Vietnamese Health Care Providers ’ Preferences Regarding Recommendation of HPV Vaccines

Globally, cervical cancer is the second most common cancer in women and is the most common cancer among women in developing countries. Approximately half a million women are newly affected by cervical cancer each year and more than half die from the disease (International Agency for Research on Cancer [IARC], 2012). Of the estimated 266,000 annual cervical cancer deaths in 2012, almost nine out of ten occur in less developed countries (IARC, 2012). In Vietnam, cervical cancer is the fifth most frequent cancer among women, with approximately 5,174 new diagnoses and 2,472 deaths each year (Bruni et al., 2014). This largely preventable cancer has the highest age-specific incidence of all cancers among women 15-44 years of age, at 9.6 per 100,000 women in Vietnam (Bruni et al., 2014). Vaccines to prevent human papillomavirus (HPV), the causative agent in invasive cervical cancer (Walboomers et al., 1999), are potentially life-saving Abstract


Introduction
Globally, cervical cancer is the second most common cancer in women and is the most common cancer among women in developing countries. Approximately half a million women are newly affected by cervical cancer each year and more than half die from the disease (International Agency for Research on Cancer [IARC], 2012). Of the estimated 266,000 annual cervical cancer deaths in 2012, almost nine out of ten occur in less developed countries (IARC, 2012). In Vietnam, cervical cancer is the fifth most frequent cancer among women, with approximately 5,174 new diagnoses and 2,472 deaths each year (Bruni et al., 2014). This largely preventable cancer has the highest age-specific incidence of all cancers among women 15-44 years of age, at 9.6 per 100,000 women in Vietnam (Bruni et al., 2014). Vaccines to prevent human papillomavirus (HPV), the causative agent in invasive cervical cancer (Walboomers et al., 1999), are potentially life-saving as the advice of health workers (Cover et al., 2012).
Health care providers' acceptance and recommendation of vaccines may vary by country, provider characteristics, and by patient characteristics such as sex, age, and race/ ethnicity (Cover et al., 2012;Ylitalo et al., 2013). In the U.S., physician recommendation has been identified as an important predictor of uptake of HPV and other adolescent vaccines (Do et al., 2009;Gargano et al., 2013;Ylitalo et al., 2013), with a 4-fold greater likelihood of vaccination resulting from a strong recommendation . Similarly, in Vietnam, the advice of health workers is a primary motivator for parents to participate in HPV vaccine programs (Cover et al., 2012). Other influences in Vietnam include vaccination being a recognized way of preventing disease and trusted involvement of the government in vaccination programs (LaMontagne et al., 2011;Cover et al., 2012). It is important to understand physician attitudes and beliefs regarding vaccination against HPV to ensure consistent recommendation and administration of the vaccine by health care providers. The objective of this study was to compare Vietnamese health care providers' preferences and the tradeoffs that they are willing to make to recommend HPV vaccination with regard to three factors: patient age, type of HPV vaccine, and cost.

Study design and participants
The study design was cross-sectional and used survey methodology. Data collection occurred in December 2009 in Da Nang, Vietnam. A convenience sample of Vietnamese health care providers at Da Nang General Hospital and its associated clinics were recruited in person to participate in an anonymous survey about HPV vaccination. Individuals were eligible if they were at least 18 years of age, self-identified as a health care provider, and provided oral assent to participate. The study was approved under expedited review procedures by a U.S. Institutional Review Board and by the Director of Da Nang General Hospital.

