Perceptions of Zika virus risk in Germany in 2016

Perceptions of Zika virus risk in Germany in 2016 Abstract Background Risks associated with Zika virus (ZIKV) transmission in the Americas have been discussed widely in the media as several European athletes declined to participate in the 2016 Summer Olympic Games. Since risk perceptions of individuals in unaffected areas are unknown, we assessed the risk perceptions of ZIKV and related behaviour in Lower Saxony, Germany, with a specific focus on pregnant women and their partners. Methods In May 2016, we surveyed 1,037 participants aged 15-69 years of an online panel (addressing hygiene and preventive behaviour regarding infections) in Lower Saxony with respect to their risk perceptions related to ZIKV. We additionally included 26 expectant parents who were recruited at antenatal preparation courses in Braunschweig and Hannover between May and July 2016. Results Six hundred fifty-five (69.1%) of the panel participants had ever heard about ZIKV. About 8% of the study participants reported to be concerned about ZIKV. Pregnant women had the highest odds of reporting concern about ZIKV (OR: 6.24; 95% CI: 2.94–13.26, reference: non-pregnant women). The vast majority of participants (79%) would travel to the Olympics if they won a free trip; this proportion was lower in currently pregnant women (46%). Risk perceptions towards ZIKV were considerably lower than those towards Ebola during the 2014 epidemic. Conclusion This study showed that fear of contracting ZIKV is not a major deterrent for travelling to high-risk areas. Pregnant women are appropriately concerned about the risk of ZIKV. Studies modelling the further spread of ZIKV need to account for these results. Introduction Zika virus (ZIKV) has received considerable media coverage due to the epidemic in South America, particularly in Brazil, the site of the 2016 Summer Olympic and Paralympic Games.1 The virus spreads via bites from infected Aedes aegypti mosquitoes, but can also be transmitted sexually.2–4 It causes asymptomatic to mild infections in children and adults, but can cause Guillain-Barré syndrome in some cases and microcephaly in newborns infected in utero.4,5 Due to this risk, and fear of global spread caused by returning travellers, some researchers have called for the World Health Organization (WHO) and the International Olympic Committee to postpone, relocate or cancel the 2016 Olympics.6 However, the organizations refused, arguing that risk of infection is low and the number of ZIKV cases declined sharply in recent months.7 Others contend that ZIKV fears are no reason for individuals from unaffected countries to avoid the Olympic games, except in the case of pregnant women, for whom the severity of possible outcomes outweigh the relatively low risk.8 In May 2016, WHO published a risk assessment projecting the possible spread of ZIKV to the European Union throughout the summer of 20169 and acknowledged that the risk of a ZIKV outbreak may be increased by travellers returning from the 2016 Olympic and Paralympic Games.10 As of April 2016, there had been no reported cases of locally acquired ZIKV in the European Union, but 452 cases in returning travellers.11 Despite receiving a lot of attention, there is a lack of research on the risk perceptions in currently unaffected areas, regarding both ZIKV and the risk associated with travel to the Olympics. Research on public reactions during the 2014 Ebola virus disease (EVD) outbreak in countries not directly affected revealed that, although the level of worries about EVD was low, misperceptions regarding transmission were common and could trigger inappropriate behaviour changes.12,13 In contrast to EVD, the risk posed by ZIKV may be interpreted as low because the virus does not spread directly from one person to another and only a subgroup of the population (pregnant women) is at increased risk. To understand whether the general public and pregnant women specifically, are concerned about the spread of ZIKV in the European Union and during the 2016 Olympics and Paralympics, and whether they would change their behaviour, we conducted a survey among participants in an on-going online panel of residents in Lower Saxony, Germany. We also aimed to compare individual risk perceptions in Germany between the current ZIKV outbreak and the EVD outbreak in 2014 in West Africa. Methods Study population We implemented this survey in the online panel Hygiene and Behaviour Infectious Diseases Study (HaBIDS), which started in March 2014 and aims to address human hygiene and preventive behaviour regarding infectious diseases.12,14 The panel consists of 1037 participants who complete short, online questionnaires every two to three months. Panel members come from four districts in Lower Saxony, Germany (Braunschweig, Salzgitter, Vechta and Wolfenbüttel). In each, participants were invited by proportional stratified random sampling from the population registry in the age range of 15–69 years. Since pregnant women present a special risk group for consequences of ZIKV infection and only 3.0% of the panel participants were pregnant, we additionally recruited expectant parents at antenatal preparation courses, guided delivery room tours, and information evenings at birth clinics in Braunschweig and Hannover. In May 2016, the members of the HaBIDS panel received the ZIKV online questionnaire. The separately recruited expectant parents were invited by handing out flyers including the link to the questionnaire and a unique token during recruitment. We started recruiting expectant parents at the same time at which we sent the questionnaire to the panel, and continued recruiting for two months (until July 2016). Questionnaire The ZIKV questionnaire (Supplementary figure S1) was developed based on a questionnaire about Ebola risk perception that we used in a survey in the HaBIDS study in November 2014.12 Questionnaire items were adapted to ZIKV. The questionnaire was then piloted using cognitive interviews15 in five unrelated persons. The survey consisted of 21 questions. To assess knowledge, participants were asked how they rated their personal knowledge of ZIKV (on a five-point scale from ‘very poor’ to ‘very good’). They were also asked how ZIKV can be transmitted (10 answer options which were either true or false). A score was created by adding the correct responses (1 point for each correct answer, range 0–10). In addition, we asked if they had a potential risk of contracting ZIKV during travel to affected countries, with the responses on a four-point scale from ‘highly unlikely’ to ‘highly likely’. They were also asked to rate whether the information they had received from the media was reliable on a five-point scale from ‘very poor’ to ‘very good’. Participants were also asked from which source they had heard about ZIKV. They could choose up to seven sources (print media, online media, social media, TV, radio, information from public institutions, and friends/colleagues). To assess whether the risk perception was modified by pregnancy, the participants were asked if they or their partners were currently pregnant or planning to become pregnant. Sociodemographic data (age, sex and highest educational level) of panel members was known from the baseline questionnaire of the HaBIDS study. No sociodemographic data was available for additionally recruited expectant parents, so we asked them for information about age, sex, and highest educational level. To compare risk perceptions towards ZIKV with risk perceptions towards EVD in the same study population, we merged the individual data from the ZIKV and an EVD survey12 within the same online panel. The study was approved by the Ethics Committee of Hannover Medical School and the Federal Commissioner for Data Protection and Freedom of Information in Germany. Statistical analyses To assess associations in the univariable analysis, we used chi-squared testing for categorical variables. Wilcoxon rank-sum tests were applied for continuous variables because data were not normally distributed. Logistic regression modelling was used to build two predictive models. The first was a model of concern about ZIKV. Possible predictors included age, sex, education, pregnancy status, knowledge score and perceived reliability of information from the media. The second was a model for whether participants would travel to the Olympics if they won a free trip. The initial variables were sex, age, education, pregnancy status, perceptions of ZIKV risk during travel to an affected country and perceived media reliability. Possible predictors with a p value ≤ 0.25 in the univariable analyses were included for multivariable model building. Both models were built using