The contribution of risk perception and social norms to reported preventive behaviour against selected vector-borne diseases in Guyana

doi:10.21203/rs.3.rs-3128723/v1

Download PDF

Article

The contribution of risk perception and social norms to reported preventive behaviour against selected vector-borne diseases in Guyana

https://doi.org/10.21203/rs.3.rs-3128723/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 06 Oct, 2023

Read the published version in Scientific Reports →

You are reading this latest preprint version

Preventing vector-borne diseases (VBDs) mainly relies on effective vector control tools and strategies, which in turn depend on population acceptance and adherence. Inspired by the abundant recent literature on SARS-COV-2, we investigate the relationship between risk perception and preventive behaviour for selected VBDs and the extent to which risk perception is determined by social norms.

We use cross-sectional data collected from 497 individuals in four regions of Guyana in 2017. We use a conditional mixed process estimator with multilevel coefficients, estimated through a Generalized Linear Model (GLM) framework, applying a simultaneous equation structure.

We find robust results on malaria: risk perception was significantly influenced by the risk perception of the reference group across different definitions of the reference group, hinting at the existence of social norms. Risk perception significantly increased the likelihood of passive behaviour by 4.48%. Less clear-cut results were found for dengue.

This study applies quantitative social science methods to public health issues in the context of VBDs. Our findings point to the relevance of tailoring communications on health risks for VBDs to groups defined at the intersection of socio-economic and demographic characteristics. Such tailored strategies are expected to align risk perception among reference groups and boost preventive behaviour.

Vector-borne diseases (VBDs) account for more than 17% of all infectious diseases worldwide, causing more than 700,000 deaths annually (WHO, 2020). Despite being endemic to many areas of the globe, the morbidity and mortality burden associated with VBDs is disproportionately concentrated in low- and middle-income countries. VBDs constitute a significant threat among the poorest [1]. Due to climate change and the movement of people, the burden of disease associated with VBDs is expected to increase in endemic areas and expand geographically [2, 3]. As a matter of fact, the World Health Organization has recently warned about the expansion of arboviruses in the American region well beyond the historical transmission areas [4].

The prevention of VBDs relies on a few pharmacological measures, such as preventive chemotherapy and a series of vector control tools and strategies that either control the proliferation of vectors or avoid contact between humans and vectors (bed nets, indoor residual spraying, house improvement, skin repellents, coils, stagnant water management, etc.). Notably, most VBDs are not yet vaccine-preventable. The effectiveness of the available tools and strategies depends on individuals’ “preventive behaviour”, that is, on the acceptance of and adherence to the existing measures.

The SARS-COV-2 pandemic has re-focused the attention of researchers on the relationship between preventive behaviour and risk perception [5, 6, 7]. Although risk perception has several definitions, it is typically defined as the subjective probability of the occurrence of an adverse event in the future and the expected magnitude of its consequences [8]. Risk perception refers, therefore, to uncertain future events [9]. For this reason, preventive behaviours against uncertain adverse health events are conditioned to risk perception [10, 11, 12].

In the context of SARS-COV-2, preventive behaviour refers to a series of measures such as vaccination, mask-wearing, hand washing, and “social” distancing. It has been proven that adherence to these measures depends on how high people perceive the risk of infection or, even more impactfully, the risk of severe disease or death [13, 14, 15]. Noticeably, risk perception has also been seen as a significant determinant of preventive behaviour for VBDs [16, 17].

Several factors underlie risk perception. Besides knowledge of the mechanisms that affect the occurrence of diseases, risk perception about infections and diseases is jointly determined by personal traits and other factors such as previous life experiences or information received from one's peers. Thus, risk perception is determined at the individual level and conditioned to both personal characteristics and each individual’s peer group. Specifically, the perception of risk goes beyond individuals and constitutes a social and cultural construct reflecting values, symbols, history, and ideology [18], which reflect social norms [19]. Social norms have their origins within social groups [20]. There are two types of social groups worth distinguishing: (i) those formed by individuals that actively participate in the enforcement of social norms (e.g., family members, friends, colleagues); (ii) those formed by individuals that conform to similar others based on characteristics such as age, gender, or education, but who are not necessarily related to each other on the social sphere. The latter is named in the Economics literature as the reference group [21, 22].

The SARS-COV-2 pandemic has also drawn attention to the social norm component of preventive behaviour, where social norms can be either enforced through social interactions within groups of individuals sharing a common social space (social group (i)) and sustained by feelings of anxiety, guilt, embarrassment or shame if violated [23, 22], or defined by the most common and expected behaviour from the social group (ii). In the context of SARS-COV-2, it has been highlighted that social norms are constitutive of compliance with preventive measures [24, 25]. We argue that social norms also apply to VBDs’ preventive behaviour. However, social norms are likely to influence behaviour differently for SARS-COV-2 and VBDs, possibly due to the different transmission mechanisms of the infection. SARS-CoV-2 transmission occurs directly through social contact among human beings, while a vector mediates VBD transmission. For this reason, SARS-CoV-2-related behaviour is directly influenced by social norms emerging from both types of social groups [26] while VBD preventive behaviour is likely to be indirectly affected by social groups through their effect on risk perception [27].

This work focuses on the relationship between preventive behaviour, risk perception and social norms for selected VBDs, namely Zika virus (ZIKV) disease, malaria, dengue and cutaneous leishmaniasis (CL), in a highly understudied setting such as Guyana, South America. We gather knowledge of social science, public health, epidemiology, and entomology to generate innovative policy indications to tackle VBDs. Specifically, we sourced both the theory on social norms and the empirical methods for their identification and measurement from the social science literature and we apply those to VBD infection mechanisms and prevention.

