Understanding Knowledge,Attitudes,and Behaviors Toward West Nile Virus Prevention: A Survey of High-Risk Adults in Maryland

Introduction

West nile virus (WNV) remains the leading cause of domestically acquired mosquito-borne disease in the United States since its emergence in 1999 (Petersen and Fischer 2012). Following early epidemics from 1999 to 2003, WNV was established as endemic in Maryland. Dramatically increased WNV activity in 2012 illustrated the potential for resurgence following several years of declining arboviral activity (Petersen and Fischer 2012). WNV incidence increases 20-fold after the age of 50 years, particularly among persons aged 60–89 years old (Huhn et al. 2003, Hayes et al. 2005). In fact, persons at least 50 years old are more likely to develop encephalitis, aseptic meningitis, or meningoencephalitis when infected with WNV.

Health officials have identified specific personal protective behaviors (PPBs) that aid in WNV prevention. They include: (1) using insect repellent on exposed skin when outside, (2) dressing in long-sleeved shirts and long pants when outdoors, (3) avoiding outdoor activities during peak mosquito feeding hours (dusk and dawn), and (4) draining or emptying areas of standing water (potential mosquito breeding sites) around personal property (Zielinski-Gutierrez 2004, CDC 2012). Little is known about the knowledge, perceptions, and factors that shape older adults' WNV preventive behaviors. Understanding barriers to PPB practice in older adults may enable public health officials to address those barriers.

The health belief model (HBM), a widely recognized model of health behavior developed in the 1950s, examines perceptions of risk and barriers to engagement in protective health behaviors. It features six constructs: perceived susceptibility, perceived severity, perceived benefits, perceived barriers, cues to action, and self-efficacy. Since its inception, the HBM has been applied to various disciplines to characterize such behaviors as cancer screening and sexual risk behavior (Lollis et al. 1997, Yarbrough and Braden 2001). The HBM was validated for examining WNV risk behaviors by Yerby (2007), but its use in examining knowledge, awareness, and perceptions of adults toward WNV prevention has been limited.

This study aims to identify HBM constructs that predict adoption of prevention behaviors among Maryland adults 60 years of age and older. We targeted adults of at least 60 years of age due to their significant vulnerability to severe WNV disease (Murray et al. 2006, Racsa 2014, CDC 2017). For each of the six HBM constructs, six hypotheses explored the associations between that construct and each of the six WNV preventive behaviors. For example, the hypotheses for perceived susceptibility stated that persons with a high perceived susceptibility to WNV are more likely to use insect repellent, dress in long-sleeved shirts and long pants, avoid the outdoors at prime mosquito feeding hours, remove standing water, accept a WNV vaccine, and support mosquito control programs.

Few studies examine the knowledge, awareness, and perceptions of adults toward WNV prevention; most do not explore the cognitive underpinnings of WNV prevention behaviors among high-risk groups, with rare exceptions (Adams et al. 2003, Herrington 2003, Aquino et al. 2004, Trumbo and Harper 2015). This study addresses these gaps and has implications for development of behavioral interventions to reduce the WNV disease burden in this population.

Methods

We conducted a cross-sectional study of WNV knowledge, attitudes, and behaviors among adults ≥60 years old living independently in Maryland, between October and December 2012. Advanced age has been consistently cited as one of the primary risk factors associated with encephalitis and death due to WNV disease, with particular emphasis on those patients over 60 years of age (Murray et al. 2006, Gray and Webb 2014, Racsa 2014). In addition, adults ≥60 years old may encounter more physical, mental, and financial barriers when engaging in WNV prevention behaviors.

We identified 11 Maryland zip codes with at least two confirmed or probable human WNV cases reported during 2006–2011. The sampling frame of 34,000 households was purchased from a sales marketing company and contained head of household name, address, county, phone number, and age. Stratified (by zip code) random sampling generated a sample of 1700 Maryland heads of household meeting the eligibility criteria, which included age ≥60 years old, English speaking, and living independently (not in a long-term care facility or other institution). Letters describing the study were mailed to the 1700 households with the target of enrolling 256 participants.