forward stepwise selection in which variables were added individually and kept based on a p value < 0.05 (likelihood ratio test). After variable selection, we added interactions between sex and each selected variable to the models and estimated sex-specific odds ratios.16 To check for collinearity, we assessed models at each step for changes in estimates or standard errors and, when those appeared, we created a correlation matrix to explore relationships between individual variables. In all descriptive analyses, only participants of the online panel were included to give unbiased estimates on population level. For all association analyses, the additionally recruited pregnant participants were included. To compare individual risk perceptions between ZIKV and EVD in the study population, we used McNemar’s tests for paired nominal data. Analyses were performed in Stata 12.17 Results Study population We received responses from 717 of the 1037 (69.1%) panel participants. Of those, 655 (91.4%) have ever heard about ZIKV (table 1). Almost 42% of the participants were male and 44.7% had university degrees. The median age was 50 (interquartile range 39–59). Twenty-two pregnant women, two women planning to become pregnant and two men with a pregnant partner were identified through the additional recruitment of expectant parents. Taking these additional participants into account, 5.0% of the sample was either currently pregnant or had a pregnant partner. Another 3.0% were planning to become pregnant. Table 1 Characteristics of study participants from the original online panel and from the additional sample of expectant parents Characteristics  Online panel (n = 655a) no. (%)  Additional sample (n = 26) no. (%)  Sex      Female  380 (58.0)  24 (92.3)      Male  273 (41.7)  2 (7.7)      Missing  2 (0.3)  0  Education          Lower secondary education  175 (26.7)  2 (7.7)      Still at upper secondary  12 (1.8)  0      University entrance Qualification  169 (25.8)  3 (11.5)      University degree  288 (44.0)  21 (80.8)      Missing  11 (1.7)  0  Pregnancy      Not pregnant  615 (93.9)  0      Planning to become pregnant  20 (3.1)  2 (7.7)      Currently pregnant  13 (2.0)  24 (92.3)      Missing  7 (1.1)  0  Age in years      Median (IQR)  50 (39–59)  32 (30–35)  Knowledge score      Median (IQR)  6 (3–8)  8.5 (5–10)  Characteristics  Online panel (n = 655a) no. (%)  Additional sample (n = 26) no. (%)  Sex      Female  380 (58.0)  24 (92.3)      Male  273 (41.7)  2 (7.7)      Missing  2 (0.3)  0  Education          Lower secondary education  175 (26.7)  2 (7.7)      Still at upper secondary  12 (1.8)  0      University entrance Qualification  169 (25.8)  3 (11.5)      University degree  288 (44.0)  21 (80.8)      Missing  11 (1.7)  0  Pregnancy      Not pregnant  615 (93.9)  0      Planning to become pregnant  20 (3.1)  2 (7.7)      Currently pregnant  13 (2.0)  24 (92.3)      Missing  7 (1.1)  0  Age in years      Median (IQR)  50 (39–59)  32 (30–35)  Knowledge score      Median (IQR)  6 (3–8)  8.5 (5–10)  IQR, Interquartile range. a Those replying that they have not heard from ZIKV (n = 62, 8.6%) were not reported, and excluded from all following analyses. Risk perceptions Of those who responded to the original online panel, 7.8% reported worries about ZIKV. Reporting concern about ZIKV varied by sex and pregnancy status. Women currently pregnant or planning to become pregnant had the highest percentages of concern (table 2). Asked if they would travel to the Olympics if they won a free trip, 79.1% responded ‘yes’ (table 2). Although only 36.4% of women currently pregnant were willing to travel to the Olympics, 78.2% of non-pregnant women answered that they would go (p value from chi-squared test = 0.01). The response also differed by ZIKV concern, with those concerned about ZIKV more likely to respond that they would not go (p value from chi-squared test < 0.001). Approximately 66% of participants felt that risk of contracting ZIKV in an affected country was ‘quite likely’ or ‘highly likely’ but, of those, 75.6% of individuals answered that they were willing to travel to the 2016 Olympics under our proposed scenario. Table 2 Characteristics of participants stratified by concern about ZIKV and by willingness to take free trip to Olympics Characteristics  Concerneda (n = 51) no. (%)  Not concerneda (n = 603) no. (%)  P-value  Would travelb (n = 516) no. (%)  Would not travelb (n = 136) no. (%)  P-value  Sex      0.039c      0.081c      Female  36 (70.6)  343 (56.9)    292 (56.6)  87 (64.0)        Male  14 (27.5)  259 (43.0)    224 (43.4)  47 (34.6)        Missing  1 (2.0)  1 (0.2)    0  2 (1.5)    Education      0.159c      0.160c      Lower secondary education  19 (37.3)  155 (25.7)    129 (25.0)  46 (33.8)        Still at upper secondary  1 (2.0)  11 (1.8)    10 (1.9)  2 (1.5)        University entrance Qualification  14 (27.5)  155 (25.7)    134 (26.0)  35 (25.7)        University degree  15 (29.4)  273 (45.3)    235 (45.5)  50 (36.8)        Missing  2 (3.9)  9 (1.5)    8 (1.6)  3 (2.2)    Pregnancy status      0.027c      0.002c      Male with non-pregnant partner  14 (27.5)  252 (41.8)    220 (42.6)  46 (33.8)        Male with pregnant partner  0 (0)  2 (0.3)    2 (0.4)  0        Non-pregnant female  33 (64.7)  333 (55.2)    286 (55.4)  80 (58.8)        Currently pregnant female  3 (5.9)  8 (1.3)    4 (0.8)  7 (5.2)        Missing  1 (2.0)  8 (1.3)    4 (0.8)  3 (2.2)    Age in years      0.111d      0.263d      Median (IQR)  48 (32-55)  50 (40-59)    50 (40-58)  52 (39-61)    Knowledge score      0.398d      0.046d      Median (IQR)  5 (2-9)  7 (3-8)    7 (3-9)  6 (2-8)    Infection during travel to affected country            0.004c      Likely        236 (45.7)  76 (55.9)        Unlikely        140 (27.1)  21 (15.4)        Missing        140 (27.3)  39 (28.7)    Concerned about ZIKV            <0.001c      Yes        25 (4.8)  26 (19.1)        No        491 (95.2)  109 (80.2)        Missing        0 (0.0)  1 (0.7)    Characteristics  Concerneda (n = 51) no. (%)  Not concerneda (n = 603) no. (%)  P-value  Would travelb (n = 516) no. (%)  Would not travelb (n = 136) no. (%)  P-value  Sex      0.039c      0.081c      Female  36 (70.6)  343 (56.9)    292 (56.6)  87 (64.0)        Male  14 (27.5)  259 (43.0)    224 (43.4)  47 (34.6)        Missing  1 (2.0)  1 (0.2)    0  2 (1.5)    Education      0.159c      0.160c      Lower secondary education  19 (37.3)  155 (25.7)    129 (25.0)  46 (33.8)        Still at upper secondary  1 (2.0)  11 (1.8)    10 (1.9)  2 (1.5)        University entrance Qualification  14 (27.5)  155 (25.7)    134 (26.0)  35 (25.7)        University degree  15 (29.4)  273 (45.3)    235 (45.5)  50 (36.8)        Missing  2 (3.9)  9 (1.5)    8 (1.6)  3 (2.2)    Pregnancy status      0.027c      0.002c      Male with non-pregnant partner  14 (27.5)  252 (41.8)    220 (42.6)  46 (33.8)        Male with pregnant partner  0 (0)  2 (0.3)    2 (0.4)  0        Non-pregnant female  33 (64.7)  333 (55.2)    286 (55.4)  80 (58.8)        Currently pregnant female  3 (5.9)  8 (1.3)    4 (0.8)  7 (5.2)        Missing  1 (2.0)  8 (1.3)    4 (0.8)  3 (2.2)    Age in years      0.111d      0.263d      Median (IQR)  48 (32-55)  50 (40-59)    50 (40-58)  52 (39-61)    Knowledge score      0.398d      0.046d      Median (IQR)  5 (2-9)  7 (3-8)    7 (3-9)  6 (2-8)    Infection during travel to affected country            0.004c      Likely        236 (45.7)  76 (55.9)        Unlikely        140 (27.1)  21 (15.4)        Missing        140 (27.3)  39 (28.7)    Concerned about ZIKV            <0.001c      Yes        25 (4.8)  26 (19.1)        No        491 (95.2)  109 (80.2)        Missing        0 (0.0)  1 (0.7)    a 26 participants of the additional sample were not considered for this description; 1 did not answer this question. b 26 participants of the additional sample were not considered for this description; 3 did not answer the question. c P-value from chi-squared test (not including the missing categories). d P-value from Wilcoxon rank-sum test. Knowledge Three participants (0.5%) rated their personal knowledge of ZIKV as ‘very good’ and 8.6% as ‘good’. Conversely, 44.1% felt their knowledge was ‘poor’, or ‘very poor’. However, 38.5% of the participants answered at least 8 out of 10 questions correctly. The most common incorrect answer (by 51.8% of participants) was that ZIKV is spread through material contaminated with bodily fluids of infected persons; the most common correct answer (82.1% of participants) was that ZIKV is spread by insects in South America (Supplementary figure S2). Some participants rated the information from the media about ZIKV as ‘very good’ or ‘good’ (22.0%) while the majority rated the information they