Ethics statement

The study protocol with the reference number 265 was reviewed and approved by the Institutional Review Board (IRB) of the Ministry of Public Health of Guyana. The research was conducted in accordance with the principles of the Helsinki Declaration. All adult subjects provided written informed consent prior to participating to the study.

Data and outcome variables

We undertook a cross-section survey among 845 individuals. Of the 845 observations, the analysis was conducted on the 497 collected from private houses; we dropped, therefore, the 348 observations collected from schools, workplaces, health facilities and businesses, as there was no information on socio-economic and demographic characteristics for these observations. Information was collected between August and December 2017 in four regions of Guyana (South America), Region 1 (Barima-Waini), Region 4 (Demerara-Mahaica, the region of the country capital (Georgetown), Region 6 (East Berbice-Corentyne) and Region 8 (Potaro-Siparuni). In the regions of Demerara-Mahaica and East Berbice-Corentyne, information was uniquely collected in the main urban areas, that is, Georgetown and New Amsterdam, respectively. Regions 4 and 6 are coastal; Regions 1 and 8 are in the interior. Data were collected in different regions of the country to capture the heterogenous endemicity of the infections included in this study. In interior regions 1 and 8, where mining is the main economic activity, there is a higher concentration of malaria and CL than in regions 4 and 6 [28, 29]. While in the coastal regions, arboviruses such as dengue and ZIKV are more prevalent than in regions 1 and 8 [30]. Interviews were conducted through a structured questionnaire, which included questions on disease knowledge, risk perception and preventive behaviour for malaria, dengue, ZIKV and CL, together with questions on interviewees' socio-economic characteristics. Additional information on the study can be found in a previously published study that used the same data [16].

Individuals can rely on non-exclusive active or passive behaviour measures to face risk. Passive behaviour refers to using measures the government provides free of charge as part of the national vector-control programme. On the contrary, active behaviour involves preventive measures that individuals actively acquire. For this study, we considered active behaviour the use of any of the following preventive measures: screened windows, skin repellent, mosquito zapper rackets, beeper mosquito, mosquito coils, bracelets, or simply sitting next to a fire at night. On the other hand, passive behaviour was considered when the reported measures were indoor residual spraying (IRS), fogging or bed nets.

Empirical strategy

We base our empirical strategy on three postulates: (i) risk perception on different diseases should be estimated jointly given that they depend on common individual traits; (ii) active and passive preventive behaviours are conditioned on risk perception and jointly determined at the individual level, and (iii) social norms play a relevant role when the reference group affects individual risk perception. From an empirical perspective, the first two postulates imply that risk perception and behaviour should be jointly estimated. As for the third one, the reference group for each individual needs to be defined.

According to Manski (1993)[31] and Etilé (2007)[22], identifying the effects of social norms require defining the reference group to which each individual identifies and hinges upon a critical assumption, the exogeneity of the characteristics used to identify the social norm. Within this study, the social norm involves individuals comparing their risk perception to the group’s commonly held norm, which in turn is assumed to influence the individual’s reported behaviour. In other words, being more or less risk averse than the average risk aversion within the reference group affects self-reported risk perception. Hence, we hypothesise that one’s self-reported risk perception is conditioned by a reference group's inherent risk perception.

According to the literature, those labelled as the reference group share similar characteristics, such as ethnicity, age, gender, or education, but are unrelated to each other in the social sphere. Consequently, this group is intrinsically (more) exogenous than the group formed by those who can actively enforce social norms. In the study, following standard practice, we constructed the risk perception of the reference group as the average reported perception of the reference group once the individual risk perception is removed. Based on the literature, social norms are confirmed by the positive correlation between the group and individual risk perceptions, as introduced by postulate (iii).

Equations (1) to (6) show the details of the simultaneously estimated equations. Alternative specifications were used to accommodate the distribution of each endogenous variable. Equations (1) and (2) were estimated using Tobit, ordered Probit and ordinary least squares (OLS) methods. Equations (3) to (6) were estimated using Probit. To obtain a more straightforward interpretation and convergence of Maximum Likelihood estimates, we applied OLS estimations of risk perception (Eqs. (1) and (2)) and a Probit model to estimate active/passive behaviour (Eq. (3) to (6)).

\({d}_{r}={S}_{j}^{d}\delta +{r}_{i}+{ϵ}_{i}\)	(1)
\({m}_{r}={S}_{j}^{m}\delta +{r}_{i}+{\nu }_{i}\)	(2)
\({d}_{act}={X}_{i}^{{\prime }}\gamma +{d}_{r}{\beta }_{1}+{u}_{i}\)	(3)
\({m}_{act}={X}_{i}^{{\prime }}\gamma +{m}_{r}{\beta }_{2}+{\mu }_{i}\)	(4)
\({d}_{pass}={X}_{i}^{{\prime }}\gamma +{d}_{r}{\beta }_{3}+{\epsilon }_{i}\)	(5)
\({m}_{pass}={X}_{i}^{{\prime }}\gamma +{m}_{r}{\beta }_{4}+{\xi }_{i}\)	(6)

where d represents dengue, m malaria, S is the average perception risk for the reference group j excluding individual i, X_i is the k-vector of explanatory variables, r_i represents precipitation levels during the previous 15 (4) days for malaria (dengue). The last term of each equation represents the residual terms. We considered two dichotomous variables representing an active or passive behaviour against the stated disease risk perception. We excluded from X_i those variables that formed the reference group to avoid collinearity and allow for convergence of estimates. The analysis focuses on malaria and dengue only as both risk perception and behaviour were too low for ZIKV and CL.

