items to assess providers’ level of confidence to discuss childbearing and provide SCC to different types of couples. A mean item score was computed with higher scores representing greater confidence.

Intentions to Provide SCC.

We used five items to assess providers’ intention to provide specific aspects of SCC. Three items were used descriptively and two on intention to provide counseling on TUI and MSI were used in the analyses. Here again, we explored the development of a five-item scale, but likely due to the variety of SCM covered, the psychometrics were poor and thus we opted to use individual items in the analyses.

Data Analysis

Descriptive statistics (frequencies, means, standard deviations, ranges) were used to describe sample characteristics and findings. Spearman correlations were used to examine associations between 24-month follow-up predictors and providers’ ratings of self-efficacy.

Ethical Approval

The study protocol was reviewed and approved by Institutional Review Boards at Makerere University School of Biomedical Sciences and RAND Corporation, as well as the Uganda National Council for Science and Technology.

RESULTS

The baseline sample consisted of 57 providers (29 from Kampala and 28 from Jinja), including 10 medical/clinical officers (6 female), 13 nurses (10 female), and 34 counselors (17 female). A total of 46 of these providers contributed data to the one-year follow-up. At 24 months, responses were available from 35 providers (10 from Kampala and 25 from Jinja), of whom 5 were medical/clinical officers (3 female), 6 were nurses (5 female), and 23 were counselors (11 female). In the second year of the study, TASO experienced a significant funding cut that resulted in a 40% reduction in staff positions across all sites. A complete description of the entire baseline sample is provided in an earlier publication [10]. Analyses for this study were restricted to the 35 providers who contributed data at baseline and 24-months. At baseline, providers in this study were on average 35 years of age (SD=5.4, range 24–50 years), with just over half (57%) being female and averaging 7.5 years of experience working with clients living with HIV (SD=3.9, range 1–25). A comparison of providers who dropped out to those who completed the 24-month evaluation revealed that dropouts were demographically similar to completers. Dropouts evidenced similar responses to almost all variables, but were slightly more likely to be interested in providing SCC to serodiscordant couples and had greater intentions to counsel on TUI at baseline than those who completed the 24-month assessment. Changes over time for most of the questionnaire items and all items that formed scales or inventories are displayed in Table 1. Results for additional items are presented in the text below.

Frequency of childbearing discussions with clients.

At the 24-month assessment, all providers (100%) responded “Yes” to the question “Have you ever discussed childbearing plans with an HIV-infected client?” as compared to a similarly high 97% at baseline. Nevertheless, on average providers had discussed childbearing with only 37% of reproductive aged clients in the past month, a statistically non-significant increase from the reported 28% at baseline. Discussions regarding childbearing occurred with 49% of female and 25% of male clients, up from 36% and 20% respectively at baseline (no significant differences). When these conversations did occur, female clients initiated 60%, a non-significant increase from the 52% observed at baseline. Conversations initiated by male clients significantly increased from 21% at baseline up to 40% at the 24-month follow-up (p=0.02). Providers reported that they initiated these discussions only 33% of the time, down from 36% observed at baseline. Providers who initiated childbearing discussions reported higher awareness of SCM (r_s = .49, p = .021), interest in providing SCC (r_s=0.423, p=0.014), self-efficacy (r_s=0.66, p <.001), and fewer barriers (r_s= −0.552, p<0.001).

Provider Stigma of Childbearing Among PLHIV scale.

As displayed in Table 1, average stigma scores did not change significantly over time (p=0.964) and hovered around the midpoint of this four point scale. Most items stayed about the same or got more positive, but providers endorsed more negative attitudes on two items focusing on PLHIV who want children as being selfish and ensuring that all clients are having protected sex as being more important than helping couples conceive.

Awareness of Safer Conception Methods.

Overall awareness significantly improved over time (p<0.001) with 100% reporting awareness of the steps required for TUI (up from 69% at baseline) and 72% displaying an understanding of MSI (baseline=49%). At 24 months, 41% of completers reported inadequate information to counsel their clients (baseline=45%), 91% reported need for more training on SC options (baseline=100%), and 91% reported they would like to receive this type of training (baseline=100%). Providers reporting awareness of comprehensive reproductive guidelines (which do not exist in Uganda) dropped to 15% at 24 months, down from 23% at baseline.

Perceived Value of Providing SCC scale.

Providers saw increasing value over the 24-month study period (p<0.001) with less than a quarter (24%) worrying that clients would struggle to resume using condoms after unprotected intercourse during the fertile periods as compared to 57% who voiced this concern at baseline.

Perceived Acceptability of SCM to Clients.

The proportion of providers agreeing that SCM would be acceptable to clients mostly increased over time. Except for sperm washing (71% vs. 82%), this proportion for all SCM was higher at 24 months compared to baseline.

