Abstract
In anticipation of a future HIV vaccine, we mapped the different personal positions regarding HIV vaccination of people in Togo. In early 2014, 363 adults indicated their willingness to receive a future HIV vaccine under different conditions varying as a function of five factors: perceived susceptibility to HIV, vaccine effectiveness, perceived severity of AIDS, vaccine cost, and family’s influence. We found five qualitatively different positions: unconditional acceptance (49%), depends on cost/effectiveness ratio (20%), depends on cost (18%), total indecision (10%), and complete reluctance (3%). Accordingly, HIV vaccination strategies in Togo and elsewhere in Africa need to be multifaceted and tailored.
Introduction
The development of a preventive HIV vaccine offers the best long-term hope of managing the impact of the AIDS pandemic that is ravaging sub-Saharan Africa (International AIDS Vaccine Initiative, 2009), where 70 percent of the 2.7 million new HIV infections worldwide in 2012 and 70 percent of all AIDS deaths occurred (UNAIDS, 2013). Progress is being made in the development of an effective HIV vaccine (Rappuoli and Aderem, 2011; Ross et al., 2010); however, findings from studies about the willingness of African people to participate in HIV vaccine trials (Jaspan et al., 2006; Smit et al., 2006) and about the acceptability of a potential HIV vaccine (Bishai et al., 2004; Sayles et al., 2010) strongly suggest that the advent of an HIV vaccine would not guarantee its uptake. It is important, therefore, to begin planning how to promote its widespread uptake.
A recent systematic review suggests that people vary considerably in their level of willingness to be vaccinated against HIV, ranging from 37.2 to 94 on a 100-point scale (Newman and Logie, 2010). In addition, the review has identified five broad factors impacting acceptability of the vaccine, as suggested by health-protective behavior theories (Weinstein, 1993): perceived susceptibility to HIV infection, effectiveness of the vaccine, perceived severity of AIDS, cost of the vaccine, and perceived social approval of vaccination. Nevertheless, very little is known about the relative contributions of these factors on African people’s willingness to receive the vaccine and about the possibly diverse vaccination positions of these people. Addressing this gap in knowledge is important in order to be able to tailor promotion of HIV vaccination to different people’s views and needs (Butler et al., 2015).
Methods
The study was conducted from January to April 2014. Ethics approval for the study was obtained from the Institutional Review Board of the University of Québec-Teluq.
Study setting
The study was conducted in Togo, a sub-Saharan African country with a population of 7 million (World Health Organization (WHO), 2015) that faces a generalized HIV epidemic (Conseil National de Lutte Contre le SIDA et les Infections Sexuellement Transmissibles—Togo (CNLS-IST), 2015). In Togo, the HIV prevalence is 2.5 percent, with an estimated incidence of 6649 in 2011; AIDS remains the most common cause of death, accounting for 17 percent of all deaths (CNLS-IST, 2015). The study site was Lomé, the capital city, with a population of 1.5 million (United Nations—Statistics Division, 2015), where the prevalence of HIV is the highest in the country (3.5%) (CNLS-IST, 2015). In Lomé, HIV is predominantly transmitted through heterosexual contacts; women are twice as likely to be infected as men, and sex workers are the most seriously affected (UNAIDS, 2009). People aged between 30 and 39 years face the most severe burden of HIV (CNLS-IST, 2015).
Study design
The study was designed in accordance with Anderson’s Functional Theory of Cognition (Anderson, 2008). The procedure had two phases. In the familiarization phase, the investigator explained to participants what was expected of them, that is, that for each scenario they were to indicate their level of willingness to receive the vaccine. Next, each participant was presented with 18 of the 48 scenarios, in random order. The participant then provided ratings of willingness to receive the vaccine. After completing the 18 ratings, the participant was allowed to compare responses and make changes until satisfied with the entire set of ratings. In the experimental phase, the whole set of 48 scenarios was presented. Each participant provided ratings at his or her own pace, but was neither allowed to compare responses nor allowed to go back and make changes. Each participant was tested individually.
Participants
The participants were unpaid volunteers recruited and tested by five research assistants, well-trained in the research technique used. The assistants contacted a total of 500 people walking along the streets of Lomé. The diversity of the population of the capital enabled the recruitment of participants with various demographic characteristics. The assistants explained to them the purpose of the study, sought their participation, and obtained informed consent. This recruitment technique was an efficient and inexpensive means of revealing the basic trends needed for what was fundamentally an exploratory study.
Material
The material consisted of 48 vignettes composed of all combinations of the five main constructs of health-protective behavior theories (Weinstein, 1993): perceived susceptibility to infection (1 chance in 10 or in 50), effectiveness of the vaccine (70% or at least 90%), perceived severity of AIDS (lethal or chronic owing to the availability of treatment), cost of the vaccine (free, US$100, or US$200), and influence of the family (encourages or does not encourage). The design was, consequently, a five within-subject factor design: susceptibility × effectiveness × severity × cost × influence of the family, 2 × 2 × 2 × 3 × 2. Under each story, there was the question “to what extent would you take the vaccine in this case?” as well as a 11-point response scale with a left-hand anchor of “certainly not” (0) and a right-hand anchor of “certainly yes” (10). The cards were arranged by chance and in a different order for each participant.