received as ‘moderate’ or worse. Most participants had heard about ZIKV from two (31.0%) or three sources (32.1%): TV was the most frequent source (76.2%) while information from public institutions (4.9%) and social media (6.4%) were least frequent. There was no significant association between source of information and concern about ZIKV except for social media, with a higher percentage of concerned participants in the group that had heard about ZIKV from social media compared with the group that had not (16.7 vs. 7.2%, p value from chi-squared test = 0.027). Knowledge score value were significantly higher among participants who had heard about ZIKV from print media compared with those who had not (median knowledge score 4 vs. 2, p value from Wilcoxon rank-sum test = 0.017). Knowledge score was also significantly higher among participants who had heard about ZIKV from online media (median 4 vs. 2, P value from Wilcoxon rank-sum test < 0.001) or from public institutions (median 6.5 vs. 3, p value from Wilcoxon rank-sum test = 0.008). Multivariable analyses Model for concern about ZIKV In the univariable analysis, only sex, pregnancy and age met the statistical cut-off to be included in the model. During model building sex, age and pregnancy exhibited collinearity so only pregnancy (as the strongest predictor) was left in the model. Individuals who responded that they or their partner were currently pregnant had more than five times the odds of reporting concern about ZIKV and those who were planning to become pregnant had three times the odds, as compared with those who reported no current pregnancy (table 3). The magnitude of the odds ratios did not differ by sex. Table 3 Final predictive model for the two outcomes of interest (‘concern about ZIKV’ and ‘travel behaviour regarding the 2016 Olympics’)   Odds ratio among women (95% CI)  P-valuea  Odds ratio among men (95% CI)  P-valuea  Concern about ZIKV   Pregnancy status          Not pregnant  1 (reference)    1 (reference)            (Partner) planning to become pregnant  3.15 (0.82–12.07)  0.093  2.94 (0.33–26.13)  0.33          (Partner) currently pregnant  5.25 (2.33–11.84)  <0.001  5.88 (0.57–60.22)  0.14  Declining travel to the 2016 Olympics   Pregnancy status          Not pregnant  1 (reference)    1 (reference)            (Partner) planning to become pregnant  0.60 (0.11–3.16)  0.55  0.63 (0.12–3.31)  0.59          (Partner) currently pregnant  11.26 (3.91–32.42)  <0.001  7.42 (0.48–113.81)  0.15   Perception of risk during travel to affected countries          Unlikely  1 (reference)    1 (reference)            Likely  1.40 (0.74–2.66)  0.31  2.66 (1.10–6.43)  0.03   Concern about ZIKV          Not concerned  1 (reference)    1 (reference)            Concerned  4.32 (2.00–9.31)  <0.001  8.00 (2.32–27.62)  0.001    Odds ratio among women (95% CI)  P-valuea  Odds ratio among men (95% CI)  P-valuea  Concern about ZIKV   Pregnancy status          Not pregnant  1 (reference)    1 (reference)            (Partner) planning to become pregnant  3.15 (0.82–12.07)  0.093  2.94 (0.33–26.13)  0.33          (Partner) currently pregnant  5.25 (2.33–11.84)  <0.001  5.88 (0.57–60.22)  0.14  Declining travel to the 2016 Olympics   Pregnancy status          Not pregnant  1 (reference)    1 (reference)            (Partner) planning to become pregnant  0.60 (0.11–3.16)  0.55  0.63 (0.12–3.31)  0.59          (Partner) currently pregnant  11.26 (3.91–32.42)  <0.001  7.42 (0.48–113.81)  0.15   Perception of risk during travel to affected countries          Unlikely  1 (reference)    1 (reference)            Likely  1.40 (0.74–2.66)  0.31  2.66 (1.10–6.43)  0.03   Concern about ZIKV          Not concerned  1 (reference)    1 (reference)            Concerned  4.32 (2.00–9.31)  <0.001  8.00 (2.32–27.62)  0.001  a P-value from Wald test. Model for travel to the 2016 Olympics For the univariable analyses, sex, pregnancy status, proportions of ZIKV knowledge questions answered correctly, perceptions of infection risk when travelling to an affected country and concern about ZIKV all met the inclusion criteria for the multivariable model regarding willingness to travel to the 2016 Olympics under our hypothetical situation. The final multivariable model contained pregnancy status, perceptions of infection risk when travelling to an affected county, and concern about ZIKV (table 3). Women who were pregnant were 11 times more likely to refuse the trip; concern about ZIKV led to a 4-fold increase in odds of refusal and reporting that infection was likely during travel to an affected country led to a 1.4-fold increase in odds of refusal. Men responding that their partner were currently pregnant were seven times more likely to refuse the trip. The odds ratios of being concerned about ZIKV and reporting that infection was likely during travel were twice as high in men. Comparison between ZIKV and EVD risk perception Among the 655 participants of the ZIKV survey, 459 had also filled in the EVD survey in 2014. Of those, 26.1% had been concerned about EVD (figure 1); this is more than three times the proportion of panel participants concerned about ZIKV in the current survey. Nearly 64% of the participants who were concerned about ZIKV had also been concerned about EVD, while only 23.2% of those not concerned about ZIKV had been concerned about EVD (p value from McNemar’s test < 0.001). Moreover, 79.0% of the EVD survey participants said they would cancel travel to an affected country, even if it meant losing 100% of the travel costs. This is higher than the 30.0% who refused a free trip to the 2016 Olympics in the ZIKV survey. The perceived chance of contracting the virus in public places was also higher for EVD than for ZIKV. Travel to an affected country was seen as most risky for contracting EVD, followed by being at an airport in Germany. For ZIKV, the second highest category in perception of personal risk was also German airports. Knowledge of EVD transmission was comparable to knowledge of ZIKV transmission (median number of correct answers was 7 out of 11 for Ebola and 7 out of 10 for ZIKV using similar items). Figure 1 View largeDownload slide Personal concern and risk perception in study participants during the 2016 ZIKV epidemic in the Americas when compared with the 2014 EVD epidemic in West Africa Figure 1 View largeDownload slide Personal concern and risk perception in study participants during the 2016 ZIKV epidemic in the Americas when compared with the 2014 EVD epidemic in West Africa Discussion In looking at ZIKV risk perceptions in a country that is currently unaffected by the outbreak, we found that rates of concern about the virus were generally low. Only few participants reported worries about ZIKV. The majority of the sample reported that they would travel to the 2016 Olympics in Brazil if they won the trip. Although many participants felt that the risk of contracting ZIKV during travel to an affected country was high, they would still attend the 2016 Olympics if they won a trip. This indicates that, even though individuals think the chance of infection is high, they perceive the consequences of contracting ZIKV as mild enough to justify visiting affected countries. The notable exception was individuals reporting they were currently pregnant or had a pregnant partner. Of those participants, one third reported worries about ZIKV. The majority said they would refuse the trip to the Olympics. By the time of our survey, it was known that pregnant women can be infected with ZIKV in any trimester 18, but a causal relationship between ZIKV and microcephaly had not yet been found.19 Since there is not yet local transmission in Germany and the real threat from ZIKV results from infection during foetal development, low levels of concern while still declining to attend the Olympics may represent an appropriate understanding of the risk and consequences of infection. Of interest, the odds for refusing the trip to the Olympics was twice as high for men concerned about ZIKV than for women in the respective groups. Once men understood the risk of infection, they seemed more likely to advocate for conservative and risk-avoidant behaviour. The difference in knowledge was highest between participants who had received information about ZIKV directly from public institutions and those who had not. However, less than one tenth had received information about ZIKV directly from public institutions, while the majority had heard about ZIKV from print or online media. This suggests that information from public institutions should be spread more widely to print and online media. The perceptions of the risk associated with ZIKV were lower than those for EVD during the outbreak in West Africa in this same