Measures
The survey instrument was developed in English, translated into Vietnamese by a bilingual individual whose native language was Vietnamese, and backtranslated into English by a second bilingual individual who was unfamiliar with the original survey content. Item translation was refined during this process, and one item with no meaningful Vietnamese translation was omitted. Surveys were self-administered and anonymous.
Questions assessed socio-demographic variables, including age, marital status, number of children, and medical specialty (internal medicine, family medicine, obstetrics/gynecology [OBGYN], pediatrics, other), as well as awareness ("Have you heard about the human papillomavirus [HPV] before today?"), and knowledge regarding HPV infection (three true/false questions). Additional questions assessed the likelihood of recommending vaccination for boys (7-point Likert scale), the ideal age for vaccination (9-12, 13-15, 16-18, >18, other), and a rank-ordering (1-"most important to you" to 5-"least important to you") of the importance of different endorsements (National Department of Health, hospital colleagues, Director of the hospital, medical societies in France or the U.S., medical societies in Vietnam) on providers' willingness to recommend HPV vaccination for their patients. Overall endorsement toward recommending HPV vaccination was assessed with a single item (7-point "extremely unfavorable" to "extremely favorable"). A question assessing perception of the type of physician who would be "most likely to convince parents to vaccinate their children against HPV" was also posed. Finally, as supported by the literature (Molano et al., 2003;Miyashita et al., 2009;Poulos et al., 2011;Hoang et al., 2013), we presented 12 different scenarios manipulating three key factors (attributes) that may be associated with provider willingness to vaccinate against HPV: patient age (12, 16, or 22 years of age), vaccine cost (0 Vietnamese Dong [VND]/free vs 1,000,000 VND [about 63 United States Dollars, USD, at the time of the study]), and vaccine type (HPV2 vs HPV4). Using a full factorial design (3×2×2), we assessed providers' willingness to vaccinate in each scenario using a 10-point Likert scale where 10=extremely willing to vaccinate. All scenarios reflected a female patient and were presented in the same order for all providers.

Analysis
Descriptive statistics included frequencies, means, and standard deviations. Conjoint analysis was performed to examine providers' willingness to vaccinate across different hypothetical vaccine scenarios presented in the survey. Conjoint analysis is a methodological technique often used for assessing consumers' preferences for products, however this technique has been applied to health related studies (Stockwell et al., 2011;Poulos et al., 2011;Tsunematsu et al., 2013). The technique allows the relative importance of an attribute or factor to be estimated thus showing how the factors of interest influence each other, and assesses the weights/preferences assigned to those various factors. A conjoint analysis of 12 scenarios was used to examine the relative importance providers assigned to each of the levels within each factor. An overall willingness to vaccinate score was calculated by assessing the mean and standard deviation of each attribute. Providers' socio-demographic characteristics (age, being a physician, marital status, medical specialty, and having children) were examined with linear regression analysis. Statistical significance was declared at p<0.05.

Respondent characteristics
Surveys were returned by a total of 244 health care providers during the one-month data collection period. Nearly half (n=114, 47%) were between 18-29 years of age, with a mean age of 34±12 range of 19-80 years (Table 1). More than half of the participants (n=146, 60%) were married and 56% had at least one child. Twenty-six percent (n=64) of survey respondents were physicians; approximately one-third (36%) of the participants DOI:http://dx.doi.org/10.7314/APJCP.2015.16.12.4895 Vietnamese Health Care Providers and HPV Vaccine Recommendation identified their medical specialty as OBGYN, 24% as family practice, and 11% as pediatrics. Over sixty percent (n=152) of the providers reported spending more than half of their time working in direct patient care activities.

Awareness, knowledge, and perceptions surrounding HPV vaccination
A majority (n=207, 85%) of providers were aware of HPV; 72% recognized that HPV causes genital warts in males and females, while 49% indicated that HPV is an uncommon sexually transmitted infection (STI). Overall, 80% (n=195) of providers indicated a favorable attitude toward recommending a vaccine to prevent HPV to their patients; over three-quarters of respondents (n=189, 77%) indicated that OBGYNS are the physicians most likely to convince parents to vaccinate their children against HPV followed by pediatricians at 11%. Of the 244 respondents, 28% indicated that the ideal age range for HPV vaccination is 9-12 years. Twenty-four percent chose 13-15 years, 24% chose 16-18 years, and 17% indicated that the ideal age for HPV vaccination is greater than 18 years. Sixty-eight percent were likely to recommend HPV vaccine for boys ("extremely likely" [31%], "quite likely" [15%], and "slightly likely" [22%]). Twelve percent of providers were unlikely to recommend HPV vaccine for boys while 18% were indifferent about recommending the vaccine for boys.
Regarding external influences on providers' own opinions as to whether or not to recommend HPV vaccination to their patients, 59% (n=145) of providers gave a ranking of 1 (most important) to "endorsement by international medical societies in France or the U.S". Other entities which providers gave a ranking of 1 included endorsements by the National Department of Health (n= 64, 26%), medical societies in Vietnam (n=7, 17%) and colleagues in the hospital (n=4, 2 %). However, 72% of providers gave a ranking of 5 (least important) to "endorsement by colleagues in the hospital".