After obtaining verbal informed consent from participants, we administered a telephone survey that took ∼15 min. A 36-item questionnaire containing Yes/No, Likert-type, categorical, and open-ended questions captured respondents' WNV knowledge, attitudes, and practices and HBM constructs such as perceived susceptibility, perceived severity, perceived benefits, perceived barriers, cues to action, and self-efficacy. Demographic characteristics, including gender, age, race, and household income (dichotomized as <$70,000 and ≥$70,000), were collected for all respondents.

Survey questions designed to measure perceived susceptibility sought to capture feelings of vulnerability to WNV disease and included measurements of both worry and likelihood of getting infected in the next year. Responses to the question examining feelings of worry were dichotomized from a 4- to 2-point scale of “worried” and “not worried.” Perceived benefits indicated an individual's perception of the value of engaging in WNV preventive behaviors and were measured with two questions: (1) how much respondents were willing to pay for a WNV vaccine and (2) whether they would support a community program to assist persons over 60 years old with window repair and draining standing water. Survey questions measuring perceived barriers to WNV prevention were both categorical and open ended and asked about failure to engage in certain WNV PPBs, namely using insect repellent when outdoors and draining standing water from objects around homes. Choices included “Nothing in my yard collects water,” and “I do not have anyone to help me.”

Qualitative open-ended responses captured other reasons for not engaging in those behaviors, besides the designated categorical choices. Those responses were captured in an “Other” response item and logically distributed into personal, environmental, and financial categories. For instance, comments about how the repellent product made the participants feel personally (e.g., “It feels sticky”; “It will make me sick”) fell into the “personal” category. Comments about its cost (e.g., “It is too expensive”) were categorized as “financial.” And comments that alluded to the environment (e.g., “It's too cold for mosquitoes”) were categorized as “environmental.”

Descriptive analyses were performed to describe demographic characteristics of the respondents and examine the frequency of self-reported protective behaviors and knowledge about and attitudes toward WNV. Pearson's chi-square tests examined associations between the six HBM constructs and demographic variables and the six outcome measures: (1) acceptance of a WNV vaccine, (2) avoidance of the outdoors at dusk and dawn, (3) dressing in long-sleeved shirts and pants when outdoors, (4) draining standing water on property, (5) use of insect repellent in the previous 90 days, and (6) support for community mosquito control programs. The relationships between each of the six behavioral and attitude outcomes and the six HBM constructs and demographic variables were further explored using multivariate logistic regression models constructed containing variables significant on bivariate analysis, potential confounders, and selected sample characteristics that were potentially meaningful. The models were built manually and using forward selection and backward elimination procedures. Final models were reviewed for convergence to a single parsimonious model. All analyses were conducted using STATA versions 12.1 and 14.2.

The study was approved by the Institutional Review Boards at the Maryland Department of Health and Mental Hygiene and the University of Maryland College Park.

Results

Table 1 summarizes the demographic characteristics of the 211 Maryland adults enrolled in the study. Respondents comprised 51% females and ranged in age from 60 to 99 years old (mean age 70 years; standard deviation = 8.22), with just over half (55%) falling between 60 and 69 years old (Table 1). The majority of enrollees were White (79%) and almost exclusively non-Hispanic (98%), and most, 53%, had a Bachelor's degree or higher.

Seventy-five percent of respondents correctly indicated that WNV is caused by mosquito bites; and 25% correctly identified adults ≥50 years of age as the age group most likely to get seriously ill from WNV infection (Table 2).

Most respondents (70%) expressed willingness to accept a WNV vaccine, if one was available (Table 2). Those who were uncomfortable receiving vaccination expressed concerns about its safety and potential side effects, citing fear that it might make them sick or interact with other medications.

Just over half of all respondents (51%) reported avoiding the outdoors during mosquito feeding hours of dusk and dawn, and 56% reported dressing in long-sleeved shirts and long pants when outdoors. The majority of respondents (52%) reported never using insect repellent on exposed skin while outdoors during the previous 90 days (Table 2).

Perceptions of WNV risk, barriers to prevention, and behavior as measured by the constructs of perceived susceptibility, perceived severity, perceived barriers, perceived benefits, cues to action, and self-efficacy are shown in Table 3. The perceived severity construct revealed that nearly all study participants (96%) believed that WNV causes serious illness. Regarding perceived susceptibility, most participants did not consider themselves susceptible to WNV disease; 34% and 38%, respectively, indicated that they had few or no worries about getting sick with WNV. Likewise, 53% reported that it was not likely and 31% reported that it was not very likely that they would get WNV disease in the next calendar year. Regarding perceived benefits of receiving the WNV vaccine, slightly more than half of respondents (56%) expressed willingness to pay for a WNV vaccine if it cost $25 or less. A small proportion of respondents perceived benefits in a community program to help seniors with window repair; of the 39% of respondents who answered that question, 19% of them expressed interest in the program.