When constructing a relevant reference group, we acknowledge that risk perception in health is influenced by factors such as education, income, age, or gender [32]. The size of our dataset constrains the definition of a reference group: while several factors determine the group, the group size needs to be large enough to allow for identification. Therefore, the number of individuals categorised into each combination of conditioning variables limits the number of factors considered when constructing the reference group. Given the characteristics of our population of interest, we tested for several factors that might jointly condition individual risk perception from a social and cultural perspective such as gender, age (quartiles), ethnic group and educational level.

Accordingly, we constructed the reference group, j, for each individual, i, based on all plausible combinations of these factors that showed enough comparability power, that is, enough sample size within the reference group, resulting in two potential reference groups based on: (1) age and ethnic group, and (2) age and educational level. Our main specification, which considers age and ethnic group as the determinants of the reference group, was chosen based on performance (logarithm of likelihood function) and identification of social norms. Straightforwardly, we chose the best-performing model with a reference group that significantly predicted individual behaviour. The choice of the reference group is aligned with what is consistently depicted in the literature as common determinants of homophily, major sociodemographic dimensions that stratify society, and natural identities that affect behaviour [33, 34, 35].

In addition to the reference group, we included the moving average level of precipitation over a few days preceding the interview, \(r\), as an additional covariate that may explain individual risk perception. This additional covariate satisfies the exclusion restriction. That is, we assume that precipitation records could condition risk perception but not active or passive behaviour.

Beyond risk perception, the selection of covariates, \(X\), consists of factors that may mediate how individuals react to risk. Such factors are mostly related to economic conditions. Thus, in addition to considering the floor and walls quality of dwellings, we included a measure of deprivation based on a list of 30 reported owned assets. Specifically, we used a Rausch model (a joint maximum likelihood estimator for Rausch Item Response Theory models of dichotomous items) to compute “wealth” scores for each individual [36]. Also, we included demographic characteristics, such as gender, ethnic group, marital status, education, and having relatives living abroad (the diaspora outside Guyana has been a major phenomenon for many years [37]). Finally, we controlled for region-fixed effects. For estimation purposes, when a reference group included any of these characteristics, we excluded them from the list of covariates.

Robustness checks

We performed a secondary analysis and sensitivity analyses to test the robustness of our results to some of the definitions adopted in our main specification. As a secondary analysis, we estimated the same model by altering the reference group: age and educational level. In this case, we excluded education from the list of covariates, X, in regressions (3)-(6) but included the deprivation measure. The deprivation index was not considered in our main specification because of our sample's high correlation between wealth and educational levels. As a sensitivity analysis, we first consider active/passive behaviour as an index of reported behaviours instead of a dichotomous variable. That is, we constructed a count of all reported active/passive behaviours for each disease. In this case, all six regressions were estimated by ordered probit.

Second, we tested the sensitivity of our results to the exclusion of bednets as a (passive) measure. The reason for such exclusion is bednets can also be directly bought by individuals as a substitute or in addition to those distributed by (1) the Ministry of Health during mass campaigns, at the health facilities or on health outreaches; (2) camp managers to miners on the workplace. In our main specification, the reason why we considered bed nets as passive behaviour is two-fold: (i) liquidity constraints have been seen as major barriers towards bednet purchase and, in the absence of subsidies, the ownership rate is likely to be low [38]; (ii) a large effort is put in place by the government of Guyana to distribute highly effective long-lasting impregnated nets (LLINs) to fight, primarily, malaria. A large governmental distribution campaign was conducted in 2018, one year after our survey. Despite that, anticipatory effects may have shaped the population's behaviour towards not actively acquiring bed nets as the government conducted sensitisation campaigns before the actual bednet distribution.

Risk perception across the selected diseases showed different patterns. Average/median risk scores on a 0–10 scale were the following: ZIKV (1.68/0), dengue (3.39/3), malaria (4.94/5) and CL (0.63/0). The overall unweighted average risk perception turned out to be 2.66 (median = 2.75). Figure 1a displays boxplots of the risk perception distribution for each disease. We observe that CL, mostly unknown across regions, was not perceived as a risk. In contrast, individuals were more risk-averse against malaria, showing the highest heterogeneity across the four diseases (largest interquartile range (IQR)). Dengue and ZIKV were next in risk perception descending order. To correct for misreporting, we set risk perception as 0 when the individual had no knowledge about the disease. As could be expected, the magnitude of pairwise correlations between knowledge and risk are remarkable: ZIKV (0.68), dengue (0.71), malaria (0.50) and CL (0.61). Figure 1b shows that individuals in coastal areas (regions 4 and 6) have higher relative risk perception for ZIKV and Dengue fever compared to hinterland regions, accordingly to the incidence of these diseases in those regions.

