Table 2.
Description of independent variables
| Independent variable | Explanation | Note |
|---|---|---|
| Parental health care utilization | 1. Parental outpatient visit: The paired parent used outpatient services in the past 2 weeks (Yes = 1, No = 0) 2. Parental infusion use: The paired parent used infusion during outpatient visits in the past 2 weeks (Yes = 1, No = 0) 3. Parental inpatient visit: The paired parent used inpatient services in the last year (Yes = 1, No = 0) |
Binary instead of count variables were employed, since only 1.71% of the surveyed parents used outpatient services more than once in the past 2 weeks and 0.46% used inpatient services more than once in the last year. |
| Parental socioeconomic level | Per capita household expenditure last year (five equal quantiles for the poorest, poor, average, rich, and the richest) | It has shown to be a better proxy for resources available than self-reported income which is more likely to be misreported [65, 66]. |
| Children’s health insurance | A child was covered by any insurance scheme (Yes = 1, No = 0) | Among the surveyed children, 91.6% of the urban children were covered by URBMI and 96.9%% of the rural children were covered by NCMS. Since whether a child was under URBMI or NCMS depended on his/her residential area controlled later, we used an indicator variable for insurance status instead of specifying different insurance schemes. |
| Children’s need for care | 1. Days of illness if a child was reported to have any illnesses or injuries in the past 2 weeks 2. Children’s height and weight (self-reported) |
We used the days of illness to indicate the severity of the illness and therefore the need for health care services. As the recall period for the days of illness was 2 weeks, this variable was not able to be taken into account in the analysis of the use of inpatient services in the last year. |
| Parental need for care | Days of illness if the parent was reported to have any illnesses or injuries in the past 2 weeks | Due to time inconsistency, this variable was not used in the inpatient model. |
| Children’s age | Grouped into 0–2, 3–5, 6–8, 9–11, 12–14, 15–17 years | |
| Children’s gender | Male or female | |
| Children’s ethnicity | Han or others | |
| Parental age | Parental age in the survey year | |
| Residential areas | Urban or rural | |
| Living region | Shanbei (north), Guanzhong (central), and Shannan (south) | |
| Family size | The number of members in a family |