In the developing and under-developed countries, contagious diseases are main indicating factors of poor health status and poverty (Kumar et al., 2007). The recent Chikungunya viral epidemic in Bangladesh affected the strength and ability of working populations to carry on with regular activities which resulted in damaging socioeconomic back grounds. Chikungunya is a mosquito-borne (Aedes species) disease caused by Chikungunya virus (CHIKV), which is transmitted through the bite of an infected mosquito. It creates an acute infection of sudden onset, characterized by high fever, headache, joint swelling, photophobia, rash and retro-orbital pain (Abedin et al., 2021; Hossain et al., 2018). Chikungunya virus was first identified in 1953, in Tanzania (Goh et al., 2015). After that, Chikungunya disease caused many large-scale outbreaks in India, Southeast Asia, Western Pacific, America, and Africa (Burt et al., 2017; Nunes et al., 2015; Thiberville et al., 2013). In Bangladesh, Chikungunya was first recog-nized and reported in 2008 (Icddr, 2009). In 2011 and 2012, rural Bangladesh experienced  outbreaks of Chi-kungunya (Grainger, 2008; Khatun et al., 2015). 

The peak time of Chikungunya outbreak as identified by IEDCR is between early May and end of July (Hossain et al., 2018). Capital city Dhaka had ex-perienced a large-scale Chikungunya outbreak in 2017. The main objective of this study was to determine the socio-economic impact of Chikungunya in Dhaka City. The socio-economic impact was considered through the absence in working place of the affected people and percentage of monthly family income spends to monthly family expenditure due to Chikungunya dis-ease. The risk factors of disease were also identified. Various suggestions for knocking the disease off had been analyzed and then a concluding remark ended the study.

For the study, primary data was collected; a ques-tionnaire was designed cautiously with respect to study objectives. Direct interview (face-to-face) method was used in this study, the study covered a total of 272 Chikungunya affected and 272 unaffected respondents. Data on economic variables were collected from only the earning members of the   families. The question-naire covered eight major sections: personal (demo-graphic) information of the respondents, economic condition, environmental condition, disease incidence, treatment, preventive measure, knowledge on Chikun-gunya, suggestions.

Two most popular regression models, multiple linear regression and binary logistic regression, had been employed. When there is more than one covariate and the response variable is of continuous type having linear relation to the covariates and follows some other specific assumptions then multiple linear regression is used (Nathans et al., 2012). For the vector of response variable Y and vector of covariatesx, the functional form of the model in matrix notation becomes (Brown, 2009)

Y= x^ β+ ϵ,

Where β and ϵ be the vector of regression coefficient and random error term, respectively. Now, if the res-ponse vector Y be of binary type i.e. referring to whether an event has occurred or not, binary logistic regression is used for modeling purpose, which has the form (Sarkar & Midi, 2010)

π(x)=e^(x^ β)/(1+e^(x^ β) )

Where  π(x) represents the conditional mean of Y given x i.e., E(Y| x). For both model, the unknown parameters (β) are estimated by the method of maxi-mum  likelihood estimation (Albert & Anderson, 1984; Myung, 2003). The former model has been used for relating family cost and productivity to Chikungunya incidence, while the factors behind occurring Chikun-gunya disease was identified by the later model. Also, multiple response analysis had been incorporated in this paper to assess various suggestions from the study participants to reduce the disease in future. 

Model 1

The first model was a multiple linear regression model that determined, primarily, the effect of Chikungunya infection on the monthly expenditure of the patient. Here, dependent variable family cost [percentage of monthly income spent on expenditure] was used as a proxy for the measure of monthly expenditure of the patient. The main independent variable was Chikun-gunya incidence and some other independent variables related to spend on expenditure were considered in this model. The model took the form as –

〖family cost 〗_i= β_0+ β_1 〖incidence〗_i+ β_2 〖living status〗_i+ β_3 〖family size〗_i+ β_4 〖marital status〗_i+ β_5 〖age〗_i+ β_6 〖gender〗_i+ β_7 〖treatment cost〗_i+ ϵ_i,  

i=1,2,⋯,n;

Where β denotes regression coefficient and ϵ_i is the random error term of the i^th respondent. 