As expected, active measures were, on average, used by a significantly lower proportion of individuals compared to passive ones: zika (26.2% vs 37.2%, Chi² = 97.34, p < 0.001), dengue (24.1% vs 60.1%%, Chi² = 63.97, p < 0.001), malaria (44.2% vs 82.2%%, Chi² = 42.66, p < 0.001) and leishmaniasis (4.3% vs 11.7%%, Chi² = 132.38, p < = 0.001). Figure 2 displays heterogeneity across regions. Notably, Region 8 showed the lowest active average behaviour compared to regions where the prevalence of Malaria is lower. ZIKV and CL were excluded from the analysis because the observed risk perception for both diseases does not show enough statistical variation.

Figures A1-A3 (Supplementary materials A) show risk perception according to education, ethnicity and age. While dengue risk perception is not correlated with age (p > 0.05), it is significantly correlated with ethnicity and education (Chi² = 24.77, p = 0.006 and Chi² = 84.17, p < 0.001, respectively). In the case of malaria, risk perception is significantly correlated with the age group (Chi² = 66.56, p < 0.001), education (Chi² = 31.63, p < 0.001) and ethnicity (Chi² = 119.08, p < 0.001), with Amerindians having the higher risk perception in comparison the other ethnic groups.

Figures A4-A6 (supplementary materials A) show similar behaviour coverages, active and passive, for dengue disease across age groups. We find significant correlations between dengue active behaviour and education (Chi² = 10.07, p = 0.002), and dengue passive behaviour and ethnicity (Chi² = 24.66, p < 0.001). For malaria, we observe differing active and passive behaviour across age groups (Chi² = 15.60, p = 0.001 and Chi² = 12.91, p = 0.005, respectively) and ethnicity (Chi² = 13.35, p = 0.004 and Chi² = 40.52, p < 0.001, respectively). The highest coverage (96%) is recorded for malaria passive behaviour among the Amerindian ethnic group. In addition, the two education groups differ in the reported active behaviour (Chi² = 17.97, p < 0.001).

Table 1 shows the list of covariates that may condition risk perception and preventive health behaviours included in our empirical analysis. The average age of our sample is approximately 42 years, with a significantly higher prevalence of female individuals (76%). Our sample is evenly spread across the four regions and across four ethnicities: African (26%), Amerindian (24%), East Indian (20%), and mixed (30%). The deprivation index categorised individuals into five wealth groups: low (22%), low-middle (25%), middle (18%), upper-middle (22%), and high (14%). Floors were reported of higher average quality than walls, with considerably lower reporting of rotten floors (3%) compared to bad quality walls (36%). In our sample, 35% reported being married, 33% in a common law partnership and 24% single. A marginal number of individuals reported being either separated/divorced (4%) or widowed (4%). In terms of education, the highest proportion of individuals reported having either secondary or tertiary level (69%). Sixty-five% reported having at least one household member living abroad.

Table 1. Summary statistics of covariates

Variables

Mean/Frequency (std. dev.)

Age

Female

Ethnic group

African

Amerindian

East Indian

Mixed

Region

Region 1

Region 4

Region 6

Region 8

Deprivation Index

Low income

Low-middle income

Middle income

Upper middle income

Higher income

Floor quality

Bad floor quality

Average floor quality

Good floor quality

Walls quality

Bad walls quality

Average walls quality

Good walls quality

Abroad

No relatives abroad

Relatives abroad

Marital status

Common law

Married

Separated/divorced

Single

Widow/Widower

Educational level

Primary or less

Secondary/tertiary

41.84 (15.83)

0.76 (0.43)

0.26 (0.44)

0.24 (0.43)

0.20 (0.40)

0.30 (0.46)

0.29 (0.45)

0.27 (0.44)

0.21 (0.41)

0.23 (0.42)

0.22 (0.41)

0.25 (0.43)

0.18 (0.38)

0.22 (0.41)

0.14 (0.35)

0.03 (0.18)

0.51 (0.50)

0.46 (0.50)

0.36 (0.48)

0.37 (0.48)

0.27 (0.45)

0.35 (0.48)

0.65 (0.48)

0.33 (0.47)

0.35 (0.48)

0.04 (0.20)

0.24 (0.42)

0.04 (0.21)

0.31 (0.46)

0.69 (0.46)

Figure 3 shows the marginal effects of the six-equations model. Both precipitation and the risk perception of the reference group were relevant and positively associated with the individual’s risk perception (equations 1 and 2). Thus, the more abundant the rain on the days prior to the interview, the higher the risk perception. The greater the risk perception of the reference group, the higher the individual risk perception.

After considering a simultaneous joint decision by individuals, risk perception was positively statistically associated with passive measures adopted for malaria (p < 0.001). Hence, the higher the risk perception for malaria, the greater the likelihood of accepting passive measures against the infection. Per each additional point increase in the 10-point Likert scale, the likelihood of acceptance of passive malaria preventive measures increases by 4.48% (95% CI: 2.10%-6.86%). Risk perception does not predict behaviour in the case of dengue and only weakly predicts malaria active behaviour (p = 0.074).

Regarding other predictors of active/passive behaviour, higher education levels are associated with active preventive behaviours against infections but not with passive ones. There are regional differences compared to the hinterland base category (Region 1). Specifically, individuals based in Region 8 showed lower probabilities of carrying out both active and passive measures and individuals from coastal areas (Regions 4 and 6) were less willing to tackle dengue passively. People in Region 6 were also less prone to uptake passive protective measures against malaria. In coastal areas, no statistical differences were found for active behaviour.