Barriers to Providing SCC.

Providers reported a slight yet statistically significant reduction (p<0.006) in barriers to providing SCC at 24 months. Nevertheless, as compared to baseline, a greater percentage of providers viewed the lack of established SCC guidelines, lack of training, client reluctance to discuss childbearing needs, not having enough time and their own personal reluctance to broach the subject with clients as greater barriers at the 24-month follow-up.

Peer Support for Providing SCC.

Providers’ reports of peer support stayed consistently high over time (94% or greater endorsing strong peer support).

Interest in Providing SCC scales.

Similarly, providers’ interest on all subscales stayed consistently high.

Self-efficacy for Proving SCC scale.

Providers’ self-efficacy significantly increased over time (p=0.022) with the greatest improvements observed in providers’ confidence in advising serodiscordant couples where the man is infected, when disclosure had not happened, or with clients who did not have a committed partner.

Intentions to Provide SCC.

Providers’ intentions to talk to clients about childbearing desires, share the availability of SCM, and counsel on TUI and MSI were all high at baseline (all averaged 7 or better on 10 point scale). Changes over time were generally small and mixed. Providers’ intentions to talk to male and female clients about childbearing desires, discuss the availability of SCM and counsel on TUI fluctuated slightly. However, providers’ intentions to counsel on MSI decreased over the 24-month follow-up period.

Correlates of Self-efficacy for the Provision of SCC at 24 months

At the 24-month follow-up, providers who saw fewer barriers (p=0.002), had greater interest in providing SCC to serodisordant couples (p=0.002), had communicated with a greater proportion of clients about childbearing in the last 30 days (p=0.004), reported greater peer support (p=0.01), and had fewer years of experience as a provider (p=0.045) reported greater self-efficacy to provide SCC (Table 2).

Table 2.

Spearman correlations at 24 months (completers only, N = 35).

	Self-efficacy to provide SCC
	rs (p value)
Sex (female)	−0.009 (.959)
Age	−0.307 (.077)
Years worked as provider	−0.347 (.045)
Years worked with HIV clients	−0.157 (.375)
Proportion of patients communicated with about childbearing in last 30 days	0.485 (.004)
Provider stigma of childbearing scale	−0.178 (.314)
Awareness of SCM	0.151 (.394)
Perceived value of providing SCC scale	0.299 (.086)
Perceived acceptability of TUI	−0.03 (.87)
Perceived acceptability of MSI	0.05 (.78)
Barriers to providing SCC	−0.504 (.002)
Peer support for providing SCC	0.435 (.01)
Interest in Providing SCC...  to Serodiscordant Couples scale	0.518 (.002)
regarding Specific SCM scale	−0.011 (.95)
in Context of Relational Factors scale	−0.038 (.831)
Intentions to counsel on TUI	0.277 (.112)
Intentions to counsel on MSI	−0.039 (.828)

DISCUSSION

Providers reported having childbearing discussions with about the same proportion of clients at 24 months as they reported at baseline. However, a greater percentage of these conversations were initiated by male clients at 24-month follow-up as compared to baseline. Consistent with other studies, this increase is likely linked to the general increased awareness among clients of the availability, effectiveness and acceptability of SCM [13,17]. It also highlights the importance of the issue to male clients and the role that they can play in safer conception. Consistent with baseline, providers reported initiating these conversations about a third of the time. This lack of change indicates that rates of provider-initiated support for childbearing discussions is not likely to increase without a dedicated effort to address providers concerns and training needs. Providers who initiate childbearing discussions reported greater awareness of SCM, interest in and self-efficacy for providing SCC and saw fewer barriers for the provision of SCC. Providing high quality training that promotes increases in these important constructs among providers who are not currently initiating childbearing conversations with their clients might greatly improve the rate of childbearing discussion occurring in clinical care.

In general, providers’ endorsement of negative stigmatizing attitudes towards clients who desire a child stayed the same or slightly improved. Nevertheless, increases in the proportion of providers who view clients with childbearing desires as selfish or see promoting safer sex as more important than safer conception counseling reveal that there are lingering negative attitudes that may negatively impact care provision. Consistent with the eIMB model, education on the availability of effective SCM alone will likely not impact these provider attitudes. Rather, effective training will need to employ strategies to guide providers in exploring their own underlying reasons for holding these negative attitudes. Increasing providers’ understanding of and empathy for clients while providing a more realistic understanding of the risks involved in assisting couples versus doing nothing will be necessary [28,32]. Exposing providers to findings from our pilot intervention studies, where we found that 40% of couples provided with unbiased SCC chose not to pursue conception and where we observed no seroconversions among discordant couples [21], might help providers who fear doing harm by providing SCC. Guiding providers to shift their focus from feeling responsible for potentially negative outcomes to their role in assisting clients in making their own informed decisions will be critical. Facilitating the sharing of success stories through provider and client testimonials will also likely help to further reduce stigma and reinforce the importance of providers’ role.

