Why are people with dengue dying? A scoping review of determinants for dengue mortality

Age was described as a determinant related to immunological status and type of infection [1, 8, 9, 107]. Interestingly, despite the fact that reported mortality rates were higher in adults, children’s cases seemed to be more sensitive from a sociocultural perspective and generated greater concern. Moreover, fatal outcomes in children were described as also being dependent on health staff and parents’ awareness of the disease, implying that children’s outcomes depend not only on age, but also on adequate management by health staff and opportune care-seeking behavior by parents [29, 37, 38, 93, 107]. Likewise, ethnicity was considered to be a determinant for dengue mortality despite the known biological implications. For instance, while some papers have posited African ancestry or being ‘Black’ as a protective factor for severe forms and dengue mortality [107, 109‐112], this review found a paper from Brazil contradicting that position [88]. We observed that when the protective role of African ancestry was demonstrated in other studies, socioeconomic information was not taken into account, and that when this information was considered as a possible risk factor, the social aspect appeared as a co-determinant. This could be related to the fact that people of African ancestry in the Americas tend to live in unfavorable socioeconomic conditions [88]. Hence, it is not the fact of being ‘Black’ in Brazil that increases the chances of dying from dengue, but rather living in unfavorable socioeconomic conditions, as presented by Blanton et al. in 2008 [109]. When both ethnicity and income were addressed, and even after controlling for income, African ancestry (both self-reported and genotyped) was a protective factor for severe forms of dengue. As such, ethnicity would not be a determinant on its own but is linked with socioeconomic position and opportunity for access to qualified services [18, 19, 88].

Education is a determinant with a twofold effect on dengue mortality. First, educated people will understand the importance of the disease and its risks, seek care opportunely and adhere to treatment [1, 17, 88]. Second, educated health staff will manage patients accurately, thereby decreasing their chances of developing severe forms, which in turn reduces dengue mortality [32, 91, 93, 107]. Furthermore, the level of education (and awareness) of individuals could be a proxy for poverty (and poor quality of facilities), since it is related to the quality of life of individuals and societies and to the quality of the health system [16‐19].

Type of dengue infection, virus characteristics and immunological status of the individual are known biological determinants of severe forms and fatal outcomes of dengue [3, 6, 8, 9, 13, 33, 37, 67, 107, 110, 113]. However, it is known that individuals’ immunological status is also related to socioeconomic status (SES) [19, 104, 105]. Moreover, in the case of dengue, type of infection and immunological status will depend on previous exposure to the virus. Acting as a selective pressure, previous exposure induces changes in the virus and allows it to remain in a determinate region, leading to endemicity [1, 6, 7, 9, 19, 23]. As such, those determinants, along with their biological implications, could also be considered an outcome of the socioeconomic and environmental conditions in which people are living.

Poverty and social behavior were considered to be general (wide-ranging) social dimension determinants for dengue mortality, related to income, education and health systems. Lack of economic resources is often associated with the presence of dengue [1, 6, 7, 19, 109]. Poverty will also have an impact on access to health care, since in many developing countries access is determined by capacity to pay [16, 18]. Likewise, poverty could limit access to care in remote areas [17‐19]. It is also known that lack of economic resources may lead to fatal outcomes and have devastating economic implications [11, 21, 87, 92, 94]. Furthermore, society is moved by cultural or idiosyncratic patterns, whether hierarchical, economic, power or gender-oriented, that affect people’s health status [16‐19, 104]. Some documents in this review observed that care-seeking behavior may determine the possibility of receiving opportune treatment and thereby avoiding fatal outcomes [29, 30, 76, 77, 80, 93]. A relationship was also observed between care-seeking behavior patterns and gender, in accordance with societal cultural behaviors [18, 19, 35, 93, 94].

The health systems determinants related to dengue mortality were access to health care, opportunity and quality of care, and health staff knowledge. Indeed, health systems play a determinant role in the health outcomes of people, as mentioned by the CSDH [16‐18]. Access to health care is nevertheless mediated by other SDHs, such as income and poverty [16, 18‐20, 104, 114], and it is therefore essential to recognize that limited access to health care may contribute to the development of dengue fatal outcomes [3, 19, 21, 30‐32, 40, 92]. In addition, opportunity and quality of care and health staff knowledge are determinants related to education in several ways. In the case of opportunity for care, empowered and disease-aware individuals tend to seek care opportunely [19] and knowledgeable staff tend to offer opportune management to dengue patients [18, 19]. When care is not offered opportunely, the risk of dengue fatal outcomes is theoretically doubled [27, 31, 55, 61, 80, 88, 93, 102]. As for quality of care and health staff knowledge, trained personnel translate into better case diagnosis and management, and consequently low mortality rates [32, 91, 93, 107, 114]. Therefore, the above-mentioned health systems related determinants must figure prominently in any strategy targeting a decrease in dengue mortality.

Although sex/gender was not considered a determinant for dengue mortality in this review, some authors observed that its role in fatal outcomes depended on care-seeking behavior, which in turn depended on sociocultural aspects [27, 88, 94]. As gender, being socially constructed, is differentiated from sex, and since gender roles interact in different forms with NTDs [18, 19], further research on this topic would be interesting. In the case of dengue, it might not be sex, but rather care-seeking behavior driven by gender, that determines the outcome [4, 88]. With regard to the presence of comorbidities or pre-existing conditions, the information provided in the reviewed documents was not specific and therefore not conclusive enough to consider it a determinant for dengue mortality. Nonetheless, it is important to acknowledge that comorbidities might aggravate the disease, as has already been stated [8, 12, 110]. Likewise, although income and occupation were not described as determinants, these two variables are related with socioeconomic status and will depend on education as well. Hence they are considered social determinants in relation to many infectious diseases, including dengue [18, 19, 109].

Very few documents provided information on environmental factors, and neither the few environmental factors described nor vector presence, were considered direct determinants for dengue mortality. Aspects such as urban residence or level of urbanization were reported only when describing dengue cases but not when informing about fatal cases. Nonetheless, vector presence could depend on virus presence and social behaviors (community or government actions). Moreover, vector control is the current feasible strategy for dengue control while waiting for vaccine development [1, 3, 6, 9, 13, 14]. Neither health care coverage nor hospital length of stay was conclusive as a determinant for dengue mortality. Opportunity for care by health staff might be related to coverage and access to health care, and hospital length of stay might be related to quality of care and health staff knowledge. These conditions, along with the presence of a surveillance system, are useful aspects to consider when analyzing the presence of dengue and its outcomes.

This review’s inclusiveness and the comprehensiveness of the analysis are major strengths. Including a great variety of documents and conducting a review in four languages helped reduce selection bias and increased the scope of the information in this type of review [22]. However, the fact that we did not include documents written in native languages from the Asia region (Thai, Vietnamese, Indonesian, etc.) could have limited the number of documents reviewed from that region. Likewise, it is possible that the Americas region was over-represented by the fact that the review included documents written in Spanish and Portuguese. A summative content analysis [25] of dengue mortality documents provided comprehensive and detailed information on the topic over the past 16 years. It is important to acknowledge that, despite the substantial dengue literature, information on fatal cases is still limited and is mostly of a biomedical nature. Moreover, access to some documents was not possible. An added challenge is the fact that it is not always possible to distinguish between SDHs and biological factors, nor to identify their determinant roles separately [18, 19, 104]. Notwithstanding the above-mentioned limitations, this scoping review offers valuable information on the role of SDHs in dengue mortality.