sample population.12 Since ZIKV infection is generally less severe than EVD and less likely to be transmitted person-to-person, it is not surprising that risk perceptions are lower although the total number of cases and the true risk of infection might be higher. However, the low levels of risk perception may have changed if infected travellers return from the Olympics or local transmission of ZIKV occurs in the European Union. Strengths and limitations The main strength of our survey is the timeliness with respect to media reports in Germany: We sent invitations to the survey shortly after the first report of sexual transmission of ZIKV in Germany on May 13.5 We were also able to compare risk perceptions with respect to EVD and ZIKV by implementing the questionnaires in our online panel. One of the limitations of this study is that data were collected from a regional sample that might not be completely generalizable to the population of Germany. Although the study population was recruited from population registries, those participating in HaBIDS are still not completely representative for the underlying population (e.g. higher education levels). There were more women enrolled in our study than there are in the general population. There were differences in age between the study participants who responded to our survey and those who did not (Supplementary table S1). Non-responders had a median age of 46 while responders had a median age of 50. Non-responders and responders did not, however, differ significantly on sex or education. Conclusion This survey, which was implemented in the spring and summer before the 2016 Olympics, shows that perceptions of risk are appropriate and that the fear of contracting ZIKV, even when travelling to affected countries, is not a major deterrent for most people. Pregnant women and their partners are likely to be appropriately concerned about the relevant risks of ZIKV, and respond accordingly. Nevertheless, adequate risk communication is important in this subgroup to prevent misconceptions about ZIKV. The comparisons with EVD reveal that ZIKV is perceived as less threatening than EVD, which may be appropriate considered the less severe population-level outcomes. The results of this study allow understanding how the ongoing ZIKV epidemic might change behaviour in not yet affected countries and will therefore be of high importance as input data for studies modelling the worldwide spread of ZIKV. Supplementary data Supplementary data are available at EURPUB online. Acknowledgements The authors acknowledge feedback received from colleagues of the Department of Epidemiology of the Helmholtz-Centre for Infection Research. Funding This study was funded by intramural funds of the Helmholtz Centre for Infection Research and did not receive external funding. Conflicts of interest: None declared. Key points Taking advantage of a population-based online panel in Germany with more than 1,000 participants we investigated risk perceptions and risk behaviour towards Zika virus (ZIKV) in an unaffected country during the ongoing epidemic in the Americas; while 69.1% of the study participants had ever heard about ZIKV, only 7.8% of those were concerned about ZIKV. Pregnancy status was the main predictor (OR 6.2 [95% CI: 2.9–13.3] for individuals currently pregnant or with pregnant partner; OR 3.3 [95% CI: 1.0–10.1] for individuals planning a pregnancy) for being concerned about ZIKV in multivariable logistic regression analyses. Although the majority of participants (79.1%) would travel if they won a free trip to the 2016 Olympics in Brazil; this proportion was significantly lower in currently pregnant women (46.2%). In a direct comparison to perceived risks during the Ebola virus outbreak in 2014/2015 concerns about EVD in 2014 (26.2%) were considerably higher than those about ZIKV in 2016 (7.8%). The results of our study need to be taken into account for informing mathematical modelling studies investigating the further international spread of ZIKV. References 1 Chang C, Ortiz K, Ansari A, Gershwin ME. The Zika outbreak of the 21st century. J Autoimmun  2016; 68: 1– 13. Google Scholar CrossRef Search ADS PubMed  2 Frank C, Cadar D, Schlaphof A, et al.   Sexual transmission of Zika virus in Germany, April 2016. Euro Surveill  2016; 21: 30252. Google Scholar CrossRef Search ADS   3 Turmel JM, Abgueguen P, Hubert B, et al.   Late sexual transmission of Zika virus related to persistence in the semen. Lancet  2016; 387: 2501. Google Scholar CrossRef Search ADS PubMed  4 Rabaan AA, Bazzi AM, Al-Ahmed SH, et al.   Overview of Zika infection, epidemiology, transmission and control measures. J Infect Public Health  2017; 10: 141– 9. Google Scholar CrossRef Search ADS PubMed  5 Robert Koch-Institute. Erste sexuelle Übertragung von Zikavirus in Deutschland (First sexual transmission of Zika virus in Germany). Epidemiol Bull . 2016; 20: 179. 6 Attaran A. Off the podium: why public health concerns for global spread of Zika virus means that Rio de Janeiro’s 2016 Olympic Games must not proceed. Available at: http://harvardpublichealthreview.org/off-the-podium-why-rios-2016-olympic-games-must-not-proceed/ (23 June 2016, date last accessed) 7 Kmietowicz Z. Olympic Games are likely to be free of Zika cases, say officials. BMJ  2016; 353: i3233. Google Scholar CrossRef Search ADS PubMed  8 Codeço C, Villela D, Gomes MF, et al.   Zika is not a reason for missing the Olympic Games in Rio de Janeiro: response to the open letter of Dr Attaran and colleagues to Dr Margaret Chan, Director - General, WHO, on the Zika threat to the Olympic and Paralympic Games. Mem Inst Oswaldo Cruz  2016; 111: 414– 5. Google Scholar CrossRef Search ADS PubMed  9 World Health Organization. Zika virus technical report. Interim Risk Assessment for WHO European Region. Available at: http://www.euro.who.int/en/health-topics/emergencies/zika-virus/technical-reports-and-guidelines-on-zika-virus/zika-virus-technical-report.-interim-risk-assessment-for-who-european-region (23 June 2016, date last accessed). 10 World Health Organization Regional Office for Europe. Frequently asked questions - Zika virus expected to spread in Europe in late spring and summer: overall risk is low to moderate. Available at: http://www.euro.who.int/en/health-topics/emergencies/zika-virus/zika-virus/frequently-asked-questions-zika-virus-expected-to-spread-in-europe-in-late-spring-and-summer-overall-risk-is-low-to-moderate#311299 (23 June 2016, date last accessed). 11 European Centre for Disease Prevention and Control. Epidemiological update: Outbreaks of Zika virus and complications potentially linked to the Zika virus infection. Available at: http://ecdc.europa.eu/en/press/news/_layouts/forms/News_DispForm.aspx?ID=1437&List=8db7286c-fe2d-476c-9133-18ff4cb1b568 (17 July 2016, date last accessed). 12 Rübsamen N, Castell S, Horn J, et al.   Ebola risk perception in Germany, 2014. Emerg Infect Dis  2015; 21: 1012– 8. Google Scholar CrossRef Search ADS PubMed  13 Prati G, Pietrantoni L. Knowledge, risk perceptions, and xenophobic attitudes: evidence from Italy during the Ebola outbreak. Risk Anal 2016;36:2000–10 14 Rübsamen N, Akmatov MK, Castell S, et al.   Comparison of response patterns in different survey designs: a longitudinal panel with mixed-mode and online-only design. Emerg Themes Epidemiol  2017; 14: 15 Porst R. Pretests zur Evaluation des Fragebogen(entwurf)s (Pretests for evaluation of the questionnaire (draft)). In: Sahner H, Bayer M, Sackmann R, editors. Fragebogen: ein Arbeitsbuch (Questionnaire: a workbook) , 4th edn Wiesbaden: VS Verlag für Sozialwissenschaften, 2014: 189– 205. 16 Figueiras A, Domenech-Massons JM, Cadarso C. Regression models: calculating the confidence interval of effects in the presence of interactions. Stat Med  1998; 17: 2099– 105. Google Scholar CrossRef Search ADS PubMed  17 StataCorp. Stata statistical software: release 12 . College Station, TX: StataCorp LP.; 2011. 18 Petersen EE, Staples JE, Meaney-Delman D, et al.   Interim guidelines for pregnant women during a Zika virus outbreak — United States, 2016. MMWR Morb Mortal Wkly Rep  2016; 65: 30– 3. Google Scholar CrossRef Search ADS PubMed  19 Nunes ML, Carlini CR, Marinowic D, et al.   Microcephaly and Zika virus: a clinical and epidemiological analysis of the current outbreak in Brazil. J Pediatr (Rio J)  2016; 92: 230– 40. Google Scholar CrossRef Search ADS PubMed  © The Author 2017. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The European Journal of Public Health Oxford University Press

Loading next page...
 
/lp/ou_press/perceptions-of-zika-virus-risk-in-germany-in-2016-zyfljNqDxO
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved.
ISSN
1101-1262
eISSN
1464-360X
D.O.I.