Compared to baseline, providers evidenced increased awareness of SCM, perceived greater value in providing SCC, saw all but sperm washing as likely to be acceptable to clients, reported consistently high peer support for and interest in providing SCC, and reported fewer barriers at the 24-month follow-up. As this is the first longitudinal study, these are the first data to demonstrate improvements in these important constructs over time. Changes are likely related to increasing general awareness of SCM and recognition of reproductive rights of PLHIV [22,33]. Importantly providers still see the lack of established guidelines, inadequate training, and their own reluctance to broach the issue with clients as significant barriers. Provision of high quality training that addresses providers’ own reservations about SCC is needed to improve reproductive services for all clients living with HIV. Collaboration with the Ministry of Health to address the lack of national guidelines is also needed to ensure the regular provision of these needed services. The Uganda National Strategic Plan and National Priority Action Plan for HIV/AIDS states that a key intervention to reduce HIV transmission is the integration of sexual and reproductive health into HIV care programs [34]. Nevertheless, the current versions provide only very minimal guidance on how to support couples with childbearing desires.

Providers’ intentions stayed consistently high especially given the lack of training and support for SCC. The only noticeable change was a reduction in their intention to counsel on MSI. This is congruent with their consistent report of concern that this method will not be acceptable to clients. Although the high cost of sperm washing puts it out of reach for almost all clients, MSI was seen as the least acceptable SCM to clients at all time points. As suggested by other research [35], the perceived acceptability of MSI is likely to improve following high quality education and training that includes reports of real couples’ successful use.

Self-efficacy for providing SCC increased from baseline with the greatest improvement in providers’ confidence in advising serodiscordant couples where the man is HIV-infected. This is likely because providers are more comfortable with TUI which is the only viable option for these couples. In contrast, a more modest increase was observed for serodiscordant couples where the woman is HIV-infected. This is likely because providers view this situation as more complicated because MSI is the best option but they feel less comfortable with it. Despite removing all risk for the HIV-infected male partner, MSI is still seen by providers as unlikely to be acceptable to clients and they therefore anticipate having to offer TUI, which introduces risk for the uninfected male partner. Not surprisingly, without training and support, providers are likely to be more ambivalent about offering SCC to serodiscordant couples where the woman is HIV-infected.

Correlates of greater self-efficacy at 24 months included greater interest in providing SCC to serodiscordant couples and availability of peer support, as well as perceiving fewer barriers, having engaged in a greater proportion of childbearing discussion in the last 30 days, and having less years of experience as a provider. These important correlates of provider self-efficacy are not likely to improve to levels where SCC can become a part of normal practice without a dedicated effort to develop and test a comprehensive reproductive health program that includes high quality training which addresses providers’ ambivalence about providing SCC to all of their clients living with HIV.

Limitations

This study it is not without its limitations including a relatively small sample of providers and reliance on self-report data. Providers were drawn from two different sites within the same non-governmental organization that has already embraced progressive policies regarding fertility rights of PLHIV, so our findings may not be generalizable to all HIV providers in Uganda. Nevertheless, providers in this study still reported stigmatizing attitudes, low rates of childbearing discussion, modest self-efficacy and the desire for more training that would likely be echoed if not amplified by non-TASO providers. Our reliance on newly developed measures for some constructs is also a limitation. However, these are the first quantitative measures of these important constructs and their development was informed by extensive qualitative research, their psychometrics and preliminary validity have been explored elsewhere [29] and they may facilitate further research. Changes over time in providers’ knowledge, attitudes and self-efficacy for providing SCC may have been impacted by their participation in the study and completion of the repeat assessments. Nevertheless, as highlighted in the Discussion, similar findings have been observed in other cross sectional studies [17,22,33]. We did not correct for multiple comparisons in the interpretation of significant findings, due to the exploratory nature of the investigation and this being among the first studies to quantitatively evaluate these provider constructs. Our interpretation is based on the broad pattern of results and while we offer discussion of the potential interrelatedness of some findings, we were careful to avoid inferring causation in this observational study.

Conclusion

This longitudinal study demonstrates that providers are more interested and open to providing SCC for their HIV-infected clients than ever. However, the changes observed have not led to wide spread availability of SCC nor its integration into standard HIV services. Providers are clear that the remaining barriers of lack of high quality training and MOH guidelines stand in the way of the integration of SCC into routine HIV reproductive health services.