Brief report: A health belief model approach to men’s assessment of a novel long-acting contraceptive

Abstract We investigated predictors of men’s attitudes towards and intent to use VasalgelTM, a long-acting reversible male contraceptive under development. The health belief model (HBM), with additional constructs of interpersonal factors and social norms, provided the framework. Heterosexual men (N = 146) living in Upstate New York (age: M = 24, range = 18–48) completed a survey assessing attitudes, intent, and other psychosocial characteristics. Overall men had positive attitudes towards VasalgelTM, and reported intent to use. The HBM predicted attitudes towards and intent to use VasalgelTM, and including perceived norms and interpersonal factors significantly improved the model for intent to use but not attitudes. Results indicate positive response to VasalgelTM and the utility of theory-based models.


ABOUT THE AUTHORS
Aisha King focuses on psychological and public health issues. Her work spans many public health areas; from writing on Latin American immigrant health issues, to providing crisis intervention to students, to supporting clients with psychological illnesses. She currently works with the NGO Partnership Nepal to advocate for women and children displaced by the 2015 earthquake.
Farnaz Kaighobadi is an assistant professor of Psychology at Bronx Community College, City University New York. Her research focuses on the intersection of sexuality, sexual violence, and HIVrisk behaviors among sexual minorities. She also studies the composition of social networks and their impact on the development of norms for sexual risk behaviors and the diffusion of health information.
Amy Winecoff focuses on the physiological mechanisms of affect and cognition. She researches the neural underpinnings of emotion regulation, reward processing, and psychopathology. She is currently a scientist at Charles River Analytics.

PUBLIC INTEREST STATEMENT
Though many contraceptive options are available for women, a dearth of reliable contraceptives are available for men. Vasalgel TM is a long-acting reversible contraceptive for men that is currently being developed. In the context of health psychology, this article describes heterosexual men's attitudes towards Vasalgel TM . Survey data suggests that men have generally positive attitudes towards Vasalgel TM and that a large proportion of men would likely use Vasalgel TM if it became available. Additionally, our research supports the basis of the health belief modelthat behavior is a function of the value an individual places on pregnancy avoidance and the individual's belief that a specific preventative action-getting Vasalgel TM -will achieve that goal. This study indicates that a new contraceptive would be used by many men in the United States and could alter the way the world, and men in particular, view family planning.
contraception (Klima, 1998). However, most methods of contraception are developed for and used exclusively by women, which has led to a lack of birth control options for men-condoms and vasectomy remain the only contraceptives available to men. Thus, women disproportionately carry the responsibility for birth control. Vasalgel TM (http://www.parsemusfoundation.org/projects/vasalgel/), which is currently under development, is a long-acting reversible contraceptive (LARC) for men. Vasalgel TM involves the injection of a polymer into the vas deferens. The polymer acts as a semi-solid plug, thereby preventing viable sperm from passing through and by extension, preventing pregnancy. Though Vasalgel TM may be effective for up to ten years, the process can be reversed at any time with a second injection, which dissolves the polymer (Lohiya et al., 2009). We are unaware of any academic research on the likelihood that men would use Vasalgel TM . Before this method of contraception is introduced to the public, it is important to understand the psychosocial factors that affect men's attitudes towards and willingness to use male LARCs (Heinemann, Saad, Weisemes, White, & Heinemann, 2004;Keith, Keith, Bussell, & Wells, 1975;Weinstein & Goebel, 1979). The health belief model (HBM), a psychological model often used to predict contraceptive behavior (Chernick et al.,2015;Laraque, Mclean, Brownpeterside, Ashton, & Diamond, 1997), provides a useful framework for understanding men's thoughts and feelings towards Vasalgel TM . The HBM is comprised of five primary constructs: perceived susceptibility, perceived severity, perceived benefits, perceived barriers, and self-efficacy (Glanz, Rimer, & Viswanath, 2008). The model posits that to be motivated to use contraception, an individual must consider himself to be susceptible to pregnancy (perceived susceptibility), believe that the pregnancy has potentially serious consequences (perceived severity), understand that the negative consequences of pregnancy are avoidable through effective use of contraception (perceived benefits), and believe that he is able to perform contraceptive behavior (self-efficacy). These factors are then weighed against the costs of using contraception (perceived barriers).
Although the HBM has had remarkable success predicting sexual health behaviors such as condom use (Zhao et al., 2012), it fails to capture two important aspects of contraceptive use. First, the HBM does not account for how behavior is influenced by social norms (Conner & Norman, 2005). Perceived social norms, which are often determined by culture and group norms, are an integral part of contraceptive behaviors (Fekadu & Kraft, 2001). Second, the model does not take into account how health behaviors may be affected by an individual's significant other or spouse (Gillam, 1991). Although the decision to use Vasalgel TM could be made without regard for a partner, contraceptive behaviors and decisions can never be fully captured by characterizing the individual alone (Conner & Norman, 2005).
The current study investigated how perceptions of pregnancy and Vasalgel TM influence men's attitudes and intentions towards Vasalgel TM through the lens of the HBM. Moreover, we sought to build upon the existing HBM framework by accounting for social norms as well as interpersonal factors (i.e. communication with partner). We first hypothesized that the HBM would predict men's attitudes and intent to use Vasalgel TM . Secondly, we hypothesized that the inclusion of interpersonal factors and social norms would significantly improve the predictive ability of the HBM and therefore better capture attitude towards and intent to use Vasalgel TM .