10.1093/eurpub/ckx092
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Risks associated with Zika virus (ZIKV) transmission in the Americas have been discussed widely in the media as several European athletes declined to participate in the 2016 Summer Olympic Games. Since risk perceptions of individuals in unaffected areas are unknown, we assessed the risk perceptions of ZIKV and related behaviour in Lower Saxony, Germany, with a specific focus on pregnant women and their partners. Methods In May 2016, we surveyed 1,037 participants aged 15-69 years of an online panel (addressing hygiene and preventive behaviour regarding infections) in Lower Saxony with respect to their risk perceptions related to ZIKV. We additionally included 26 expectant parents who were recruited at antenatal preparation courses in Braunschweig and Hannover between May and July 2016. Results Six hundred fifty-five (69.1%) of the panel participants had ever heard about ZIKV. About 8% of the study participants reported to be concerned about ZIKV. Pregnant women had the highest odds of reporting concern about ZIKV (OR: 6.24; 95% CI: 2.94–13.26, reference: non-pregnant women). The vast majority of participants (79%) would travel to the Olympics if they won a free trip; this proportion was lower in currently pregnant women (46%). Risk perceptions towards ZIKV were considerably lower than those towards Ebola during the 2014 epidemic. Conclusion This study showed that fear of contracting ZIKV is not a major deterrent for travelling to high-risk areas. Pregnant women are appropriately concerned about the risk of ZIKV. Studies modelling the further spread of ZIKV need to account for these results. Introduction Zika virus (ZIKV) has received considerable media coverage due to the epidemic in South America, particularly in Brazil, the site of the 2016 Summer Olympic and Paralympic Games.1 The virus spreads via bites from infected Aedes aegypti mosquitoes, but can also be transmitted sexually.2–4 It causes asymptomatic to mild infections in children and adults, but can cause Guillain-Barré syndrome in some cases and microcephaly in newborns infected in utero.4,5 Due to this risk, and fear of global spread caused by returning travellers, some researchers have called for the World Health Organization (WHO) and the International Olympic Committee to postpone, relocate or cancel the 2016 Olympics.6 However, the organizations refused, arguing that risk of infection is low and the number of ZIKV cases declined sharply in recent months.7 Others contend that ZIKV fears are no reason for individuals from unaffected countries to avoid the Olympic games, except in the case of pregnant women, for whom the severity of possible outcomes outweigh the relatively low risk.8 In May 2016, WHO published a risk assessment projecting the possible spread of ZIKV to the European Union throughout the summer of 20169 and acknowledged that the risk of a ZIKV outbreak may be increased by travellers returning from the 2016 Olympic and Paralympic Games.10 As of April 2016, there had been no reported cases of locally acquired ZIKV in the European Union, but 452 cases in returning travellers.11 Despite receiving a lot of attention, there is a lack of research on the risk perceptions in currently unaffected areas, regarding both ZIKV and the risk associated with travel to the Olympics. Research on public reactions during the 2014 Ebola virus disease (EVD) outbreak in countries not directly affected revealed that, although the level of worries about EVD was low, misperceptions regarding transmission were common and could trigger inappropriate behaviour changes.12,13 In contrast to EVD, the risk posed by ZIKV may be interpreted as low because the virus does not spread directly from one person to another and only a subgroup of the population (pregnant women) is at increased risk. To understand whether the general public and pregnant women specifically, are concerned about the spread of ZIKV in the European Union and during the 2016 Olympics and Paralympics, and whether they would change their behaviour, we conducted a survey among participants in an on-going online panel of residents in Lower Saxony, Germany. We also aimed to compare individual risk perceptions in Germany between the current ZIKV outbreak and the EVD outbreak in 2014 in West Africa. Methods Study population We implemented this survey in the online panel Hygiene and Behaviour Infectious Diseases Study (HaBIDS), which started in March 2014 and aims to address human hygiene and preventive behaviour regarding infectious diseases.12,14 The panel consists of 1037 participants who complete short, online questionnaires every two to three months. Panel members come from four districts in Lower Saxony, Germany (Braunschweig, Salzgitter, Vechta and Wolfenbüttel). In each, participants were invited by proportional stratified random sampling from the population registry in the age range of 15–69 years. Since pregnant women present a special risk group for consequences of ZIKV infection and only 3.0% of the panel participants were pregnant, we additionally recruited expectant parents at antenatal preparation courses, guided delivery room tours, and information evenings at birth clinics in Braunschweig and Hannover. In May 2016, the members of the HaBIDS panel received the ZIKV online questionnaire. The separately recruited expectant parents were invited by handing out flyers including the link to the questionnaire and a unique token during recruitment. We started recruiting expectant parents at the same time at which we sent the questionnaire to the panel, and continued recruiting for two months (until July 2016). Questionnaire The ZIKV questionnaire (Supplementary figure S1) was developed based on a questionnaire about Ebola risk perception that we used in a survey in the HaBIDS study in November 2014.12 Questionnaire items were adapted to ZIKV. The questionnaire was then piloted using cognitive interviews15 in five unrelated persons. The survey consisted of 21 questions. To assess knowledge, participants were asked how they rated their personal knowledge of ZIKV (on a five-point scale from ‘very poor’ to ‘very good’). They were also asked how ZIKV can be transmitted (10 answer options which were either true or false). A score was created by adding the correct responses (1 point for each correct answer, range 0–10). In addition, we asked if they had a potential risk of contracting ZIKV during travel to affected countries, with the responses on a four-point scale from ‘highly unlikely’ to ‘highly likely’. They were also asked to rate whether the information they had received from the media was reliable on a five-point scale from ‘very poor’ to ‘very good’. Participants were also asked from which source they had heard about ZIKV. They could choose up to seven sources (print media, online media, social media, TV, radio, information from public institutions, and friends/colleagues). To assess whether the risk perception was modified by pregnancy, the participants were asked if they or their partners were currently pregnant or planning to become pregnant. Sociodemographic data (age, sex and highest educational level) of panel members was known from the baseline questionnaire of the HaBIDS study. No sociodemographic data was available for additionally recruited expectant parents, so we asked them for information about age, sex, and highest educational level. To compare risk perceptions towards ZIKV with risk perceptions towards EVD in the same study population, we merged the individual data from the ZIKV and an EVD survey12 within the same online panel. The study was approved by the Ethics Committee of Hannover Medical School and the Federal Commissioner for Data Protection and Freedom of Information in Germany. Statistical analyses To assess associations in the univariable analysis, we used chi-squared testing for categorical variables. Wilcoxon rank-sum tests were applied for continuous variables because data were not normally distributed. Logistic regression modelling was used to build two predictive models. The first was a model of concern about ZIKV. Possible predictors included age, sex, education, pregnancy status, knowledge score and perceived reliability of information from the media. The second was a model for whether participants would travel to the Olympics if they won a free trip. The initial variables were sex, age, education, pregnancy status, perceptions of ZIKV risk during travel to an affected country and perceived media reliability. Possible predictors with a p value ≤ 0.25 in the univariable analyses were included for multivariable model building. Both models were built using forward stepwise selection in which variables were added individually and