Model 2

The second model was again a multiple linear reg-ression model that was used to determine the effect of Chikungunya infection on the productivity of the patient. Here, dependent variable absence [number of days absent from work] was used as a proxy for the measure of productivity of the patient. The same independent variables from model 1 were accom-modated in this model, where the main independent variable was Chikungunya incidence. The model had the expression as

〖absence 〗_i= β_0+ β_1 〖incidence〗_i+ β_2 〖living status〗_i+ β_3 〖family size〗_i+ β_4 〖marital status〗_i+ β_5 〖age〗_i+ β_6 〖gender〗_i+ β_7 〖treatment cost〗_i+ ϵ_i, 

 i=1,2,⋯,n;

Where β and ϵ_i bear the same meaning as before. 

Model 3

Binary logistic regression model had been used to identify the factors which were responsible for occur-ring Chikungunya disease. The acting response variable in this model was incidence [whether Chikungunya disease occurred or not], which was of binary type. The relationship between these response variable and related covariates took the functional form as –

¬¬¬¬¬¬¬¬¬

〖incidence 〗_i= β_0+ β_1 〖providing spray〗_i+ β_2 〖water existence〗_i+ β_3 〖area〗_i+ β_4 〖gender〗_i+ β_5 〖age〗_i+ β_6 〖education status〗_i+ β_7 〖marital status〗_i+β_8 〖living status〗_i+β_9 〖drainage system〗_i+ ϵ_i,  

i=1,2,⋯,n;

Where  β and ϵ_i bear the same meaning as before. 

Multiple linear regression and binary logistic regression models had been employed in this paper for analyzing Chikungunya patients data in Dhaka city. Family cost related factors are presented in Table 1. The obtained significant variables are living status, family size, age, gender, and treatment cost, while incidence and marital status are insignificant at 5% level of significance.

Table 1: Multiple linear regression model estimates of the selected covariates for family cost in Dhaka city along with standard error (SE) and p-value.

Living “with others” increases the percentage of monthly income spent on expenditure by 13.055 per-cent compared to living “alone” keeping all other co-variates constant. Living with family and friends has no significant effect compared to living alone at 5% level of significance. Extra one person increase of family member increases expenditure by 1.377 percent. Per-centage of monthly income spent on expenditure increases by 6.218 percent in 15-30 age group than0-15 age groups keeping all other independent variables constant. Expenditure decreases by 3.657 percent in the female patient group compared to the male patient group keeping all other independent variables constant.  For one-unit increase in treatment cost, Percentage of monthly income spent on expenditure increases by 0.001 percent, keeping all other variables constant. So, treatment cost has significance here. 

Table 2: Multiple linear regression model estimates of the selected covariates for productivity in Dhaka city along with standard error (SE) and p-value.

The covariates which affect productivity of the res-pondents are shown in Table 2, it reflects that the covariates incidence, living status, and age are signi-ficantly affecting productivity of the respondents. In-cidence has a significant effect on productivity with p-value less than 0.001. Chikungunya incidence hampers approximately 5 working days, keeping all other vari-ables constant. Patients living with friends/colleagues and others remain approximately 3 and 5 days respect-ively, less absent from workplace than patients living alone, keeping all other variables constant. So, living status has significance in the model. Patients having age group 15-30, 30-45 and 45-60 remain approxi-mately 3 days on an average less absent from work-place than 0-15 aged patients, keeping all other vari-ables constant. So, age is a significant factor to the productivity of the respondents. The variables family size, marital status, gender, and treatment cost do not affect productivity significantly. 

The potential determinants that are acting to occur Chikungunya disease are shown in Table 3. The vari-ables marital status, living status, providing spray, and area are found significant; while gender, age, education status, and drainage system are insignificant in the model. 

Table 3: Binary logistic regression model estimates of the selected covariates for chikungunya incidence in Dhaka city along with standard error (SE), hazard ratio (HR), and p-value.

