Our case-control study revealed that age and living area were associated with risk for dengue infection in the Hanoi metropolitan city. The results showed that older people had a lower risk for dengue infection as compared to younger people. In particular, people aged 31–45; 46–60 and above 60 years had 57%, 62% and 95% lower risk, respectively, to get DF than patients aged 16–30. Those results are similar to earlier studies: People aged 15–34 years has been found to be the most infected age group in the dengue outbreaks in Hanoi city in particular and in Vietnam and Singapore in general (20) (21) (22) (23). This could be explained by the fact that the older people tend to pay attention on their health and have better health protective measures as compared to young adults, e.g. by sleeping under bed-nets at all time during the day. Also, it is possible that a larger proportion of the older people had been infected by dengue in the past and were immune. Immunity would mainly decrease the probability of getting the same serotypes of dengue virus (20). In addition, young people may spend much more time on out-door activities, leading to higher risk of exposing with outside mosquitoes. Nevertheless, an earlier study in Singapore showed that the incidence rate for dengue was highest in the age group of 55 years and above, which is contradictory to our result (23). An earlier systematic review and meta-analysis also revealed that the mean age of dengue patients reported after 2010 tends to be higher as compared to dengue cases reported before 2010 (24). Therefore, it is suggested that health information, education and communication (IEC) program on dengue prevention and control should be disseminated to all age groups but focused on the adolescents and young adults.
Our study further indicated that people living in central districts have a 3.2 times higher risk to get an infection by dengue virus, as compared to people living in peri-urban districts. This is consistent with the epidemiological findings of Duong et al. (25) where 77.2% of the total dengue cases between 2006 and 2011 were concentrated to urban areas of Hanoi. Studies by Toan et al. (11) and Cuong et al. (20) showed similar results in that most cases of dengue were found in the inner districts of Hanoi. In Ho Chi Minh city, another big metropolitan area of Southern Vietnam, Raghwani et al. (26) also found that the more densely populated inner districts contributed significantly to DENV-1 transmission as compared to the suburban districts. Hanoi city is an economic center of Vietnam where migrants are populous. A lot of students and labor people from other provinces have been entering the central districts of Hanoi for living and working (27). In addition to the low awareness on dengue prevention and control and limited resources, they may live in more unhygienic conditions that may increase the probability of creating breeding sites for dengue mosquitoes. This implies the higher risk of having DF and spreading the disease to neighboring people (20). Our study also showed that students and office employees accounted for the highest percentage amongst the dengue patients. This result was in line with other studies conducted in Hanoi (20) (28). Globally, 50% of the dengue outbreaks during 1990–2015 were recorded in urban areas, followed by 28.6% in rural areas, and 21.4% in both urban and rural areas (24). Hence, living in urban areas is one of the driver of dengue dispersion due to the urbanization and huge population growth in the metropolitan regions (29), although it is reported that there was a movement of dengue morbidity from urban to rural settings (6). The recommendation is that more communication campaigns should be organized in the central urban districts, targeted towards specific groups of landlords and their tenants.
Regarding other possible risk factors, this study suggested that gender was not a risk factor of having DF. However, it is noticeable that a study of Lien et al. (21) which was also carried out in Hanoi city, showed that males accounted for the majority of the dengue case in the 2011 outbreak. Studies by Guo et al. and Ler et al. also depicted that the incidence of getting dengue infection was significantly higher in males than in females (24) (23). Therefore, it is hard to draw a conclusion on this association. It could also be that knowledge and education differs between gender, but this study had too low power to investigate that. KAP on dengue prevention and control are vital to measure the risk of getting DF. Interestingly, our findings indicated that individuals with dengue infection had better mean score of KAP as compared to the control group. This could be explained by the higher education in the dengue case group; or dengue patients may have improved their knowledge before the study started through the information by doctors, nurses or internet. This study could not show any association between KAP score and dengue status, which was not in accordance with other studies in Vietnam (30); in Malaysia (31) (32) and in Brazil (33), where the results confirmed a correlation between dengue infection and lacking of preventive measures such as not keeping the house and environment clean, storing water with uncovered containers, not wearing long sleeve clothes, or not using screen windows. In the era of information technology, it is easy to search for health information related to DF which is the most common mosquito-borne disease in Hanoi and Vietnam. In a KAP cross sectional study conducted in the same site in Hanoi during the 2017 dengue outbreak, the mean knowledge score of participants was 4.6 out of 19 points; lower than the mean score obtained in our study (7.9/19) (34). It is understandable that people in Hanoi at present have a better knowledge on dengue prevention and control as compared to earlier studies (28). However, knowledge alone does not produce individual behavior change (4). In routine life, dwellers have numerous concerns rather than doing preventive measures for mosquito-borne diseases. Previous studies have proved that the behavior of water storage in uncovered containers was a high-risk factor for Aedes breeding (24) (33) (35). Nevertheless, our study did not find any links between storing water without lid or in the buckets/barrels at home with the probability of being infected by dengue virus in the two patient groups. Since Hanoi is a capital city with a considerate speed of urbanization, the tap water system helps dwellers to change their storage habit and limit the mosquito breeding sites (20). This may explain why few patients in our study had that behavior. However, Thang et al (28) assumed that the breeding sites of mosquito could include public places like cemeteries with empty vases on the graves, temples and pagodas with many vases, the Bonsai or construction projects, and abandoned houses with stored garbage. Thus, our implication is that repeated messages on dengue prevention and control focusing on personal protections and environmental clean-up activities should be implemented at various time points of the year, not only when the dengue season starts in July.
In our study, livestock keeping was considered as a risk factor, as this has been found to be contributing to the risks of several vector-borne diseases (36)(37). However, our results revealed no association between livestock keeping and dengue infections, perhaps due to the low number of individuals keeping livestock in this study. This finding was also demonstrated similarly in another study implemented in Hanoi city (16). Further studies should be deployed to explore this issue in more detail, since urban livestock keeping is popular in many developing countries.
One strength of our study is that we used Cronbach alpha to determine the internal validity and reliability of KAP items in the questionnaire. The Cronbach alpha scores demonstrated satisfactory internal consistencies in our study. To our knowledge, this is first case-control study included all KAP components other than practice only to identify the risk factors of DF. However, this study still had several limitations that needs to be considered. Firstly, we could not match case and control patients which could lead to the confounder and interaction of some variables including age, gender, education level and knowledge. However, we statistically controlled for possible confounders in the logistic regression model. Secondly, sampling was based on the selection of doctors at the Department of Infectious Diseases, and it is possible that this caused some selection bias. In order to minimize the bias, our study employed the updated national case definition of Vietnam Ministry of Health that all doctors in the Department of Infectious Disease were capable of and experienced in diagnosing the dengue case. Our research group also had a clearly documented selection procedure, and this was explained to all doctors before the start of the study. Thirdly, the patients may not be representative of the population of Hanoi because we excluded the children from 1–15 years old who is a vulnerable group suffering from DF. We only carried out the study in one hospital of Hanoi so that we could not infer these findings to the whole city’s population. Fourthly, the patients’ living place was not directly observed, leading to bias in their answers in some housing and environmental factors. Nevertheless, the interviewers were well trained and had experiences in data collection to get the validity of all participants’ responses.
In conclusion, our study found that younger adults aged 16–30 and people living in central urban areas have higher risk of getting dengue infection than older people and those living in peri-urban areas. KAP on dengue prevention and control and other demographic, environmental, housing variables were not related factors of dengue infection. Any IEC program on dengue prevention and control should be focused on specific groups of adolescents, younger adults, landlords and migrants and implemented many times of each year to improve the KAP of citizens.