kept based on a p value < 0.05 (likelihood ratio test). After variable selection, we added interactions between sex and each selected variable to the models and estimated sex-specific odds ratios.16 To check for collinearity, we assessed models at each step for changes in estimates or standard errors and, when those appeared, we created a correlation matrix to explore relationships between individual variables. In all descriptive analyses, only participants of the online panel were included to give unbiased estimates on population level. For all association analyses, the additionally recruited pregnant participants were included. To compare individual risk perceptions between ZIKV and EVD in the study population, we used McNemar’s tests for paired nominal data. Analyses were performed in Stata 12.17 Results Study population We received responses from 717 of the 1037 (69.1%) panel participants. Of those, 655 (91.4%) have ever heard about ZIKV (table 1). Almost 42% of the participants were male and 44.7% had university degrees. The median age was 50 (interquartile range 39–59). Twenty-two pregnant women, two women planning to become pregnant and two men with a pregnant partner were identified through the additional recruitment of expectant parents. Taking these additional participants into account, 5.0% of the sample was either currently pregnant or had a pregnant partner. Another 3.0% were planning to become pregnant. Table 1 Characteristics of study participants from the original online panel and from the additional sample of expectant parents Characteristics  Online panel (n = 655a) no. (%)  Additional sample (n = 26) no. (%)  Sex      Female  380 (58.0)  24 (92.3)      Male  273 (41.7)  2 (7.7)      Missing  2 (0.3)  0  Education          Lower secondary education  175 (26.7)  2 (7.7)      Still at upper secondary  12 (1.8)  0      University entrance Qualification  169 (25.8)  3 (11.5)      University degree  288 (44.0)  21 (80.8)      Missing  11 (1.7)  0  Pregnancy      Not pregnant  615 (93.9)  0      Planning to become pregnant  20 (3.1)  2 (7.7)      Currently pregnant  13 (2.0)  24 (92.3)      Missing  7 (1.1)  0  Age in years      Median (IQR)  50 (39–59)  32 (30–35)  Knowledge score      Median (IQR)  6 (3–8)  8.5 (5–10)  Characteristics  Online panel (n = 655a) no. (%)  Additional sample (n = 26) no. (%)  Sex      Female  380 (58.0)  24 (92.3)      Male  273 (41.7)  2 (7.7)      Missing  2 (0.3)  0  Education          Lower secondary education  175 (26.7)  2 (7.7)      Still at upper secondary  12 (1.8)  0      University entrance Qualification  169 (25.8)  3 (11.5)      University degree  288 (44.0)  21 (80.8)      Missing  11 (1.7)  0  Pregnancy      Not pregnant  615 (93.9)  0      Planning to become pregnant  20 (3.1)  2 (7.7)      Currently pregnant  13 (2.0)  24 (92.3)      Missing  7 (1.1)  0  Age in years      Median (IQR)  50 (39–59)  32 (30–35)  Knowledge score      Median (IQR)  6 (3–8)  8.5 (5–10)  IQR, Interquartile range. a Those replying that they have not heard from ZIKV (n = 62, 8.6%) were not reported, and excluded from all following analyses. Risk perceptions Of those who responded to the original online panel, 7.8% reported worries about ZIKV. Reporting concern about ZIKV varied by sex and pregnancy status. Women currently pregnant or planning to become pregnant had the highest percentages of concern (table 2). Asked if they would travel to the Olympics if they won a free trip, 79.1% responded ‘yes’ (table 2). Although only 36.4% of women currently pregnant were willing to travel to the Olympics, 78.2% of non-pregnant women answered that they would go (p value from chi-squared test = 0.01). The response also differed by ZIKV concern, with those concerned about ZIKV more likely to respond that they would not go (p value from chi-squared test < 0.001). Approximately 66% of participants felt that risk of contracting ZIKV in an affected country was ‘quite likely’ or ‘highly likely’ but, of those, 75.6% of individuals answered that they were willing to travel to the 2016 Olympics under our proposed scenario. Table 2 Characteristics of participants stratified by concern about ZIKV and by willingness to take free trip to Olympics Characteristics  Concerneda (n = 51) no. (%)  Not concerneda (n = 603) no. (%)  P-value  Would travelb (n = 516) no. (%)  Would not travelb (n = 136) no. (%)  P-value  Sex      0.039c      0.081c      Female  36 (70.6)  343 (56.9)    292 (56.6)  87 (64.0)        Male  14 (27.5)  259 (43.0)    224 (43.4)  47 (34.6)        Missing  1 (2.0)  1 (0.2)    0  2 (1.5)    Education      0.159c      0.160c      Lower secondary education  19 (37.3)  155 (25.7)    129 (25.0)  46 (33.8)        Still at upper secondary  1 (2.0)  11 (1.8)    10 (1.9)  2 (1.5)        University entrance Qualification  14 (27.5)  155 (25.7)    134 (26.0)  35 (25.7)        University degree  15 (29.4)  273 (45.3)    235 (45.5)  50 (36.8)        Missing  2 (3.9)  9 (1.5)    8 (1.6)  3 (2.2)    Pregnancy status      0.027c      0.002c      Male with non-pregnant partner  14 (27.5)  252 (41.8)    220 (42.6)  46 (33.8)        Male with pregnant partner  0 (0)  2 (0.3)    2 (0.4)  0        Non-pregnant female  33 (64.7)  333 (55.2)    286 (55.4)  80 (58.8)        Currently pregnant female  3 (5.9)  8 (1.3)    4 (0.8)  7 (5.2)        Missing  1 (2.0)  8 (1.3)    4 (0.8)  3 (2.2)    Age in years      0.111d      0.263d      Median (IQR)  48 (32-55)  50 (40-59)    50 (40-58)  52 (39-61)    Knowledge score      0.398d      0.046d      Median (IQR)  5 (2-9)  7 (3-8)    7 (3-9)  6 (2-8)    Infection during travel to affected country            0.004c      Likely        236 (45.7)  76 (55.9)        Unlikely        140 (27.1)  21 (15.4)        Missing        140 (27.3)  39 (28.7)    Concerned about ZIKV            <0.001c      Yes        25 (4.8)  26 (19.1)        No        491 (95.2)  109 (80.2)        Missing        0 (0.0)  1 (0.7)    Characteristics  Concerneda (n = 51) no. (%)  Not concerneda (n = 603) no. (%)  P-value  Would travelb (n = 516) no. (%)  Would not travelb (n = 136) no. (%)  P-value  Sex      0.039c      0.081c      Female  36 (70.6)  343 (56.9)    292 (56.6)  87 (64.0)        Male  14 (27.5)  259 (43.0)    224 (43.4)  47 (34.6)        Missing  1 (2.0)  1 (0.2)    0  2 (1.5)    Education      0.159c      0.160c      Lower secondary education  19 (37.3)  155 (25.7)    129 (25.0)  46 (33.8)        Still at upper secondary  1 (2.0)  11 (1.8)    10 (1.9)  2 (1.5)        University entrance Qualification  14 (27.5)  155 (25.7)    134 (26.0)  35 (25.7)        University degree  15 (29.4)  273 (45.3)    235 (45.5)  50 (36.8)        Missing  2 (3.9)  9 (1.5)    8 (1.6)  3 (2.2)    Pregnancy status      0.027c      0.002c      Male with non-pregnant partner  14 (27.5)  252 (41.8)    220 (42.6)  46 (33.8)        Male with pregnant partner  0 (0)  2 (0.3)    2 (0.4)  0        Non-pregnant female  33 (64.7)  333 (55.2)    286 (55.4)  80 (58.8)        Currently pregnant female  3 (5.9)  8 (1.3)    4 (0.8)  7 (5.2)        Missing  1 (2.0)  8 (1.3)    4 (0.8)  3 (2.2)    Age in years      0.111d      0.263d      Median (IQR)  48 (32-55)  50 (40-59)    50 (40-58)  52 (39-61)    Knowledge score      0.398d      0.046d      Median (IQR)  5 (2-9)  7 (3-8)    7 (3-9)  6 (2-8)    Infection during travel to affected country            0.004c      Likely        236 (45.7)  76 (55.9)        Unlikely        140 (27.1)  21 (15.4)        Missing        140 (27.3)  39 (28.7)    Concerned about ZIKV            <0.001c      Yes        25 (4.8)  26 (19.1)        No        491 (95.2)  109 (80.2)        Missing        0 (0.0)  1 (0.7)    a 26 participants of the additional sample were not considered for this description; 1 did not answer this question. b 26 participants of the additional sample were not considered for this description; 3 did not answer the question. c P-value from chi-squared test (not including the missing categories). d P-value from Wilcoxon rank-sum test. Knowledge Three participants (0.5%) rated their personal knowledge of ZIKV as ‘very good’ and 8.6% as ‘good’. Conversely, 44.1% felt their knowledge was ‘poor’, or ‘very poor’. However, 38.5% of the participants answered at least 8 out of 10 questions correctly. The most common incorrect answer (by 51.8% of participants) was that ZIKV is spread through material contaminated with bodily fluids of infected persons; the most common correct answer (82.1% of participants) was that ZIKV is spread by insects in South America (Supplementary figure S2). Some participants rated the information from the media about ZIKV as ‘very good’ or ‘good’ (22.0%) while the majority rated the information they received as ‘moderate’ or worse. Most participants had heard about ZIKV from two (31.0%) or three