Table 2: Summary of measurement variable and main findings of the reviewed papers.

No.Dependent variableIndependent variableFindings

1Willingness-to-buy(1) CHI premium
(2) Public health insurance coverage
(3) Age
(4) Employment, educational and health status
(5) Income
(1) The demand for CHI in urban areas had great potential.
(2) Individuals was more likely to buy and pay more for MCDI and IEI than OEI.
(3) Determinants of CHI demand were similar in the three programs (employed by private enterprises or self-employed, aged under 40, college educated, and higher income).

2Purchase of CHI(1) New Cooperative Medical Scheme (NCMS)(1) Adults were 2.1 % more likely to purchase CHI when NCMS became available.
(2) NCMS had a larger positive effect on adult private coverage in higher income groups and in communities with CMS.
(3) For adults and children, risk preferences and socio-economic status are important predictors of commercial insurance take-up.
(4) No evidence for adverse selection in the demand for private health insurance.

3Health Protection Level(1) Provincial GDP
(2) Number of hospital bed per 1000 persons
(3) Number of doctors per 1000 persons
(4) Depth and density, premium of CHI
(1) GDP, coverage level of social health insurance, number of hospital bed per 1000 persons, depth of CHI, density of CHI was positively related to the health protection level.
(2) Number of doctors per 1000 persons and premium of CHI was negatively related to the health protection level.
(3) Tangible healthcare resource boosted the healthcare efficiency but the intangible one like doctors reduced due to moral issues.
(4) Cooperation of CHI and Social health insurance was good in the region with high health protection level.

4CHI premium(1) Disposable income
(2) Family annual health expenditure per urban resident
(3) Number of social health insurance participant
(4) Social health insurance premium
(5) Elderly dependency rate
(1) CHI development level is higher in Beijing and Shanghai than in Hubei.
(2) In Hubei region, CHI premium is positively related to disposable income, number of social health insurance participant but negatively related to family annual health expenditure per urban resident and elderly dependency rate.
(3) In Shanghai region, CHI premium is positively related to number of social health insurance participant and elderly dependency rate.
(4) In Beijing region, CHI premium is positively related to disposable income, number of social health insurance participant, social health insurance premium and family annual health expenditure per urban resident.

5Health expenditure per personCHI premium per person - indicator of level of CHI development(1) CHI could promote the health protection level.
(2) The effect of CHI in promoting health protection level is different among regions.

6Total CHI premium(1) Income per person
(2) Health expenditure per person
(3) Percentage of governmental health expenditure
(4) Percentage of elderly population
(5) Total population: resource level
(6) Depth of CHI
(1) Disposable income and insurance awareness significantly promote the development of CHI
(2) Level of social health insurance is inversely correlated to the development of CHI.
(3) Health expenditure per persons, total population and the proportion of elderly have no significant effect on development level of CHI.

7CHI premium per person(1) Social health insurance premium
(2) Herfindahl-Hirschman index
(3) Proportion of foreign insurance company
(4) Health expenditure per person
(5) Disposable Income
(6) Proportion of elder population
(7) Gender
(8) Education level
(1) Social health insurance had significantly driven the development of CHI.
(2) Market competition contribute to the development of CHI.
(3) Both the influence of participation of foreign company and the professional health insurance company in development of CHI is insignificant

8CHI premium – indicate the demand of CHI(1) Disposable income
(2) Elderly population
(3) Insurance awareness
(4) Health expenditure
(1) Increase in health expenditure, disposable income and insurance awareness significantly boost the development of CHI
(2) The demand of CHI is higher among elderly.

9CHI premium(1) Disposable income per person
(2) Family health expenditure
(3) Number of social health insurance participant
(4) Elderly dependency rate
(1) Increase in disposable income, family health expenditure, number of social health insurance participants and elderly dependency rate can promote the CHI demand due to raised awareness on health.
(2) There is great difference in different areas in China regarding to the purchasing power of the residents.

10(1) Insurance coverage
(2) Gross medical cost
(3) Net out-of-pocket payment
(1) Households
(2) Income
(3) Medical expense
(4) Presence of at least one inpatient treatment
(5) Presence of chronic disease,
(6) Living in urban/rural areas
Commercial insurance coverage was significantly associated with medical expense.

11Kakwani index –(1) Concentration index of health care payments
(2) Gini coefficient of gross income
(1) The finances of private health insurance were progressive among the rich and poor.
(2) In both cities and villages, the healthcare financing channels of private health insurance and OOP payments were equitable
(3) The financing performance of private health insurance in urban areas was inferior to that in rural areas.

12CHI compensation rate(1) Number of doctor per 1000 persons
(2) Hospital bed occupancy rate
(3) Outpatient health expenditure per person
(4) Hospital bed per 1000 persons
(5) Inpatient health expenditure per person
(6) Hospital day
(1) The effect of inpatient health expenditure on the claim cost is significant but not high.
(2) The number of A&E patient admitted to inpatient department has great effect on the claim cost.