When estimating the model with age and education as an alternative definition of the reference group, results remain consistent in describing the relationship between malaria risk perception and passive preventive behaviour (Fig. 4). Nevertheless, malaria risk perception becomes statistically significant both for passive and active behaviour (p < 0.001). Increased malaria risk perception is associated with an increased (willingness to) uptake of both active and passive behaviours against the disease by 5.50% (95% CI: 2.67%-8.33%) and 4.42% (95% CI: 1.96%-6.88%), respectively, per unitary increase in the risk perception scale. Moreover, the risk perception of the reference group based on age and education ceases to explain dengue individual risk perception, pointing, thus, to the absence of a social norm. Last, results show that higher income levels were positively correlated to active behaviour exclusively for malaria, in contrast to education which influenced active behaviour irrespective of the disease.

Figures B1 and B2 show that results are also consistent when considering active/passive behaviour as an index: risk perception predicts the willingness to use passive measures against malaria. However, when we exclude bednets from the analysis, leaving IRS and fogging as the only passive behaviour, results vary (figures B3 and B4, in the supplemental material B). Specifically, irrespective of the definition of the reference group, passive behaviour against dengue is explained by risk perception (p < 0.001), while risk perception ceases to be significant in the malaria passive behaviour equation.

Understanding the extent of the social determinants of preventive behaviour against vector-borne diseases (VBDs) is key to drawing effective health policies. When social norms are proven to influence individual behaviour, tailored interventions can be targeted to identified social groups, potentially multiplying the effects of policy interventions [39]. This is particularly important in areas of low transmission where disease elimination is within reach. Still, the risk perception is likely to be low, and the path towards the last mile entails the need for nudging preventive behaviours through, for example, the spread of information on the risk of outbreaks and resurgences [40, 41].

This study analysed the relationship between individual risk perception and preventive behaviour towards selected VBDs. In addition, we investigated the social dimension of individual preventive behaviour by testing, within identified reference groups, the existence of social norms influencing individual risk perception and, consequently, individual preventive behaviour. Due to the very nature of vector-mediated transmission mechanisms of VBDs, we considered social norms that emerge from the reference group, that is a group of individuals that share similar traits but are not necessarily linked in the social sphere. We hypothesised that social norms exist through the mediation of risk perception.

Descriptively, we found that both risk perception and behaviour, particularly passive behaviour, were higher for malaria than for any other VBD. The infection that ranked second was dengue. Risk perception and passive behaviour were the highest in regions 1 and 8 for malaria and dengue. In contrast, region 8 had the lowest reporting of active behaviour, probably due to regional wealth disadvantages (Supplementary material A, table A1) [42, 43, 44]. The risk perception on CL was the lowest and so was behaviour. These figures reflect the available information on the epidemiology of the infections. As reported in a previous study, malaria was on the rise between 2016 and 2017, with the number of yearly cases being around 13,000 [16] and there is also evidence of increasing incidence at the borders between Brazil, Guyana and Venezuela [45]. The number of cases detected is far lower for dengue and ZIKV and even lower for CL. Although the number of cases detected is most likely a consequence of the testing/laboratory capacity of the country, these figures are a signal of the risk that people may perceive.

We found that the reference group influences individual risk perception both for malaria and dengue. In the case of malaria, this was apparent for the two definitions of reference group: (i) age and ethnic group, and (ii) age and education. The results show that risk perception significantly and positively affects malaria preventive behaviour, both passive and active, when the reference group is based on age and education; and passive malaria behaviour when the reference group was based on age and ethnic group. However, the association with passive behaviour disappeared when bed nets were withdrawn from the analysis. That is, risk perception only predicted the uptake of passive malaria behaviour when it included the acceptance and usage of bednets. This highlights the perceived high effectiveness of bednets as a tool against malaria. Similarly, the exclusion of bednets as a form of passive behaviour points out the importance of IRS/fogging as a prevention tool against dengue. As results show, risk perception is a positive and significant determinant of dengue passive behaviour when bednets are excluded from the analysis, and only fogging is considered. The sensitivity analysis results showcase that interviewees correctly disentangle which preventive tools are more efficacious for each disease. In fact, while there is some evidence pointing to the efficacy of bednets in reducing the Aedes aegypti mosquito’s population, which is responsible for dengue transmission and bite during daylight time [46], bednets remain a main entomological strategy for impeding contact between humans and the mosquito responsible of malaria transmission - predominantly at night-time, the Anopheles.

Less clear-cut results were found for the remaining VBDs. ZIKV and CL were excluded from the analysis as both risk perception and behaviour were too low. For dengue, the presence of social norms in determining risk perception was weak: the relationship was highly dependent on the definition of the reference group and the dependent variable. While results show that risk perception is a relevant factor in the uptake of passive behaviours such as fogging, the risk perception seems to be explained by factors other than the risk perception of the reference group.

In the previous study that used the same data [16], the aim was to understand the contribution of disease knowledge and risk perception on passive behaviour, while no social norm dimension was included in the analysis. Analytically, a different model was estimated from the current one and all four VBDs were included in the analysis. Despite the conceptual and empirical strategy differences, individual risk perception was also found to positively and significantly contribute to passive behaviour, for all four diseases.