sources (32.1%): TV was the most frequent source (76.2%) while information from public institutions (4.9%) and social media (6.4%) were least frequent. There was no significant association between source of information and concern about ZIKV except for social media, with a higher percentage of concerned participants in the group that had heard about ZIKV from social media compared with the group that had not (16.7 vs. 7.2%, p value from chi-squared test = 0.027). Knowledge score value were significantly higher among participants who had heard about ZIKV from print media compared with those who had not (median knowledge score 4 vs. 2, p value from Wilcoxon rank-sum test = 0.017). Knowledge score was also significantly higher among participants who had heard about ZIKV from online media (median 4 vs. 2, P value from Wilcoxon rank-sum test < 0.001) or from public institutions (median 6.5 vs. 3, p value from Wilcoxon rank-sum test = 0.008). Multivariable analyses Model for concern about ZIKV In the univariable analysis, only sex, pregnancy and age met the statistical cut-off to be included in the model. During model building sex, age and pregnancy exhibited collinearity so only pregnancy (as the strongest predictor) was left in the model. Individuals who responded that they or their partner were currently pregnant had more than five times the odds of reporting concern about ZIKV and those who were planning to become pregnant had three times the odds, as compared with those who reported no current pregnancy (table 3). The magnitude of the odds ratios did not differ by sex. Table 3 Final predictive model for the two outcomes of interest (‘concern about ZIKV’ and ‘travel behaviour regarding the 2016 Olympics’)   Odds ratio among women (95% CI)  P-valuea  Odds ratio among men (95% CI)  P-valuea  Concern about ZIKV   Pregnancy status          Not pregnant  1 (reference)    1 (reference)            (Partner) planning to become pregnant  3.15 (0.82–12.07)  0.093  2.94 (0.33–26.13)  0.33          (Partner) currently pregnant  5.25 (2.33–11.84)  <0.001  5.88 (0.57–60.22)  0.14  Declining travel to the 2016 Olympics   Pregnancy status          Not pregnant  1 (reference)    1 (reference)            (Partner) planning to become pregnant  0.60 (0.11–3.16)  0.55  0.63 (0.12–3.31)  0.59          (Partner) currently pregnant  11.26 (3.91–32.42)  <0.001  7.42 (0.48–113.81)  0.15   Perception of risk during travel to affected countries          Unlikely  1 (reference)    1 (reference)            Likely  1.40 (0.74–2.66)  0.31  2.66 (1.10–6.43)  0.03   Concern about ZIKV          Not concerned  1 (reference)    1 (reference)            Concerned  4.32 (2.00–9.31)  <0.001  8.00 (2.32–27.62)  0.001    Odds ratio among women (95% CI)  P-valuea  Odds ratio among men (95% CI)  P-valuea  Concern about ZIKV   Pregnancy status          Not pregnant  1 (reference)    1 (reference)            (Partner) planning to become pregnant  3.15 (0.82–12.07)  0.093  2.94 (0.33–26.13)  0.33          (Partner) currently pregnant  5.25 (2.33–11.84)  <0.001  5.88 (0.57–60.22)  0.14  Declining travel to the 2016 Olympics   Pregnancy status          Not pregnant  1 (reference)    1 (reference)            (Partner) planning to become pregnant  0.60 (0.11–3.16)  0.55  0.63 (0.12–3.31)  0.59          (Partner) currently pregnant  11.26 (3.91–32.42)  <0.001  7.42 (0.48–113.81)  0.15   Perception of risk during travel to affected countries          Unlikely  1 (reference)    1 (reference)            Likely  1.40 (0.74–2.66)  0.31  2.66 (1.10–6.43)  0.03   Concern about ZIKV          Not concerned  1 (reference)    1 (reference)            Concerned  4.32 (2.00–9.31)  <0.001  8.00 (2.32–27.62)  0.001  a P-value from Wald test. Model for travel to the 2016 Olympics For the univariable analyses, sex, pregnancy status, proportions of ZIKV knowledge questions answered correctly, perceptions of infection risk when travelling to an affected country and concern about ZIKV all met the inclusion criteria for the multivariable model regarding willingness to travel to the 2016 Olympics under our hypothetical situation. The final multivariable model contained pregnancy status, perceptions of infection risk when travelling to an affected county, and concern about ZIKV (table 3). Women who were pregnant were 11 times more likely to refuse the trip; concern about ZIKV led to a 4-fold increase in odds of refusal and reporting that infection was likely during travel to an affected country led to a 1.4-fold increase in odds of refusal. Men responding that their partner were currently pregnant were seven times more likely to refuse the trip. The odds ratios of being concerned about ZIKV and reporting that infection was likely during travel were twice as high in men. Comparison between ZIKV and EVD risk perception Among the 655 participants of the ZIKV survey, 459 had also filled in the EVD survey in 2014. Of those, 26.1% had been concerned about EVD (figure 1); this is more than three times the proportion of panel participants concerned about ZIKV in the current survey. Nearly 64% of the participants who were concerned about ZIKV had also been concerned about EVD, while only 23.2% of those not concerned about ZIKV had been concerned about EVD (p value from McNemar’s test < 0.001). Moreover, 79.0% of the EVD survey participants said they would cancel travel to an affected country, even if it meant losing 100% of the travel costs. This is higher than the 30.0% who refused a free trip to the 2016 Olympics in the ZIKV survey. The perceived chance of contracting the virus in public places was also higher for EVD than for ZIKV. Travel to an affected country was seen as most risky for contracting EVD, followed by being at an airport in Germany. For ZIKV, the second highest category in perception of personal risk was also German airports. Knowledge of EVD transmission was comparable to knowledge of ZIKV transmission (median number of correct answers was 7 out of 11 for Ebola and 7 out of 10 for ZIKV using similar items). Figure 1 View largeDownload slide Personal concern and risk perception in study participants during the 2016 ZIKV epidemic in the Americas when compared with the 2014 EVD epidemic in West Africa Figure 1 View largeDownload slide Personal concern and risk perception in study participants during the 2016 ZIKV epidemic in the Americas when compared with the 2014 EVD epidemic in West Africa Discussion In looking at ZIKV risk perceptions in a country that is currently unaffected by the outbreak, we found that rates of concern about the virus were generally low. Only few participants reported worries about ZIKV. The majority of the sample reported that they would travel to the 2016 Olympics in Brazil if they won the trip. Although many participants felt that the risk of contracting ZIKV during travel to an affected country was high, they would still attend the 2016 Olympics if they won a trip. This indicates that, even though individuals think the chance of infection is high, they perceive the consequences of contracting ZIKV as mild enough to justify visiting affected countries. The notable exception was individuals reporting they were currently pregnant or had a pregnant partner. Of those participants, one third reported worries about ZIKV. The majority said they would refuse the trip to the Olympics. By the time of our survey, it was known that pregnant women can be infected with ZIKV in any trimester 18, but a causal relationship between ZIKV and microcephaly had not yet been found.19 Since there is not yet local transmission in Germany and the real threat from ZIKV results from infection during foetal development, low levels of concern while still declining to attend the Olympics may represent an appropriate understanding of the risk and consequences of infection. Of interest, the odds for refusing the trip to the Olympics was twice as high for men concerned about ZIKV than for women in the respective groups. Once men understood the risk of infection, they seemed more likely to advocate for conservative and risk-avoidant behaviour. The difference in knowledge was highest between participants who had received information about ZIKV directly from public institutions and those who had not. However, less than one tenth had received information about ZIKV directly from public institutions, while the majority had heard about ZIKV from print or online media. This suggests that information from public institutions should be spread more widely to print and online media. The perceptions of the risk associated with ZIKV were lower than those for EVD during the outbreak in West Africa in this same sample population.12 Since ZIKV infection is generally less severe than EVD and less likely to be transmitted person-to-person, it