Perceived barriers to practicing PPBs included personal, financial, and environmental barriers to using insect repellent on exposed skin and draining standing water. Personal barriers to repellent use (88% of responses) included sensitivity to chemicals and perceived immunity to mosquito bites, and environmental barriers accounted for 41% and included absence of mosquito activity in the respondent's neighborhood and lack of standing water on the property. Financial barriers to insect repellent use comprised 0.8% of all responses and consisted of inability or unwillingness to pay for repellent products and unwillingness to support taxation as a funding mechanism for mosquito control. Personal barriers to draining standing water on property comprised 7.5% of the total, while reported absence of standing water accounted for 92% of the environmental barriers. No financial barriers were cited for draining standing water.

Respondents indicated that they had received cues to action regarding the need for personal WNV prevention, that is, receipt of WNV information or knowledge of an infected person. The majority of participants (60%) reported having received some information about WNV in the past year, including mailings (4%), news reports on TV (32%), radio (6%), and newspaper (31%), and information from a healthcare provider (2%). In addition, 36% of participants indicated a preference for receiving WNV information through written materials, such as newspapers, magazines, and other mailings. Of those who reported receiving WNV information, however, 89% stated that they had not received the other cue to action: knowing someone who had ever gotten WNV. As an indication of self-efficacy, respondents expressed confidence in their ability to protect themselves and their household members from WNV, with most stating that they were somewhat (52%) or very (25%) confident.

Bivariate analyses revealed significant associations between selected HBM constructs and four of the six outcomes. Perceived susceptibility to WNV infection, in the form of worry about getting WNV, was significantly associated with more frequent insect repellent use during the previous 90 days (odds ratio [OR] = 3.04, 95% confidence interval [CI] = 1.65–5.60), draining standing water (OR = 3.33, 95% CI = 1.82–6.11), and acceptance of a WNV vaccine (OR = 3.29, 95% CI = 1.39–7.77). Perceived severity of WNV disease was also a significant predictor of draining standing water (OR = 10.54, 95% CI = 1.24–89.36). Perceived benefits in the form of willingness to pay for vaccine was significantly associated with support for community mosquito control programs (OR = 9.45, 95% CI = 1.63–54.88). The remaining two outcomes, dressing in long shirts and pants and avoiding the outdoors at peak mosquito feeding times, showed no significant associations with HBM constructs in bivariate analysis.

Age, race, and WNV knowledge demonstrated some significant relationships. Adults 60–69 years old were significantly more likely than their older counterparts to have used insect repellent in the previous 90 days (OR = 2.34, 95% CI = 1.31–4.19) and to regularly drain standing water around their homes (OR = 3.49, 95% CI = 1.89–6.43). White respondents were more likely to dress in long-sleeved shirts and pants when outdoors (OR = 0.36, 95% CI = 0.15–0.85) and to accept a WNV vaccine (OR = 4.20, 95% CI = 1.54–11.49). In addition, knowledge of WNV transmission was found to be significantly associated with frequent use of insect repellent (OR = 2.49, 95% CI = 1.27–4.89) and draining standing water (OR = 3.21, 95% CI = 1.65–6.24).

Multivariate analysis for the four outcomes with significant bivariate associations with HBM variables yielded varying explanatory variables for each outcome (Table 4). Perceived susceptibility (being worried about getting WNV) was a significant factor in three of the four final models as follows: willingness to accept a WNV vaccine, draining standing water, and insect repellent use.

Participants who expressed worry about getting WNV were over thrice more likely to accept a WNV vaccine (adjusted OR [aOR] = 3.46, 95% CI = 1.33–8.95). Worry about getting WNV was significantly associated (aOR = 2.86, 95% CI = 1.25–6.52) with draining standing water, and adults 60–69 years old were over four times more likely to drain standing water (aOR = 4.39, 95% CI = 2.01–9.58) than their older counterparts. Likewise, after adjusting for age, respondents who worried about getting WNV were more than twice as likely to have used insect repellent in the previous 90 days (aOR = 2.52, 95% CI = 1.34–4.72) as those not worried, and adults 60–69 years old were twice as likely to have used repellent (aOR = 2.10, 95% CI = 1.16–3.81).