An important finding is that despite not engaging in active behaviours as a reaction to higher levels of (perceived) risk of infection, individuals are willing to accept or use passive measures to protect themselves. This finding should favour the government's provision of protective tools against both dengue and malaria [47]. Furthermore, given the relevance of social norms in explaining malaria risk perception, interventions such as health communication and promotion activities tailored to groups based on key demographic and social characteristics (i.e., age, ethnicity and education) should produce a replicative effect and reinforce malaria preventive behaviours. For example, information and awareness campaigns on the actual current and future infection, morbidity and mortality risk should accommodate individual characteristics in order to reinforce the uptake of distributed protective measures against malaria. On the one hand, there is evidence that crafting messages based on the beliefs and motivations of individuals are highly effective [48]. On the other hand, it has been proven that the way people interact, interpret and participate in health promotion interventions vary depending on the intersections of ethnicity and demographic variables such as age and gender [49, 50] (Liu et al, 2016; Noar et al., 2007). The findings from this study could be extrapolated to similar settings, e.g. Suriname, with comparable disease epidemiology, common socio-economic conditions (e.g. an important economic activity, such as mining, highly linked to malaria infection), and the movement of people between countries, with the associated sharing of knowledge and experiences. All in all, evidence should be generated in Guyana and similar settings on the impact on preventive behaviour for VBDs of interventions aimed to correct risk perception tailored to groups at the intersection of certain socio-economic and demographic characteristics.

Data Availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Author contributions statement

All authors contributed to the conception of the study. T.M., E.S. and I.L-R., wrote the preliminary manuscript text. E.S. and H.C. conducted the survey. TM and I.L-R. conducted the analysis and prepared figures. E.S. and I.L-R., wrote the final manuscript text. All authors commented and reviewed the manuscript and contributed to the interpretation of results.

Additional information

The authors declare no competing interests.

