Skip to main content
. 2021 Aug 21;15:100901. doi: 10.1016/j.ssmph.2021.100901

Table 1.

Covariates included in the two-part mixed-effects models according to the level of analysis.

Variables Type Description
Level 1 covariates: Characteristics of the care-seeking events
Cause of health-seeking Categorical Health condition for which care has been sought
Sequence of health visits Categorical A dummy variable indicating the first, second, or the third episode of the healthcare-seeking event for a household member in the last six months
Healthcare provider Categorical Type of providers from whom healthcare was sought after
Level 2 covariates: Characteristics of the individual household members for whom healthcare was sought
Gender Binary Gender of an individual household member
Age categories Categorical Age of the household member as a category
Relationship with household head Categorical Relationship of the household member with the household head
Education Categorical Educational attainment of the household member
Employment status Binary Employment status of the household member
Disability Binary Presence of any disability of the household member
Level 3 covariates: Characteristics of the household heads and the household
Gender Binary Gender of the household head
Age categories Categorical Age categories of the household heads
Education Categorical Educational attainment of the household head
Occupation Categorical Occupation of the household heads
Freedom of decision making Binary Perceived level of freedom of decision making of the household heads
Household head's social capital measures Continuous Four standardized factor scores of household head's social capital constructs derived from the MCFA measured by Adapted Social Capital Assessment Tool-India (SASCAT-I)
Religion Binary Religion of the household
Caste Categorical Social caste of the household
Caste concordance Binary An indicator which reflects if the household caste was the same as the dominant caste of the community
Wealth Categorical Asset index was developed by principal component analysis using 27 household asset-related binary variables. The standardized score of the first component was used to create five asset quintile groups, where Quintile 1 was assigned to the least wealthy household, and Quintile 5 was assigned to the wealthiest household
Financial stability Categorical Response of the household heads to the question “In the last 12 months, how has your financial situation changed overall?”
Household member number Continuous Number of members living in the household for the last six months
Monthly health expenditure Continuous Reported total monthly healthcare expenditure converted into the International.$ using Purchasing Power Parity (PPP) 2017 of India*.
This is considered as the proxy of the severity of illness.
Level 4 covariates: Characteristics of the community
Census block Categorical The administrative boundary which serves as the enumeration block of the census in each district
Community wealth Continuous Standardized average scores of the first component of principal component analysis from the households of each cluster
Community reciprocity Continuous The community-level average response of household heads of two questions: (1) “Do you think people in your village generally are willing to help each other out?” and (2) “Do you think if you help someone in your village, they will help you in return when you need it?” (response: Yes = 2, Sometime = 1, No = 0)
Community social capital Continuous Community-level standardized factor scores of four social capital constructs derived from the MCFA

Note: * = 1 International Doller = 20.648 Rupee (https://data.worldbank.org/indicator/PA.NUS.PPP?locations=IN).