Study design and sample
A convenience sample of men at a small liberal arts college in upstate New York and surrounding area was recruited for participation. Participants were recruited via flyers posted on the college campus and in local businesses. Flyers advertised the target population (men over 18 years), link to the survey, and incentive-a chance to win one of four $25 gift cards to Amazon.com. In the online survey, participants gave informed consent and were screened to ensure they were over 18 years old, interested in having vaginal sex with women, and biologically male. Once qualified, participants were provided with a brief description of Vasalgel TM , then continued on to the main survey. Upon completion, they were debriefed and informed how to enter the gift card lottery. The protocol was approved by Bard College's Institutional Review Board.
Our 28-item survey assessed seven independent variables: men's perceptions of susceptibility to pregnancy, severity of pregnancy, benefits of Vasalgel TM , barriers to using Vasalgel TM , contraceptive self-efficacy, and social norms. Participants who were in an ongoing sexual relationship with a woman were asked a series of questions regarding interpersonal factors. Survey items used to model constructs of the HBM were adapted from Bish, Sutton, and Golombok (2000), which investigated whether the HBM could predict increased cervical smear testing in women. For a full list of survey items used in our analyses, see Table 1. Our dependent variables were men's attitudes towards and intent to use Vasalgel TM Attitude towards Vasalgel TM was measured by averaging participants' responses to the following statements: "I think that Vasalgel TM is an important invention," and "Vasalgel TM should be put on the market as soon as possible" [1 = strongly disagree to 6 = strongly agree]. Participants were also asked to rate the statements: "I would use Vasalgel TM if it became available," [1 = strongly disagree to 6 = strongly agree] and "Using Vasalgel TM as a long-term contraceptive method for me would be: [1 = very awful to 6 = very nice]." Participants' responses on these items were averaged to create a composite score for intent to use.

