1. **Merging individual data with Household data**

set maxvar     10000  , permanently

open lasi_individual data

merge m:1 hhid using “lasi_household_file_path” 


  Result                           # of obs.
    -----------------------------------------
    not matched                         2,212
        from master                     1,116 (_merge==1)
        from using                      1,096  (_merge==2)

    matched                            71,134  (_merge==3)
    -----------------------------------------

tab _merge
keep if _merge==3
drop _merge
save the merged individual_household file.

                              2. Response Variable

Recoding Outcome variable (Dengue & CHIKV)**

*   gen dengue_R=.
    
 replace dengue_R=1 if ht209a==1
   
  replace dengue_R=0 if ht209==2 | ht209a==2

 * gen chikun_R=.
                 
 replace chikun_R=1 if ht208a==1
                
 replace chikun_R =0 if ht208==2 | ht208a==2


                             3. Recode variables (SES variables)**

*Age
recode dm005  (min/54=0 "45-54 years") (55/69=1 "55-69 years") (70/max=2 "70+ years"), gen(Age_Group)

*Residence
recode residence (1=0 "Rural") (2=1 "Urban"), gen(Residence_R)

*Sex
recode dm003 (1=1 "Male") (2=0 "Female"), gen(Sex)

*MPCE
recode mpce_quintile (1=0 "poorest") (2=1 "poorer") (3=2 "middle") (4=3 "richer") (5=4 "richest"),gen(MPCE_R)

*School Years
recode dm007 (.=0 "No school years") (1/5=1 "1-5 school years") (6/12=2 "6-12 school years") (13/26=3 "college") (.d=.), gen(Education)

*Caste
recode dm013 (1=1 "SC") (2=0 "ST") (3=2 "OBC") (4=3 "forward"), gen(Caste)

*Work
recode we016_mainjob (1/2=0 "Agricultural and Allied") (3/6=1 "self-employed/ wage")  (.=3 "Never/past work"), gen(Occupation)


                        4.Recode Variables (Housing conditions)**

*household_size 
recode household_size (1/5=0 "1-5" ) (6/max=1 "6+") , gen(HH_Size)

*House_Type
recode he024 (1/2=0 "pucca/semi pucaa")  (3=1 "Kutcha"), gen(House_Type)

*Drinking_Water_Location
recode he007 (1=0 "own dwelling") (2=1 "own yard/plot") (3=2 "outside dwelling"), gen(Drinking_Water_Location)

*Toilet_Sanitation

recode he004 (1=0 "improved sanitation") (2/4=1 "unimproved") (5=2 "open depecation") , gen(Toilet_Sanitation)

*Fuel_Type

recode he014 (1 2 4=0 "clean") (3 5 6 7 8 9 10=1 "unclean"), gen(Cooking_Fuel)

*Damp wall/ceiling

recode he021s4 (0=0 "no") (1=1 "yes"), gen(wall_ceiling_damp)

                           
                   5. Prevalence of dengue and CHIKV based on SES AND Household variables**

*dengue
 For eg: tab Education dengue_R [aw=indiaindividualweight],r ;for all the variables.
 
*chikungunya
For eg: tab Education chikun_R [aw=indiaindividualweight],r ;for all the variables.

                6.**Prevalence of dengue and CHIV based on states (For GIS mapping)**

              tab stateid dengue_R [aw=stateindividualweight],r
              tab stateid chikun_R [aw=stateindividualweight],r




                    7.**Univariable analysis**

logistic dengue_R  i.Age_Group [pw= indiaindividualweight] and same for all other SES variables

Ex: logistic dengue_R  i.Age_Group   [pw= indiaindividualweight], only variables with <0.25 significance taken for next analysis

logistic chikun_R  i.Age_Group [pw= indiaindividualweight] and same for all other SES variables

Ex: logistic dengue_R  i.Age_Group   [pw= indiaindividualweight], only variables with <0.25 significance taken for next analysis
 
                  8. **Multiple logistic regression**

* logistic deng_R  i.Age_Group i.Residence_R i.Sex i.MPCE_R i.Education i.Cooking Fuel i.House_type i.Occupation [pw=indiaindividualweight], only variables with <0.10 significance 

 retained;variables with >0.10 significance added one at a time; Occupation was retained in the model (>15% change in the odds ratio of other variables)

Variables >0.25 in the univariable analysis added one at a a time 

Reduced model for dengue

logistic dengue_R  i.Age_Group i.Residence_R i.MPCE_R i.Education i.Occupation i.Drinking_Water_Location [pw=indiaindividualweight]

Checking for multicollinearity

collin Age_Group Residence_R MPCE_R Education Occupation Drinking_Water_Location 



* logistic chikun_R i.Age_Group i.Residence_R i.Sex i.MPCE_R i.Education i.Caste i.House_type i.wall_ceiling_damp  i.Toilet_Sanitationi.Cooking Fuel [pw= indiaindividualweight], 

only variables with <0.10 

significance retained.variables with >0.10 significance added one at a time; No variables retained.

Variables >0.25 in the univariable analysis added one at a a time 

Reduced model for chikungunya

logistic chikun_R  i.Age_Group i.Residence_R i.MPCE_R i.Education i.Caste i.House_type i.wall_ceiling_damp i.Drinking_Water_Location ii.Toilet_Sanitation [pw= indiaindividualweight]

Checking for multicollinearity

collin Age_Group Residence_R MPCE_R Education House_type wall_ceiling_damp Drinking_Water_Location Caste Toilet_Sanitation