Willingness to vaccinate
Across all 12 scenarios, the mean willingness to vaccinate score was 7.62±2.62. Estimated means for each scenario are presented in Table 2. The scenario with the highest mean score (most favorable toward vaccination) was as follows: "The patient is a 22 year old girl, the vaccine is effective against the HPV types that cause both cervical cancer and genital warts (HPV4), the vaccine is free." For this scenario (scenario 9), the estimated mean score was 8.76 and 64% of providers gave this scenario a score of 10 (extremely willing to vaccinate). The scenario with the lowest estimated mean (least favorable toward vaccination) was as follows: "The patient is a 12 year old girl, the vaccine is effective against the HPV types that cause cervical cancer only, the vaccine costs 1,000,000 VND." For this scenario (scenario 3), the estimated mean score was 6.39 and only 18% of providers gave this scenario a score of 10. Of the three factors that we varied across scenarios (patient age, vaccine type, and cost), the most important factor influencing providers' willingness to vaccinate a patient was the cost of vaccination, p<0.001 (importance score=48.1) followed by patient age, p<0.001 (importance score 37.9), and vaccine type, p<0.05 (importance score 14.0), as illustrated in the Figure and Table 3. Post-hoc pairwise comparisons of patient age showed significant differences between ages 12 and 16 years (p<0.001) and 12 and 22 years (p<0.001), but not between ages 16 and 22 years (p=0.35).
We examined associations between providers' sociodemographic characteristics (age, provider type and specialty, marital status, and number of children) and their willingness to vaccinate. Older age, being a physician, being married, and having children were all associated with significant increases in willingness to recommend HPV vaccination (p<0.05). Being an OBGYN was not associated with provider's willingness to vaccinate.