Respondents who preferred to pay less for WNV vaccine were 16 times more likely to support mosquito control programs (aOR = 16.00, 95% CI = 1.50–170.67) after adjusting for knowing someone with WNV and education. Two HBM constructs, perceived severity of WNV and self-efficacy, had no explanatory effect on any of the four outcomes. Knowledge of WNV transmission was also not retained in the final models for any of the outcomes.

Discussion

Respondents to the survey demonstrated an awareness of WNV disease and receptivity to individual and community initiatives to prevent WNV.

Although many respondents showed high levels of awareness of the mode of WNV transmission and considered the disease a serious one, most reported little or no concern about getting it and high levels of confidence that they could protect themselves from WNV infection. To some extent, this confidence and lack of concern about vulnerability to WNV were justified by respondents' reported practice of PPBs. The single exception to routine practice of PPBs was use of insect repellent; most respondents stated that they had rarely or never used insect repellent in the past 90 days. Using repellent requires applying a chemical agent on one's skin, which distinguishes it from other WNV PPBs that do not require topical application of a product. Participants' reluctance to applying repellent may be understandable, since some people experience sensitivity to chemicals (fragrance, repellent, etc.) (Caress and Steinemann 2003). Responses also revealed gaps in understanding how mosquitoes breed, despite claims to the contrary. Many respondents claimed that nothing on their property collects water to explain why they do not drain standing water. Yet, as little as half an inch of standing water can support dozens of mosquito larva, which many residents may not realize, despite continuing efforts by health officials to convey that message (CDC 2013).

The significant associations found between worry and engagement in behaviors like using insect repellent and draining standing water suggest that feelings of vulnerability to WNV can be a powerful motivator prompting people to engage in PPBs to protect themselves from WNV. The important role of worry and vulnerability to disease in shaping perceptions and behaviors has been supported by research on other communicable diseases (Bish et al. 2011, Beaujean et al. 2013, Hanisch-Kirkbride et al. 2013). Prior research on psychosocial factors associated with Lyme disease (LD) prevention found a similar association between worry about getting infected and practice of LD preventive behaviors, such as checking skin for ticks and wearing protective clothing (Herrington 2004, Beaujean et al. 2013). Although LD is tick borne, it can be prevented through some of the same actions used to prevent WNV, including use of insect repellent on exposed skin and dressing in long-sleeved shirts and long pants when outdoors (Hayes and Piesman 2003, Herrington 2004).

The discrepancy in behavior responses between the two perceived susceptibility measures (worry vs. likely to get infected) could be explained by fear or threat arousal, as worry often equates with fear and has been demonstrated to produce varying responses among individuals seeking to avoid disease (Bish et al. 2011, Nan 2011). Specifically, worry about getting a disease can result in positive protective behaviors such as taking vaccine or using insect repellent, or it can result in inaction, in which one feels too frightened to take action and avoids any preventive steps (Senay et al. 2013). A 2013 study of zoonotic disease risk perception in a stratified random sample of wildlife health professionals also noted the powerful effect of worry/fear associated with disease risk, described as dread. Such feelings can negatively impact an individual's judgment and sense of control (Hanisch-Kirkbride et al. 2013). Accordingly, any attempts to induce worry to motivate PPBs should consider the adverse effects of fear so that they do not result in inaction.

Study participants who preferred to pay $0–$49 for a WNV vaccine were significantly less likely to have used insect repellent than those willing to pay more (≥$50) for WNV vaccine. That is, those who would invest less in a vaccine were also less likely to practice other preventive behaviors. This suggests that individuals who are willing to invest more in a WNV vaccine, perceiving a benefit in protecting themselves from WNV, may be more likely to use insect repellent than those unwilling to invest as much. A recent study of willingness to pay for vaccination against tick-borne encephalitis revealed a similar finding, concluding that increasing knowledge of the disease and access to the vaccine would likely expand vaccination rates (Slunge 2015).