Data analysis
For each independent variable, a composite score was calculated by averaging the participants' ratings of items in each of the HBM subscales (negatively worded items were reverse coded). Linear regressions were then used to examine the influence of demographics on HBM constructs, social norms, and interpersonal factors. Tolerance statistics indicated that the data did not violate the assumption of multicolinearity for attitude (self-efficacy = 0.56; susceptibility = 0.63; severity = 0.75; benefits, = 0.4; barriers = 0.28; social norms = 0.63; interpersonal factors = 0.93) or intent to use (self-efficacy = 0.56; susceptibility = 0.95; severity = 0.75; benefits = 0.4; barriers = 0.54; social norms = 0.63; interpersonal factors = 0.93). The data also did not violate the assumption of independence of errors (Durbin-Watson values ranged from 1.91 to 2.24).
In order to test for scale reliability, a Cronbach's α was run for each subscale of the HBM, as well as for the scales measuring perceived social norms and interpersonal factors. Items that had intercorrelations under 0.3 (of medium strength) were removed from the scales in order to improve internal consistency of subscales. Cronbach's α for perceived severity (α = 0.77), perceived benefits (α = 0.78), perceived barriers (α = 0.83), and self-efficacy (α = 0.73) were above 0.70, indicating good scale reliability (George & Mallery, 2003). Cronbach's α for perceived susceptibility (α = 0.59), perceived social norms (α = 0.66), and interpersonal factors (α = 0.60) scales were below 0.7, which suggests questionable internal consistency (George & Mallery, 2003). However, given the centrality of these measures to this study, regression analyses retained these factors.
Two hierarchical multiple regressions were run to assess the relationship between the independent variables and men's attitudes towards and intent to use Vasalgel TM as well as the effect of any confounding demographic variables. Participant characteristics that significantly impacted attitudes towards Vasalgel TM (having heard of Vasalgel TM previously) and intent to use Vasalgel TM (having children and having heard of Vasalgel TM previously) were entered at step 1 of the hierarchical regressions. At step 2, scores on independent components of the HBM were entered. Social norms and interpersonal factors were entered at step 3 to explore whether these variables would improve model estimates above and beyond the HBM constructs.

Sample characteristics
One hundred and forty-six men completed the survey (Age: M = 24, range = 18-48 years, see Table 2 for demographic information). All participants reported being biologically male with an interest in having vaginal sex with women. Three respondents were excluded from our analyses: two outliers who scored more than 3 standard deviations from the mean on either dependent variable and one who failed to answer any items corresponding to the dependent variables. Of participants in ongoing sexual relationships with a woman (57% of the sample), condoms (used by 48%) and oral contraception (used by 44%) were the most popular forms of contraception. The next most popular method was withdrawal (used by 20%), followed by intrauterine devices (used by 8%). No participants reported using a diaphragm.

Attitude toward and intent to use Vasalgel TM
Attitudes towards Vasalgel TM were predominantly favorable (M = 4.82), as were intentions to use Vasalgel TM (M = 4.30). Specifically, 41% of all participants (n = 59) either moderately or strongly agreed with the statement "I would use Vasalgel TM if it became available" compared to 22% (n = 31) who either moderately or strongly disagreed. Fifty-seven percent of participants (n = 82) reported that using Vasalgel TM would be "very nice" or "nice," compared to only 6% (n = 9) who said it would be "awful" or "very awful." Fifty-five percent (n = 79) moderately or strongly agreed with the statement: "Vasalgel TM should be put on the market as soon as possible," whereas 6% (n = 9) moderately or strongly disagreed. Additionally, 78% of participants (n = 112) moderately or strongly agreed that "Vasalgel TM is an important invention," compared to only 3% (n = 4) who moderately or strongly disagreed.
The only demographic characteristic that had a significant effect on both intent to use and attitude towards Vasalgel TM was having heard of Vasalgel TM prior to the present study (n = 63; Intent, F (1,140) = 29.82, p < 0.001; Attitudes, F(1,140) = 32.13, p < 0.001). Currently having children (n = 5) had a significant effect on intent to use (F(1,127) = 5.99, p < 0.05), but not attitudes (F(1,127) = 1. 19, n.s.). Age, number of sexual partners in the past year or ever, ever having had sex, contraception use at last sex, current employment status, and currently being in a relationship had no significant effects (all p > 0.05).

Effect of HBM constructs, interpersonal factors and social norms on attitudes and intent to use
Scores on independent components of the HBM were entered at step 2. The hierarchical multiple regressions for both intent to use and attitude revealed that at step 2, the HBM constructs collectively were significant (full model for attitude at step 2: F(6,73) = 12.61, R 2 = 0.51, p < 0.001; full model for intent, step 2; F(7,71) = 22.17, R 2 = 0.69, p < 0.001) and that these constructs improved model predictions above and beyond any participant characteristics (See Tables 3 and 4). At step 2, self-efficacy (t(78) = 2.46, p < 0.05), perceived benefits (t(78) = 3.25, p < 0.01), and perceived barriers
At step 3, when social norms and interpersonal factors were included the full model was significant (F(9,69) = 19.37, R 2 = 0.72, p < 0.001). The overall model at step 3 was a significantly better predictor of intent to use than the overall model at step 2. However, the model at step 3 did not improve in predicting attitude (full model: F(8,71) = 9.30, R 2 = 0.51, p = 0.82; see Tables 3 and 4). Additionally, although the full model for intent to use Vasalgel TM was significant, at step 3 only the addition of social norms was independently significant (t(78) = 2.69, p < 0.01).