13Purchase of CHI(1) Personal health status
(2) Economic status
(3) Personal expected healthcare demand
(4) Personal risk preference
(5) Medical insurance market and policy
(6) Gender
(7) Marital status
(8) Number of family member
(9) Age
(1) In urban, insurance company based on the age for risk selection while in rural, insurance company based on the area
(2) The potential demand of urban residents is underestimated.
(3) People have significant adverse selection behavior both in urban and rural.
(4) Personal preference and personal purchasing power affect the demand of insurance
(5) Social health insurance boost the development of CHI.
(6) The coverage of CHI affects the personal demand.

14CHI claim cost(1) Number of hospital day
(2) Age
(3) Sex
(4) Marital status
(5) Occupation
(6) Origin of patient
All the factors affected the claim cost.

15Purchase of CHI(1) Gender
(2) Social health insurance
(1) With social health insurance, male is more willing to buy CHI than female.
(2) Without social health insurance, female is more willing to buy CHI than male.

16Purchase of CHI(1) Demographic characteristic
(2) Scio-economic characteristic
(3) Health status
(4) Behavioral issue
(5) Mental state
(1) Male are more likely to purchase CHI than female.
(2) With better health status, higher education level and being divorced, elderly are more likely to purchase CHI.
(3) Rural residents are more likely to purchase CHI than urban residents.

17N/AN/A(1) The market potential of CHI were not effectively developed
(2) There was cognitive biases about CHI among Chinese population.
(3) The claims service is mess and the satisfaction is low regarding to the claim process.

18Degree of coordination(1) Depth and density of CHI
(2) Depth and density of social health insurance
(1) The development of CHI in Beijing and Shanghai is better than the other areas.
(2) The development of CHI remains the same from 2005 to 2010.
(3) The development of social health insurance is better than that of CHI.
(4) The coordination between CHI and social health insurance is not high. It improved but not significant during 2005 to 2010.
(5) Coordination level of CHI and social health insurance is related with economic development level.

19Purchase of CHI(1) NCMS
(2) Family income
(3) Age
(4) Gender
(5) Educational level
(6) Marital status
(7) Employment status
(8) Family size
(9) Chronic disease
(10) Exercise
(11) Preventive care
(12) Smoking
(13) Family expenditure
(1) The relationship between NCMS and CHI at first is substitute and later be complementary.
(2) With increasing age, chronic diseases and large-size family, rural residents are more unlikely to purchase CHI.
(3) Increase in family income can lead to increase in CHI demand.
(4) Female, people with smoking habit and without preventive care are more unlikely to purchase CHI

20Kakwani index – indicate the health financing equity(1) Concentration index of health care payments
(2) Gini coefficient of gross income
(1) Healthcare financing distribution in China was unequal.
(2) CHI had played a minor role in Chinese healthcare financing due to its low associated coverage

21CHI Premium(1) Disposable income
(2) Outpatient expenditure
(3) Claim proportion of social health insurance
(4) Number of medical professional in per 10000 persons
(5) Proportion of age over 65 in total population
(1) The increase in disposable income and availability of medical resource significantly boosted the development of CHI.
(2) Social health insurance encouraged the development of CHI.
(3) The effect of proportion of elderly is not significant.

22Total CHI premium income(1) Total GDP
(2) Healthcare expenditure price index
(3) Medical graduate
(4) Total social health insurance premium
(1) With higher education level and depth of social health insurance, premium of CHI increase.
(2) The premium of CHI is more sensitive to the increase of education level than to the depth of social health insurance.
(3) Medical expenditure price level was not the main cause.

23CHI premium(1) Governmental health expenditure
(2) Urban and rural family saving
(1) The substitute relationship between social health insurance and CHI is not significant, which means there is still space for CHI development under high coverage of social health insurance in China.
(2) With increased family saving, people are more aware about their health and increase in the CHI demand.

24(1) Health insurance development index
(2) Depth of CHI: ratio of insurance premium and local GDP
(3) Density of CHI: ratio of insurance premium and local population
(1) Life expectancy
(2) Proportion of urban citizen
(3) Education level
(4) Disposable income
(5) Number of health organization
(6) Number of health professional
(7) Number of hospital bed
(8) Distribution of healthcare resource
(9) Healthcare expenditure
(10) Population density
(1) The development rate of Eastern area was twice of the rate of Middle and Western Area
(2) The level of insurance development was not positively related to its economic development.
(3) The level of insurance development did not gradually decrease from Eastern to Middle to Western.

25Out-of-pocket payment(1) Social health insurance
(2) Population
(3) Economy status
(1) CHI increases the out-of-pocket payment proportion due to moral risk behavior.
(2) The alleviation function of CHI on inpatient and outpatient expenditure is insignificant among low income people.
(3) CHI could not satisfy the needs of elderly as most of CHI target customers with age over 60.
(4) In 2006/2009, the influence of CHI on the medical expenditure was not significant but CHI can significantly reduce the self-payment of medical expenditure.
(5) In 2006/2009, compared to 2000/2004, the coverage of social health insurance was higher but the coverage of CHI did not drop significantly.