People who may not consider a WNV vaccine worth a significant investment of money may still be more willing to support community-wide protective measures. This suggests that a personal cost-benefit analysis may influence WNV PPBs. Fears and insecurities about contracting WNV may override reluctance to invest the time and effort needed to properly protect oneself.

Interestingly, none of the associations between WNV preventive behaviors and perceived personal, environmental, and financial barriers to repellent use and draining standing water were supported. This contradicts the findings of some studies applying the HBM to WNV prevention behavior, which found significant associations between perceived barriers and WNV prevention practices (Aquino et al. 2004, Bitto et al. 2005, Butterworth et al. 2010). The finding in this study is consistent with that of Yerby (2007), who in her dissertation conducted the first validation of a theory-based WNV knowledge and attitude survey instrument and found that perceived barriers to insect repellent use was the only construct that did not predict PPB practice (Yerby 2007).

Consistent with previous literature, no significant associations were found between self-efficacy and any of the outcomes (Yarbrough and Braden 2001, Greene and Brinn 2003, Yerby 2007). While cues to action demonstrated significant associations with vaccine acceptance and support for community mosquito control, they demonstrated inverse associations with those outcomes. These contradictory associations may be attributable to a lack of personal experience with WNV, as most respondents reported that they did not personally know anyone with the infection. Had more participants witnessed someone they knew suffering illness from WNV infection, it may have prompted personal preventive action. Regarding self-efficacy playing no significant role in PPB engagement, several respondents expressed confidence in their ability to protect themselves and their household members from WNV, but since some did not grasp who is considered at risk for WNV, incorrectly identifying young children as among those at greatest risk for severe WNV disease, their confidence in their ability to protect those in their home may have been overstated.

These findings suggest several implications for public health practice. The high percentages of study participants who expressed willingness to accept a WNV vaccine, and to support community mosquito control efforts, suggest receptivity among this population for WNV prevention interventions at both the individual and community levels. There may be a benefit in designing and implementing targeted outreach interventions to reinforce prevention messages for adults ≥60 years old. Since most seniors get their WNV information from news media, including TV, newspapers, and radio, interventions that include public service announcements targeting adults 60 years old and older with WNV prevention messages may be most effective. Preference for receiving written materials in the form of magazines and/or mailings is noteworthy. Reinforcing WNV prevention messages through written resources that clearly explain the risk to older adults may also be an effective means of increasing PPB practice among this group. Interventions should consider development of WNV-specific factsheets and brochures; these can be mailed and also distributed through local newspapers and newsletters that can be picked up at grocery stores, senior centers, and churches. Efforts should be made to increase knowledge and awareness of the potentially severe effects of WNV disease through targeted media outreach efforts that elevate feelings of vulnerability to WNV without stimulating a fear response.

Furthermore, due to the seasonal nature of WNV and other mosquito-borne diseases, it may be worthwhile for local public health officials to partner with community organizations to offer educational seminars in the spring and summer emphasizing WNV prevention measures such as dressing in long sleeved shirts and long pants when outdoors and applying insect repellent to exposed skin (Zielinski-Gutierrez 2004). These seminars could be incorporated into existing programs targeting adults over 60 years old who spend time outdoors.

This study was limited by its use of self-reported data collected through telephone, which prevented validation of responses. In addition, the sampling frame was purchased from a marketing company that compiled data from public records, voter registration files, and other sources; gaps in those sources may have contributed to selection bias during recruitment. The timing of survey administration posed another limitation, beginning near the end of the WNV surveillance season. Respondents may have had difficulty recalling feelings and actions taken earlier in the season. Despite these limitations, this study was strengthened by its application of a theoretical framework, use of stratified random sampling to increase representativeness of the sample, and examination of Maryland adults ≥60 years old, a previously unstudied group.

Conclusions

Interventions may aid public health officials' efforts to reduce WNV incidence by directly targeting persons at highest risk of severe WNV disease. Promotion of community mosquito control programs should be emphasized, as these were well received by this high-risk age group of Maryland adults. Vaccine researchers and manufacturers should be made aware of the receptivity of this population to a WNV vaccine. The significant associations identified between perceived susceptibility to WNV and perceived benefits of selected PPBs underscore the importance of communicating WNV disease risk (susceptibility and severity) and the value of PPBs to this population, while minimizing panic. The effects of such campaigns, targeted to older adults, should be examined.