Multiple response analysis is a frequency analysis, when there can be more than one response from each participant of a survey question. Multiple responses are arranged into three portions: analysis of preventive approach of the patients, analysis of suggestions about peoples steps on prevention of Chikungunya, and ana-lysis of suggestions about governments steps on pre-vention of Chikungunya. Table 4 shows that 7.5% of patients who are affected by Chikungunya dont use any approach such as coil, aerosol, liquid, mosquito racket, net, etc. 34.0% of patients use coil, 37.4% of patients use aerosol, 20.8% of patients use liquid, 23.8% of patients use Mosquito racket and 52.1% of the patients use net as a preventive approach to protect themselves from Chikungunya. 

Table 4: Multiple response analysis of used preventive approach of the Chikungunya affected respondents in Dhaka city.

It is observed from Table 5 that 6.5%  people have no suggestions to prevent Chikungunya fever, 35.4% people prefer that they can cover tightly all water con-tainers, 39.5%  people have opinion that they can bury unused tiers, 58.2% people want to keep drain free from blockage, 58.6% people said that they can change water in plant container, 41.4% people prefer to change  water in trays under the fridge, 35.4% people said that they can destroy/ burn unused containers, 63.9% people also said that they can use mosquito repellent and mosquito net when sleeping and also further 23.6% of people prefer to take others steps to prevent Chikun-gunya fever transmission during outbreak. 

Table 5: Multiple response analysis of the suggestions regarding necessary steps that people might take to prevent Chikungunya disease.

From Table 6, 44.5% people think that government can repair open manhole, 62% people want that govern-ment can develop the drainage system, 51.3% people prefer that government can remove blocked water source, 64.6% people want that government can pro-vide regular mosquito spray, and 48.3% people think that government can keep the locality clean.  

Table 6: Multiple response analysis of the suggestions regarding necessary steps that government might take to prevent Chikungunya disease.

This study initially measures the impact of Chikun-gunya both on society and economy of Dhaka city.  It is found that occurrence of Chikungunya has no signi-ficant effect on family expenditure. It may be the case that the treatment of Chikungunya is not that expensive to affect the total family cost. But, this result con-tradicts a previous work in India (Gopalan & Das, 2009; Vijayakumar et al., 2013). However, Chikun-gunya badly affects productivity of the patients, which is similar to a previous work on this disease (Gopalan & Das, 2009). So, Chikungunya has no direct economic impact in a family. But, as it affects productivity, it affects the national economy to a great extent. Again, living status and age have significant effect on pro-ductivity, while family size, gender, treatment cost, age, living status is significant to the family cost of the respondents. Family size, gender, treatment cost, and marital status have no significance to productivity, while incidence of Chikungunya fever, marital status, and living status are insignificant to the family cost of the respondents (Shazeed-Ul-Karim, 2019).

People living in slum areas are at higher risk of having Chikungunya disease, which might be happen due to the unawareness of the residents and the dirty envi-ronment of these areas that grows and reserves mos-quitoes (Rashid & Zzaman, 2017). The chance of being affected varies significantly by marital status (Panato et al., 2019). The disease is more likely to occur to the single people. Similarly for the variable living status, it shows high risk of Chikungunya incidence in the alone living group (Gérardin et al., 2014). It might be case that those who are single and/or living alone dont take enough protection against mosquito bite, as a result, experience the disease.  An unexpected result is found for providing mosquito spray. The result says that pro-viding mosquito spray increases the risk of Chikun-gunya incidence which contradicts previous literature (Lahariya & Pradhan, 2006). But, in multiple response analysis, majority respondents have suggested to provide regular mosquito spray. It may happen that the hidden mosquitoes come out into houses from bushes and drains, which indirectly increasing the disease. From the multiple response part, it is found that few patients use no preventive approach, while majority use mosquito net as use preventive approach. It is an effec-tive practice to using mosquito net as in literature (Yazdani & Kaushik, 2007). To the respondents, they suggest themselves mostly to keep drain free from blockage, regular changing water from plant container, and use mosquito repellent and net. They seek from the government mostly to provide regular mosquito spray, develop drainage system, and remove blocked water source. Chikungunya disease had no direct effect on family cost, though it affected productivity to a great extent. 

First and foremost, the author is grateful to Almighty Allah.  The  author  is  also  thankful  to  anonymous reviewers  and  editors  for  their  helpful  comments  and suggestions. 

The author declares no conflict of interest.