Discussion
HPV vaccines are licensed and available in Vietnam; however, a growing body of literature including prior work in Vietnam (Dinh et al., 2007;Breitkopf et al., 2009) suggests that patient and provider characteristics are both influential with regard to recommendation and administration of HPV vaccines (Caskey et al., 2009;Do et al., 2009;Gargano et al., 2013;Kester et al., 2013;Perkins et al., 2013). This study of Vietnamese health care providers characterizes several factors as they relate to willingness to recommend HPV vaccines to female patients, and how three factors in particular, patient age, vaccine type, and cost, influence willingness.
Although many of the providers in this study indicated that they have heard about HPV, nearly half also indicated that it is an uncommon STI. The perception that HPV is "uncommon" could arguably affect a provider's recommendation for the need to adopt a preventive measure through vaccination. This finding is congruent with other reports of misconceptions and limited knowledge about the magnitude of cervical cancer, HPV as a cause of cervical cancer, and HPV vaccines in Vietnam (Paul et al., 2012). The current findings suggest that providers may benefit from epidemiologic estimates of various STIs in Vietnam and specific guidance surrounding discussions related to sexuality that are gender, age, and culturally appropriate to maximize comfort for providers, patients and parents.
Interestingly the providers we surveyed perceived that OBGYNS, relative to other physician specialties, were "most likely to convince parents to vaccinate their children against HPV". Subsequent analysis of the data however, showed no association between specific physician specialty (i.e. OBGYN) and reported willingness to vaccinate. These findings are important when considered with regard to the target groups for HPV vaccination i.e., girls (and boys) in their adolescence years, as they typically would receive care from pediatricians as opposed to OBGYNs. In the U.S., pediatricians, relative to other physician specialties, have the highest likelihood of consistently recommending HPV vaccination (Vadaparampil et al., 2011). Although beyond the scope of this study, there are probable cultural and health care system differences between the U.S. and Vietnam that affect the delivery of adolescent vaccines that are worthy of investigation.
In this study, 80% of those surveyed held a favorable attitude toward recommending a vaccine to prevent HPV infection, confirming other reports of general support for HPV vaccines among health care providers in Vietnam (Nghi et al., 2010). Based on analysis of providers' preference and willingness to vaccinate, we found the cost of HPV vaccination (of the three factors examined) to be the factor that most influenced providers' willingness to vaccinate, followed by patient age, and vaccine type. This finding is consistent with an earlier report addressing the cost-effectiveness of cervical cancer prevention strategies and the tradeoffs between a national and region-based policy in Vietnam (Kim et al., 2008). In this study, HPV vaccination was favored only when cost was low. On the other hand, formative studies in Vietnam have reported parents' willingness to obtain HPV vaccine for their children, regardless of cost (PATH, 2009).
HPV vaccination for the primary prevention of cervical cancer is recommended at 11 or 12 years of age from the standpoint of behavior (initiation of sexual activity and subsequent HPV exposure) and biology (immune response) (Centers for Disease Control and Prevention, 2013b). Nevertheless, in the U.S., delayed initiation has been described (Hofstetter et al., 2014) and several studies have reported that health care providers' willingness to recommend HPV vaccination is directly related to patient age (Daley et al., 2006;Kahn et al., 2007). Importantly, this pattern of increasing provider willingness with increasing patient age is reflected in actual immunization rates in the U.S. (Centers for Disease Control and Prevention, 2013c) thus the implications of providers favoring vaccination for older girls in this study are likely to be similar.
We surveyed a convenience sample of health care providers in a single setting in Da Nang, Vietnam where a majority of survey respondents were female therefore our findings have limited generalizability. Furthermore, this investigation considered the two types of HPV vaccines available at the time of the study, HPV2 and HPV4. Recent reports have identified other HPV types that are common in Vietnam (HPV 52, 58), but not targeted by the HPV2 and HPV4 vaccines (Vu and Le, 2011;Vu et al., 2013). These types are, however, covered in the nonavalent vaccine recently approved for use in the U.S., but not yet approved in Vietnam. As further investigation continues with new HPV vaccines and alternative dosing schedules (Lamontagne et al., 2013;WHO, 2014), additional behavioral, cost-effectiveness, and epidemiologic studies will be essential.

Conclusion/ recommendation
The use of conjoint analysis in this study elicited providers' preferences for age, cost and type of vaccine with regard to their willingness to recommend HPV vaccine. Although over half of providers indicated that the ideal age range for HPV vaccination in Vietnam was 9-15years, providers' willingness to vaccinate was significantly associated with increasing age in the patient scenarios (12 vs 16 vs 22 years of age). Greater understanding of providers' willingness to recommend HPV vaccines and their actual HPV vaccination practices may help guide intervention programs. Furthermore, it seems important to address the intersectionality of age and sexuality in Vietnam. It is believed that adolescents in Vietnam have relatively low rates of sexual activity and available information regarding the timing of initiation of intercourse is supportive of this belief (Bruni et al., 2014;Vu and Le, 2011). However, the evidence is weak, and for many other high-risk sexual behaviors, the data are quite incomplete (Bruni et al., 2014). Social and economic transformations in recent years have fundamentally altered the experiences and environment of Vietnamese youth, including the weakening of ideals such as chastity before marriage (Long, 2000). It will be critical to prepare and train health care providers to engage younger female patients in traditionally sensitive discussions in order to provide effective care for their sexual and reproductive health. Successful implementation of HPV vaccination programs in the setting of a developing country requires an integrated and comprehensive approach that addresses each of the factors influencing provider willingness to administer vaccines and patient acceptance of HPV vaccines