is not surprising that risk perceptions are lower although the total number of cases and the true risk of infection might be higher. However, the low levels of risk perception may have changed if infected travellers return from the Olympics or local transmission of ZIKV occurs in the European Union. Strengths and limitations The main strength of our survey is the timeliness with respect to media reports in Germany: We sent invitations to the survey shortly after the first report of sexual transmission of ZIKV in Germany on May 13.5 We were also able to compare risk perceptions with respect to EVD and ZIKV by implementing the questionnaires in our online panel. One of the limitations of this study is that data were collected from a regional sample that might not be completely generalizable to the population of Germany. Although the study population was recruited from population registries, those participating in HaBIDS are still not completely representative for the underlying population (e.g. higher education levels). There were more women enrolled in our study than there are in the general population. There were differences in age between the study participants who responded to our survey and those who did not (Supplementary table S1). Non-responders had a median age of 46 while responders had a median age of 50. Non-responders and responders did not, however, differ significantly on sex or education. Conclusion This survey, which was implemented in the spring and summer before the 2016 Olympics, shows that perceptions of risk are appropriate and that the fear of contracting ZIKV, even when travelling to affected countries, is not a major deterrent for most people. Pregnant women and their partners are likely to be appropriately concerned about the relevant risks of ZIKV, and respond accordingly. Nevertheless, adequate risk communication is important in this subgroup to prevent misconceptions about ZIKV. The comparisons with EVD reveal that ZIKV is perceived as less threatening than EVD, which may be appropriate considered the less severe population-level outcomes. The results of this study allow understanding how the ongoing ZIKV epidemic might change behaviour in not yet affected countries and will therefore be of high importance as input data for studies modelling the worldwide spread of ZIKV. Supplementary data Supplementary data are available at EURPUB online. Acknowledgements The authors acknowledge feedback received from colleagues of the Department of Epidemiology of the Helmholtz-Centre for Infection Research. Funding This study was funded by intramural funds of the Helmholtz Centre for Infection Research and did not receive external funding. Conflicts of interest: None declared. Key points Taking advantage of a population-based online panel in Germany with more than 1,000 participants we investigated risk perceptions and risk behaviour towards Zika virus (ZIKV) in an unaffected country during the ongoing epidemic in the Americas; while 69.1% of the study participants had ever heard about ZIKV, only 7.8% of those were concerned about ZIKV. Pregnancy status was the main predictor (OR 6.2 [95% CI: 2.9–13.3] for individuals currently pregnant or with pregnant partner; OR 3.3 [95% CI: 1.0–10.1] for individuals planning a pregnancy) for being concerned about ZIKV in multivariable logistic regression analyses. Although the majority of participants (79.1%) would travel if they won a free trip to the 2016 Olympics in Brazil; this proportion was significantly lower in currently pregnant women (46.2%). In a direct comparison to perceived risks during the Ebola virus outbreak in 2014/2015 concerns about EVD in 2014 (26.2%) were considerably higher than those about ZIKV in 2016 (7.8%). The results of our study need to be taken into account for informing mathematical modelling studies investigating the further international spread of ZIKV. References 1 Chang C, Ortiz K, Ansari A, Gershwin ME. The Zika outbreak of the 21st century. J Autoimmun  2016; 68: 1– 13. Google Scholar CrossRef Search ADS PubMed  2 Frank C, Cadar D, Schlaphof A, et al.   Sexual transmission of Zika virus in Germany, April 2016. Euro Surveill  2016; 21: 30252. Google Scholar CrossRef Search ADS   3 Turmel JM, Abgueguen P, Hubert B, et al.   Late sexual transmission of Zika virus related to persistence in the semen. Lancet  2016; 387: 2501. Google Scholar CrossRef Search ADS PubMed  4 Rabaan AA, Bazzi AM, Al-Ahmed SH, et al.   Overview of Zika infection, epidemiology, transmission and control measures. J Infect Public Health  2017; 10: 141– 9. Google Scholar CrossRef Search ADS PubMed  5 Robert Koch-Institute. Erste sexuelle Übertragung von Zikavirus in Deutschland (First sexual transmission of Zika virus in Germany). Epidemiol Bull . 2016; 20: 179. 6 Attaran A. Off the podium: why public health concerns for global spread of Zika virus means that Rio de Janeiro’s 2016 Olympic Games must not proceed. Available at: http://harvardpublichealthreview.org/off-the-podium-why-rios-2016-olympic-games-must-not-proceed/ (23 June 2016, date last accessed) 7 Kmietowicz Z. Olympic Games are likely to be free of Zika cases, say officials. BMJ  2016; 353: i3233. Google Scholar CrossRef Search ADS PubMed  8 Codeço C, Villela D, Gomes MF, et al.   Zika is not a reason for missing the Olympic Games in Rio de Janeiro: response to the open letter of Dr Attaran and colleagues to Dr Margaret Chan, Director - General, WHO, on the Zika threat to the Olympic and Paralympic Games. Mem Inst Oswaldo Cruz  2016; 111: 414– 5. Google Scholar CrossRef Search ADS PubMed  9 World Health Organization. Zika virus technical report. Interim Risk Assessment for WHO European Region. Available at: http://www.euro.who.int/en/health-topics/emergencies/zika-virus/technical-reports-and-guidelines-on-zika-virus/zika-virus-technical-report.-interim-risk-assessment-for-who-european-region (23 June 2016, date last accessed). 10 World Health Organization Regional Office for Europe. Frequently asked questions - Zika virus expected to spread in Europe in late spring and summer: overall risk is low to moderate. Available at: http://www.euro.who.int/en/health-topics/emergencies/zika-virus/zika-virus/frequently-asked-questions-zika-virus-expected-to-spread-in-europe-in-late-spring-and-summer-overall-risk-is-low-to-moderate#311299 (23 June 2016, date last accessed). 11 European Centre for Disease Prevention and Control. Epidemiological update: Outbreaks of Zika virus and complications potentially linked to the Zika virus infection. Available at: http://ecdc.europa.eu/en/press/news/_layouts/forms/News_DispForm.aspx?ID=1437&List=8db7286c-fe2d-476c-9133-18ff4cb1b568 (17 July 2016, date last accessed). 12 Rübsamen N, Castell S, Horn J, et al.   Ebola risk perception in Germany, 2014. Emerg Infect Dis  2015; 21: 1012– 8. Google Scholar CrossRef Search ADS PubMed  13 Prati G, Pietrantoni L. Knowledge, risk perceptions, and xenophobic attitudes: evidence from Italy during the Ebola outbreak. Risk Anal 2016;36:2000–10 14 Rübsamen N, Akmatov MK, Castell S, et al.   Comparison of response patterns in different survey designs: a longitudinal panel with mixed-mode and online-only design. Emerg Themes Epidemiol  2017; 14: 15 Porst R. Pretests zur Evaluation des Fragebogen(entwurf)s (Pretests for evaluation of the questionnaire (draft)). In: Sahner H, Bayer M, Sackmann R, editors. Fragebogen: ein Arbeitsbuch (Questionnaire: a workbook) , 4th edn Wiesbaden: VS Verlag für Sozialwissenschaften, 2014: 189– 205. 16 Figueiras A, Domenech-Massons JM, Cadarso C. Regression models: calculating the confidence interval of effects in the presence of interactions. Stat Med  1998; 17: 2099– 105. Google Scholar CrossRef Search ADS PubMed  17 StataCorp. Stata statistical software: release 12 . College Station, TX: StataCorp LP.; 2011. 18 Petersen EE, Staples JE, Meaney-Delman D, et al.   Interim guidelines for pregnant women during a Zika virus outbreak — United States, 2016. MMWR Morb Mortal Wkly Rep  2016; 65: 30– 3. Google Scholar CrossRef Search ADS PubMed  19 Nunes ML, Carlini CR, Marinowic D, et al.   Microcephaly and Zika virus: a clinical and epidemiological analysis of the current outbreak in Brazil. J Pediatr (Rio J)  2016; 92: 230– 40. Google Scholar CrossRef Search ADS PubMed  © The Author 2017. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved.

Journal

The European Journal of Public HealthOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve Freelancer

DeepDyve Pro

Price
FREE
$49/month

$360/year
Save searches from
Google Scholar,
PubMed
Create lists to
organize your research
Export lists, citations
Read DeepDyve articles
Abstract access only
Unlimited access to over
18 million full-text articles
Print
20 pages/month
PDF Discount
20% off