Athni, T. S., Shocket, M. S., Couper, L. I., Nova, N., Caldwell, I. R., Caldwell, J. M., ... & Mordecai, E. A. (2021). The influence of vector-borne disease on human history: socio-ecological mechanisms. Ecology Letters, 24(5), 829-846. https://doi.org/10.1111/ele.13675
Rocklöv, J., & Dubrow, R. (2020). Climate change: an enduring challenge for vector-borne disease prevention and control. Nature immunology, 21(5), 479–483. https://doi.org/10.1038/s41590-020-0648-y
Semenza, J. C., & Paz, S. (2021). Climate change and infectious disease in Europe: Impact, projection and adaptation. The Lancet Regional Health. Europe, 9, 100230. https://doi.org/10.1016/j.lanepe.2021.100230.
World Health Organization. (2023). Disease Outbreak News; Geographical expansion of cases of dengue and chikungunya beyond the historical areas of transmission in the Region of the Americas. Retrieved from https://www.who.int/emergencies/disease-outbreak-news/item/2023-DON448.
Bavel, J. J. V., Baicker, K., Boggio, P. S., Capraro, V., Cichocka, A., Cikara, M., Crockett, M. J., Crum, A. J., Douglas, K. M., Druckman, J. N., Drury, J., Dube, O., Ellemers, N., Finkel, E. J., Fowler, J. H., Gelfand, M., Han, S., Haslam, S. A., Jetten, J., Kitayama, S., … Willer, R. (2020). Using social and behavioural science to support COVID-19 pandemic response. Nature human behaviour, 4(5), 460–471. https://doi.org/10.1038/s41562-020-0884-z
Wise, T., Zbozinek, T. D., Michelini, G., Hagan, C. C., & Mobbs, D. (2020). Changes in risk perception and self-reported protective behaviour during the first week of the COVID-19 pandemic in the United States. Royal Society open science, 7(9), 200742. https://doi.org/10.1098/rsos.200742World Health Organization. (2020). Vector-borne diseases. Retrieved from https://www.who.int/news-room/fact-sheets/detail/vector-borne-diseases#:~:text=Vector%2Dborne%20diseases%20account%20for,either%20parasites%2C%20bacteria%20or%20viruses.
Zancu, S. A., Măirean, C., & Diaconu-Gherasim, L. R. (2022). The longitudinal relation between time perspective and preventive behaviors during the COVID-19 pandemic: The mediating role of risk perception. Current Psychology, Advance online publication. https://doi.org/10.1007/s12144-022-03069-z.
Rayner, S., & Cantor, R. (1987). How Fair Is Safe Enough? The Cultural Approach to Societal Technology Choice. Risk Analysis, 7(1), 3-9.
Sjöberg, L., Moen, B. E., & Rundmo, T. (2004). Explaining risk perception. An evaluation of the psychometric paradigm in risk perception research, Risk Analysis, 24(1), 1-12.
Floyd, D. L., Prentice-Dunn, S., & Rogers, R. W. (2000). A meta-analysis of research on protection motivation theory. Journal of Applied Social Psychology, 30(2), 407-429. https://doi.org/10.1111/j.1559-1816.2000.tb02323.x
Bults, M., Beaujean, D. J., de Zwart, O., Kok, G., van Empelen, P., van Steenbergen, J. E., Richardus, J. H., & Voeten, H. A. (2011). Perceived risk, anxiety, and behavioural responses of the general public during the early phase of the Influenza A (H1N1) pandemic in the Netherlands: results of three consecutive online surveys. BMC public health, 11, 2. https://doi.org/10.1186/1471-2458-11-2
de Zwart, O., Veldhuijzen, I. K., Richardus, J. H., & Brug, J. (2010). Monitoring of risk perceptions and correlates of precautionary behaviour related to human avian influenza during 2006 - 2007 in the Netherlands: results of seven consecutive surveys. BMC infectious diseases, 10, 114. https://doi.org/10.1186/1471-2334-10-114
Lo Presti, S., Mattavelli, G., Canessa, N., & Gianelli, C. (2022). Risk perception and behaviour during the COVID-19 pandemic: Predicting variables of compliance with lockdown measures. PLoS ONE, 17(1), e0262319. https://doi.org/10.1371/journal.pone.0262319.
Schneider, C. R., Dryhurst, S., Kerr, J., Freeman, A. L. J., Recchia, G., Spiegelhalter, D., & van der Linden, S. (2021). COVID-19 risk perception: A longitudinal analysis of its predictors and associations with health protective behaviours in the United Kingdom. Journal of Risk Research, 24(3-4), 294-313. https://doi.org/10.1080/13669877.2021.1890637
Qin, H., Sanders, C., Prasetyo, Y., Syukron, M., & Prentice, E. (2021). Exploring the dynamic relationships between risk perception and behavior in response to the Coronavirus Disease 2019 (COVID-19) outbreak. Social science & medicine (1982), 285, 114267. https://doi.org/10.1016/j.socscimed.2021.114267
Aerts, C., Revilla, M., Duval, L., Paaijmans, K., Chandrabose, J., Cox, H., & Sicuri, E. (2020). Understanding the role of disease knowledge and risk perception in shaping preventive behavior for selected vector-borne diseases in Guyana. PLoS Neglected Tropical Diseases, 14(4), e0008149. https://doi.org/10.1371/journal.pntd.0008149
Dussault, J. M., Paz-Bailey, G., Sánchez-González, L., Adams, L. E., Rodríguez, D. M., Ryff, K. R., ... & Arguin, P. M. (2022). Arbovirus risk perception as a predictor of mosquito-bite preventive behaviors in Ponce, Puerto Rico. PLOS Neglected Tropical Diseases, 16(7), e0010653. https://doi.org/10.1371/journal.pntd.0010653
Weinstein, N. D. (1980). Unrealistic optimism about future life events. Journal of Personality and Social Psychology, 39(5), 806–820. https://doi.org/10.1037/0022-3514.39.5.806
Cialdini, R. B., Reno, R. R., & Kallgren, C. A. (1990). A focus theory of normative conduct: Recycling the concept of norms to reduce littering in public places. Journal of Personality and Social Psychology, 58(6), 1015-1026. https://doi.org/10.1037/0022-3514.58.6.1015
Abrams, D., Wetherell, M., Cochrane, S., Hogg, M. A., & Turner, J. C. (1990). Knowing what to think by knowing who you are: self-categorization and the nature of norm formation, conformity and group polarization. The British journal of social psychology, 29, 97–119. https://doi.org/10.1111/j.2044-8309.1990.tb00892.x
Akerlof, G. A., & Kranton, R. E. (2005). Identity and the Economics of Organizations. Journal of Economic Perspectives, 19(1), 9-32. https://doi.org/10.1257/0895330053147930
Etilé, F. (2007). Social norms, ideal body weight, and food attitudes. Health Economics, 16(9), 945-966.
Elster, J. (1989). Social norms and economic theory. Journal of Economic Perspectives, 3, 99-117.
Collis, A., Garimella, K., Moehring, A., et al. (2022). Global survey on COVID-19 beliefs, behaviours and norms. Nature Human Behaviour, 6(9), 1310-1317. https://doi.org/10.1038/s41562-022-
Hensel, L., Witte, M., Caria, A. S., Fetzer, T., Fiorin, S., Götz, F. M., Gomez, M., Haushofer, J., Ivchenko, A., Kraft-Todd, G., Reutskaja, E., Roth, C., Yoeli, E., & Jachimowicz, J. M. (2022). Global Behaviors, Perceptions, and the Emergence of Social Norms at the Onset of the COVID-19 Pandemic. Journal of economic behavior & organization, 193, 473–496. https://doi.org/10.1016/j.jebo.2021.11.015