Statistical analysis
We detected strong individual differences during preliminary analysis. Accordingly, we performed cluster analysis (K-means) on the whole set of raw data as recommended by Hofmans and Mullet (2013). We then conducted separate analysis of variance (ANOVAs)—full factorial with all the interactions—on the data of each cluster, using a susceptibility × severity × effectiveness × cost × influence of the family, 2 × 2 × 2 × 3 × 2 design. Finally, we performed χ2 test to test the effects of demographic characteristics. Data were analyzed using STATISTICA 7.
Results
Participant profile
Of the 500 people contacted, 363 (205 men and 158 women) agreed to participate. The main reasons for refusal to participate were lack of available time and interest. In all, 84 percent of participants chose to be tested at their private homes and the others chose a vacant classroom at the local university. The participants took 25–35 minutes to complete the questionnaires. Their ages ranged from 18 to 62 years (mean (M) = 24.37 years, standard deviation (SD) = 4.63 years). More detailed demographic information is shown in Table 1.
Demographic characteristics of the sample and clusters.
p < .05.
Positions on vaccination
The cluster analysis gave a five-cluster solution. The main patterns of data that correspond to each cluster are shown in Figure 1. The detailed results of the corresponding ANOVAs are shown in Table 2.
Mean levels of willingness to receive the vaccine as a function of cost of the vaccine and effectiveness of the vaccine. In each panel, willingness to receive the vaccine is on the y-axis, the three levels of cost are on the x-axis, and each curve corresponds to one level of vaccine effectiveness. Each panel corresponds to one cluster: complete reluctance (Reluctant), total hesitation (Hesitant), depends on cost (Cost), depends on the cost/effectiveness ratio (Cost/Effectiveness), and unconditional acceptance (Unconditional).
Results of the ANOVAs for each cluster.
MS: mean square; df: degree of freedom.
For 178 participants (49%), the ratings were always high (M = 9.01). This cluster was called unconditional acceptance. For 71 participants (20%), ratings strongly varied as a function of cost and effectiveness; this cluster was called depends on the cost/effectiveness ratio. The cost × effectiveness interaction was significant. When the vaccine was fully effective, vaccination intention was always high (mean values ranging from 8.00 to 9.54). When it was not completely effective, intention depended strongly on cost (mean values ranging from 3.14 to 8.07). Also, when AIDS was perceived as not lethal (because of access to treatment), intention was lower (M = 6.51) than when AIDS was perceived as lethal (M = 7.53).
For 65 participants (18%), the ratings considerably varied as a function of the cost of the vaccine and this cluster was called depends on cost. When the vaccine was free, intention was higher (M = 9.03) than when it was very costly (M = 2.36) or moderately costly (M = 4.24). Effectiveness had a weak effect.
For 39 participants (11%), the ratings were concentrated in the middle of the willingness scale (M = 5.26). When family members encouraged vaccination, intention was slightly higher (M = 6.37) than when they did not (M = 4.16). This cluster was called total hesitation. Finally, for 10 participants (3%), the ratings were always low (M = 2.88). None of the factors considered in this study had a significant effect. This cluster was called complete reluctance.
The clusters differed significantly regarding educational level, χ2(4) = 11.66, p < .05, and income, χ2(4) = 35.53, p < .001. As shown in Table 1, all except one of the members of the completely reluctant cluster had not had access to post-secondary education. Also, wealthier participants (those earning more than US$100 per month, the median income in Togo) were less often members of the unconditional acceptance cluster and more often members of the depends on the cost/effectiveness ratio cluster compared to poorer ones (those earning less than US$100 per month) (World Bank, 2015).
Discussion
As expected, several qualitatively different positions were found. This finding is consistent with previous studies that show that people’s views regarding vaccination can considerably differ and reflect clearly structured personal positions that are partly associated with sociodemographic characteristics (Bynum et al., 2012; Ziemer and Hoffman, 2013).
First, vaccination acceptance is already achieved for half of the residents of Lomé. People in this group seem to be willing to get the vaccine, irrespective of its level of effectiveness, cost, or family attitudes. This result is consistent with the findings of Bishai et al. (2004) that a majority of adults in Uganda would be willing to be vaccinated against HIV even if the vaccine was not very effective. Second, for about one person in five, acceptability is determined by cost of the vaccine. This result is consistent with previous findings in South Africa (Sayles et al., 2010) and Uganda (Bishai et al., 2004) showing that cost might be a barrier to uptake of the vaccine. This suggests that the main emphasis of vaccination promotion efforts among people in this group should be on the reduction of financial hurdles. Third, for 1 person in 10, acceptability is determined by influence of the family. This suggests that community-based interventions that are engaging and persuasive to people would be necessary for people in this group. These interventions should involve community influential leaders (e.g. traditions and religious leaders) and local media. Finally, few people (3%) would be unwilling to take the vaccine, whatever its characteristics: strong resistance to its use is not expected.
Limitations of this study include the use of a convenience sample and its restriction to people living in one area. As a consequence, generalization of the results to people in Togo as a whole must be done with care. In addition, it is difficult to know how the views and demographic information of participants differed from those of the people who declined to participate. Future studies of the reproducibility of these findings in other African countries and among specific HIV-vulnerable populations are needed.
Despite its limitations, this study is the first to map the acceptability of HIV vaccine among people in Africa. It provides insights for the design and implementation of tailored HIV vaccination campaigns in Togo and probably in other sub-Saharan African countries.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.