Discussion
Overall, men reported a high level of acceptance of Vasalgel TM as a contraceptive method. More importantly, a large proportion of men (41%) claimed that they would use Vasalgel TM if it came on the market. Supporting the primary research hypothesis, the HBM significantly predicted both attitudes towards and intent to use Vasalgel TM . Within our model, intent to use Vasalgel TM was independently predicted by perceived benefits and barriers as well as by self-efficacy and social norms. Attitude towards Vasalgel TM , however, was only independently predicted by perceived benefits. These findings indicate that the HBM is an appropriate lens through which men's contraceptive intentions and attitudes can be examined, but also that perceived benefits of using a LARC may be particularly instrumental in determining men's attitudes towards Vasalgel TM . With regard to intent to use, the data indicated that the addition of interpersonal-level factors and social norms significantly improved model predictions -although only the construct of social norms was independently predictive. Therefore, our second hypothesis was also supported. That social norms influence intent to use Vasalgel TM is an important finding in the context of contraceptive research. Many studies examine how the HBM relates to contraceptive behavior (e.g. Petosa & Jackson, 1991) or how social factors influence contraceptive use (e.g. Albarracín, Johnson, Fishbein, & Muellerleile, 2001). Converging with a limited but growing literature (e.g. Condelli, 1986), our results suggest that using the HBM in concert with social and interpersonal factors may provide a more holistic account of this contraceptive behavior.
With regard to attitude, the second hypothesis was not supported. The addition of social norms and interpersonal factors to the model had no significant effect on attitudes towards Vasalgel TM . This result indicates that while the model developed for this study provides an appropriate framework for research on intent to use Vasalgel TM , it is likely that different constructs affect attitudes. That is, general attitudes towards Vasalgel TM may be independent of social norms. While this null result was unanticipated, it may indicate that attitudes towards Vasalgel TM (though not intent to use) may not be culturally bound. Thus, Vasalgel TM could potentially be viewed as a viable contraceptive alternative not only in Western men, but also in men throughout the world. As contraceptive decisions in developing countries are often made by men, the development of male methods is particularly important for improving uptake and attitude towards family planning globally (e.g. Bankole & Singh, 1998). Future research should examine how cross-cultural attitudes toward male LARCs could be translated into cross-cultural behaviors.
Our study had several limitations. The majority of men who completed the survey were young (M = 24; SD = 5.5). Although we did not find that age was associated with attitudes towards or intent to use Vasalgel TM , contraceptive use has often been found to vary with age (Baele, Dusseldorp, & Maes, 2001). As young men in the United States are at significant risk for unanticipated pregnancy (Kann et al., 2014), our study targeted a sample most likely to benefit from a novel contraceptive method. Beyond age, this study surveyed college students and male respondents from the Hudson Valley in New York. Given that class was a significant factor in previous studies on male LARCs (Balswick, 1972) and that most of our participants had relatively high socioeconomic status due to the recruitment location, our findings may not be generalizable to all men.
Yet despite these limitations, our findings have important implications for reproductive health research and developments. This study strongly implies that young men today are ready to shoulder the contraceptive responsibility that has traditionally rested upon women. If our results generalize, a new form of contraception for men could change the global approach to reproductive health by reducing unintended pregnancies and inspiring increased inclusion of men in reproductive health services. As research into male contraception advances, findings that men would actually use a new method are critical both to encourage rapid development of products and to understand how unwillingness to use a new male contraceptive can be addressed. Furthermore, the results of this study could inform interventions aimed at encouraging men to use LARCs.