Akfırat, S., Bayrak, F., Üzümçeker, E., Ergiyen, T., Yurtbakan, T., & Uysal, M. S. (2023). The roles of social norms and leadership in health communication in the context of COVID-19. Social science & medicine, 323, 115868. https://doi.org/10.1016/j.socscimed.2023.115868
Gelfand, M. J., & Harrington, J. R. (2015). The motivational force of descriptive norms: For whom and when are descriptive norms most predictive of behavior? Journal of Cross-Cultural Psychology, 46(10), 1273–1278. https://doi.org/10.1177/0022022115600796
De Salazar, P. M., Cox, H., Imhoff, H., Alexandre, J. S. F., & Buckee, C. O. (2021). The association between gold mining and malaria in Guyana: A statistical inference and time-series analysis. The Lancet Planetary Health, 5(10), e731-e738. https://doi.org/10.1016/S2542-5196(21)00203-5
Douine, M., Bonifay, T., Lambert, Y., Mutricy, L., Galindo, M. S., Godin, A., ... & Djossou, F. (2022). Zoonoses and gold mining: A cross-sectional study to assess yellow fever immunization, Q fever, leptospirosis, and leishmaniasis among the population working on illegal mining camps in French Guiana. PLOS Neglected Tropical Diseases, 16(8), e0010326. https://doi.org/10.1371/journal.pntd.0010326
Freitas, L. P., Carabali, M., Yuan, M., Jaramillo-Ramírez, G. I., Balaguera, C. G., Restrepo, B. N., ... & Cardenas, J. C. (2022). Spatio-temporal clusters and patterns of spread of dengue, chikungunya, and Zika in Colombia. PLOS Neglected Tropical Diseases, 16(8), e0010334. https://doi.org/10.1371/journal.pntd.0010334
Manski, C.F. (1993). Identification of endogenous social effects: the reflection problem. The Review of Economic Studies, 60, 531–542.
Caroli, E., & Weber-Baghdiguian, L. (2016). Self-reported health and gender: The role of social norms. Social Science & Medicine, 153, 220-229.
McPherson, M., Smith-Lovin, L., & Cook, J.M. (2001). Birds of a Feather: Homophily in Social Networks. Annual Review of Sociology, 27(1), 415–444.
Smith, J. A., McPherson, M., & Smith-Lovin, L. (2014). Social Distance in the United States: Sex, Race, Religion, Age, and Education Homophily among Confidants, 1985 to 2004. American Sociological Review, 79(3), 432–456.
Charness, G. & Chen, Y. (2020). Social Identity, Group Behavior, and Teams. Annual Review of Economics, 12(1), 691-713. https://doi.org/10.1146/annurev-economics-091619-032800
Fischer, H. G., & Molenaar, I. W. (1995). Rasch models: Foundations, recent developments, and applications. Springer.
Matera, M., Sandin, L., & Alvarez, M. (2020). An Assessment of the State of the Guyanese Diaspora. In The Guyanese Diaspora (pp. 3–16). Center for Strategic and International Studies (CSIS). http://www.jstor.org/stable/resrep26538.4
Tarozzi, A., Mahajan, A., Blackburn, B., Kopf, D., Krishnan, L., & Yoong, J. (2014). Micro-loans, Insecticide-Treated Bednets, and Malaria: Evidence from a Randomized Controlled Trial in Orissa, India. American Economic Review, 104(7), 1909-1941.
Apouey, B., & Picone, G. (2014). Social interactions and malaria preventive behaviors in sub-Saharan Africa. Health Economics, 23(9), 994-1012. https://doi.org/10.1002/hec.3055
Picone, G., Kibler, R., & Apouey, B. H. (2017). Malaria Prevalence, Indoor Residual Spraying, and Insecticide-Treated Net Usage in Sub-Saharan Africa. Journal of African Development, 19, 19–32.
Blumberg, L., & Frean, J. (2017). Malaria reduces globally but rebounds across southern Africa. Southern African Journal of Infectious Diseases, 32(2).
Pan American Health Organization, World Health Organization (WHO). (2017). Zika-Epidemiological Report. Guyana [Internet] (Vol. 2). Retrieved from https://www.paho.org/hq/dmdocuments/2017/2017-phe-zika-situation-report-guy.pdf
Pan American Health Organization, World Health Organization (WHO). (2017). Cutaneous and mucosal Leishmaniasis 2017. Guyana [Internet]. Retrieved from http://www.panaftosa.org/leish/inf2017_en/INFO_GUY_2017_ENG.pdf
Caribbean Public Health Agency. Data and Publications. CARPHA Member States Health Data Repository. Retrieved from http://carpha.org/Data-and-Publications/-Data-and-Statistics
Abdallah, R., Louzada, J., Carlson, C., et al. (2022). Cross-border malaria in the triple border region between Brazil, Venezuela and Guyana. Scientific Reports, 12, 1200. https://doi.org/10.1038/s41598-022-05205-y
Lenhart, A., Orelus, N., Maskill, R., Alexander, N., Streit, T., & McCall, P. J. (2008). Insecticide-treated bednets to control dengue vectors: Preliminary evidence from a controlled trial in Haiti. Tropical Medicine & International Health, 13(1), 56-67. https://doi.org/10.1111/j.1365-3156.2007.01966.x
Wilson, A. L., Courtenay, O., Kelly-Hope, L. A., Scott, T. W., Takken, W., Torr, S. J., et al. (2020). The importance of vector control for the control and elimination of vector-borne diseases. PLoS Neglected Tropical Diseases, 14(1), e0007831.
Lustria, M. (2017). Message Tailoring in Health and Risk Messaging. Oxford Research Encyclopedia of Communication. Retrieved May 26, 2023, from https://oxfordre.com/communication/view/10.1093/acrefore/9780190228613.001.0001/acrefore-9780190228613-e-323.
Liu, J. J., Davidson, E., Bhopal, R. S., & White, M. (2016). Adapting health promotion interventions for ethnic minority groups: a qualitative study. Health Promotion International, 31(2), 325–334. https://doi.org/10.1093/heapro/dau105.
Noar, S. M., Benac, C. N., & Harris, M. S. (2007). Does tailoring matter? Meta-analytic review of tailored print health behavior change interventions. Psychological bulletin, 133(4), 673–693. https://doi.org/10.1037/0033-2909.133.4.673

No competing interests reported.

SCIRep30062023supplementary.docx

Download PDF

Journal Publication

published 06 Oct, 2023

Read the published version in Scientific Reports →

Editorial decision: Major revision
25 Aug, 2023
Reviews received at journal
29 Jul, 2023
Reviewers agreed at journal
28 Jul, 2023
Reviewers invited by journal
27 Jul, 2023
Editor assigned by journal
27 Jul, 2023
Editor invited by journal
27 Jul, 2023
Submission checks completed at journal
27 Jul, 2023
First submitted to journal
30 Jun, 2023

You are reading this latest preprint version

The contribution of risk perception and social norms to reported preventive behaviour against selected vector-borne diseases in Guyana

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Methods

Ethics statement

Data and outcome variables

Empirical strategy

Robustness checks

Results

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1