26CHI premium(1) National GDP
(2) Total population
(3) Rural family income per person
(4) Consumer good index
(5) Urban health expenditure per person
(6) Number of health organization
(7) Geographic area
(1) For eastern regions, CHI demand elasticity is low regarding to population and national GDP per person.
(2) For western regions, CHI demand elasticity is high regarding to population but low regarding to national GDP per person.
(3) For central regions, CHI demand elasticity is high regarding to population and national GDP per person.
(4) For eastern regions, change in health expenditure per person and rural family income per person can lead to significant change on CHI demand.
(5) For western regions, only rural family income per person can lead to significant change on CHI demand.
(6) For central regions, change in health expenditure per person and rural family income per person did not lead any significant change on CHI demand.

27(1) Public insurance only
(2) Private insurance only
(3) Double coverage of public and private health insurance
(4) Double coverage of both rural and urban public health insurance
(5) No insurance
(1) Policy
(2) Health
(3) Socioeconomic
(4) Risk aversion
(1) Rural residents were more likely to participate in public health insurance than urban residents
(2) Rural-to-urban migrants were more likely to be uninsured
(3) Large-size families with more elderly members may show greater willingness of insurance participation
(4) UEBMI enrollees and high SES population were more likely to buy private insurance
(5) Public insurance coverage was associated with a reduced demand for private insurance, especially for urban residents who were covered by URBMI.

28Purchase of CHI(1) Number of relative visit during CNY
(2) Number of friend visit during CNY
(3) Household income
(4) Gender
(5) Age
(6) Martial status
(7) Education level
(8) Health status
(9) Social health insurance
(10) Smoking
(11) Risk attitudes
(1) Social network has a significant positive effect for rural people purchasing CHI.
(2) In the eastern and middle part of the high degree of marketization, there is no obvious effect.
(3) In the central and western region, network has significant positive effect for the purchase.
(4) Social network is not significant for urban people

29Consumer satisfaction(1) Gender
(2) Age
(3) Marital status
(4) Education level
(5) Occupation
(6) Annual income
(7) Understanding of CHI
(1) Gender and education level is significantly related to consumer satisfaction.
(2) With higher education level, people are more satisfied with CHI.
(3) With deeper understanding of CHI, the satisfaction is lower.

30Purchase of CHIFamily Burden
(1) Family income per person
(2) Family property per person
(3) Family dependency coefficient
(1) The family with less family burden (higher income, more property and lower family dependency) was more willing to purchase CHI.
(2) For urban family, younger and married householder is more likely to purchase CHI.
(3) For rural family, family with less family burden and younger healthy householder
(4) Higher educational level could increase the purchase of CHI in urban family but reduce in rural family.

31N/AN/A(1) CHI premium increased rapidly but the depth and premium per person were still low compared to other countries.
(2) Regarding to national premium income, there was a serious imbalance: Eastern region contributed more than 50% of total premium while western region only contributed less than 20% despite of its largest increase in the past 10 years.
(3) The unbalanced CHI development among the areas in Eastern and Western regions was more prominent while in Middle and North-Eastern region the development was even.
(4) The differences in premium between provinces was increasing, in which increased from 269 yuan in 2005 to 1,072 yuan in 2015.

32CHI Income Premium(1) Disposable income per person
(2) UEBMI premium
(3) URBMI premium
(4) Proportion of general college graduate
(5) Elderly dependency ratio
(6) Out-patient service fee
(1) Disposable income, UEBMI premium, elderly dependency ratio and outpatient services expenditure were the determinant of CHI demand.
(2) For UEBMI enrollee, increase in UEBMI premium can increase CHI premium.
(3) For URBMI enrollee, increase in disposable income can increase demand on CHI.

33CHI premium(1) Daily expenditure level
(2) Personal health expenditure
(3) Density of CHI
(4) Governmental health expenditure
(5) Depth of CHI
(6) Elderly population
(1) Daily expenditure level and elderly population can promote CHI demand
(2) The current social health insurance system has “Crowding Out Effect” on CHI.
(3) Density and depth of CHI did not highly affect the CHI demand.

34CHI density:
Total premium/population
(1) Social health insurance density: Total premium from NCMS, UEBMI and URBMI/ population(1) Social health insurance can drive the development of CHI with saving purpose but no effect on health protection-oriented CHI.

35CHI Premium(1) Total health expenditure
(2) Proportion of elderly population
(3) Total population
(4) Chinese GDP
(5) Balance of residents’ RMB savings at the end of year
(6) Education expenditure
(1) Total health expenditure, proportion of elderly population, balance of residents’ RMB saving at the end of year could have a positive effect on CHI demand
(2) GDP and education expenditure did not have significant effect on CHI demand.