Table 1.
Data extraction and the quality of the studies
| Year of publication | Title | Author | Country | Survey type | Sample size | Variables | Result (outcome) | Representativeness | Quality |
|---|---|---|---|---|---|---|---|---|---|
| 1986 | Families with catastrophic health care expenditures | Wyszewianski (1986) | Michigan, US | 1977 National Medical Care Expenditure Survey (NMCES) | 14,615 HH* |
Independent variables: Demographical characteristics, income, poverty status, HH head age and work status. Dependent variables:
|
4.2% of all HH had a CHE where the OOP was ≥ 20% of their total income and 9.6% of the HH had CHE at ≥ 10% threshold. The determinants of CHE were low income where 2/3 of them below the poverty line, HH head age > 65 years old or unemployed HH head. | Representative | Medium |
| 2006 | Catastrophic household expenditure for health care in a low income society: A study from Nouna district, Burkina Faso | Su et al. (2006) | Burkina Faso | Nouna health district household survey 2000–2001 | Sample size was 800 HH, 320 urban, 480 rural, 774 were included in the study. |
Independent variables HH characteristics (HH residence and economic status), sex and the educational status of the HH head, treatment and illness pattern variables. Dependent variables
|
CHE = 8.66% (based on ratio of health payment of 40% or more of CTP****) Seeking health care, average number of illness, chronic illness, and economic status were the factors found to be associated with CHE. |
Representative Response rate 96% |
Good |
| 2006 | Understanding the impact of eliminating user fees: Utilisation and catastrophic health expenditures in Uganda | Xu et al. (2006) | Uganda | Socioeconomic Surveys of Government of Uganda, 1997, 2000 and 2003 | 6,655, 10,691 and 9,710 households in turn, comprising 33,988, 53,761 and 47,468 individuals in 1997, 2000 and 2003, respectively. |
Independent variables Type of provider, presence of a member of 65 years old, presence of a member of less than 5 years old, HH head educational status and sex, HH residence and inpatient service use. Dependent variables
|
CHE = 2.92% (based on ratio of health payment of 40% or more of CTP) The determinants of CHE: inpatient service used among poor, HH member of > 65 years, HH head with little education, urban settlement was protective for non poor and not for poor, the elimination of fees didn’t reduce the CHE incidence. | Representative | Medium |
| 2009 | Which households are at risk of catastrophic health spending: Experience in Thailand after universal coverage | Somkotra, Lagrada (2009) | Thailand | Household Socioeconomic Surveys (SES) 2006. And data from the SES in 2000, 2002, and 2004 were also examined. |
24,747, 34,785, 34,843 and 22,547 HH collected in 2000, 2002, 2004, 2006, respectively. |
Independent variables: Sex, age and educational status of the HH head, presence of elderly, presence of children, HH economic status, HH size, insurance coverage, presence of a member with chronic illness or disability, or being hospitalised and the type of health care. Dependent variables:
|
CHE = 0.77% (2006), 0.97% (2004), 1.07% (2002), 1.23% (2000) (based on ratio of health payment of 40% or more of CTP). Based on health payment of ≥ 10% of total income; the CHE = 4.03% (2006), 4.8% (2004), 5.03% (2002), 6.44% (2000). Important determinants were inpatient care at public providers among the poor, and the outpatient care at private facilities, presence of elderly and members of chronic illness or disability or having a member hospitalised in past 12 months. Higher education HH were with less probability of having CHE. |
Representative | Medium |
| 2009 | Household catastrophic health expenditure: Evidence from Georgia and its policy implications | Gotsadze et al. (2009) | Georgia | Health Care Utilisation and Expenditure survey conducted during May–June 2007 | 2,859 households |
Independent variables: Residence, a member of chronic illness, hospitalisation and the HH economic status. Dependent variables:
|
CHE = 11.7% (based on ratio of health payment of 40% or more of CTP) CHE prevalence was 27 times with those with chronic illness and hospitalisation. The rich were less likely to have CHE. |
Representative | Medium |
| 2010 | Catastrophic health expenditure and impoverishment in Turkey | Yardima et al. (2010) | Turkey | Household Budget Survey, Consumption Expenditures, 2006. | 8,558 households |
Independent variables: Age, sex, educational and employment status of the head of the HH, family size, economic status presence of preschool kids, HH settlement, insurance coverage and presence of a member with disability. Dependent
|
CHE = 0.6% (based on ratio of health payment of 40% or more of CTP). Significant factors were HH residence, presence of a member with disability, HH head education status and work status, presence of elderly, presence of preschool children and insurance coverage. | Representative | Medium |
| 2010 | The influence of the rural health security schemes on health utilisation and household impoverishment in rural China: Data from a household survey of Western and Central China | Shi et al. (2010) | China | Community, household survey 2008 in Hebei and Shaanxi provinces, and the Inner Mongolia Autonomous Region, which represent Western and Central China | 3,340 households |
Independent variables: Age, gender, ethnicity, education level, Occupational status, marital status and religion of the HH head, insurance Status, presence of a member with chronic illness or disability, number of episodes of in-patient visits, unmet inpatient need, HH per capita expenditure, health payment, HH in poverty and household capacity to pay. Dependent variables:
|
The incidence of CHE = 14.3% (based on ratio of health payment of 40% or more of CTP) The CHE determinants were poorer HH, low education of HH head and presence of a member with chronic disease. Insurance found to be reduced the risk of CHE. Those with lower expenditure quintile were more likely to be impoverished (8%). |
Representative Response rate 99.8%. |
Medium |
| 2010 | Catastrophic out-of-pocket payment for health care and its impact on households: Experience from West Bengal, India | Mondal et al. (2010) | India | Household survey 2007 | 3,150 HH, 15,277 individuals |
Independent variables: Prevalence of illness, HH characteristics; size, residence and the economic status Dependent variables:
|
> 30% of HH spend ≥ 40% of non-food expenditure on inpatient care, those used private hospital spend 25% of their annual income on inpatient care, rural residence, birth delivery, presence of a member with chronic illness, hospitalisation, number of illness episodes, type of medical care were considered as the most important determinants of CHE. | Representative | Medium |
| 2011 | Study of catastrophic health expenditure in China’s basic health insurance | Zhou, Gao (2011) | China | Forth National Health Service Survey (NHSS) conducted in Shaanxi Province (west) 2008 | 1,215 households covered by UEMS or URMS (insurance scheme), and 2,875 households covered by NCMS were chosen in this study. |
Independent variables: Illness, presence of a member with chronic disease, outpatient and inpatient used, sex and the education level of HH head, presence of a member with 65 years old, location of HH, family size, economic status and the insurance type Dependent variables
|
CHE=16.87%–19.62% (based on ratio of health payment of 40% or more of CTP) The important determinants were presence of elderly, hospitalisation, poor health, presence of a member with chronic illness, family size and the HH economic status. |
Representative | Medium |
| 2011 | Determining factors of catastrophic health spending in Bogota, Colombia | Amaya, Ruiz (2011) | Colombia | Expenditure Survey performed by Cendix (2001) | 2,810 households |
Independent variables: Age, gender, the work status and social security of the HH head. HH income in different quintiles, HH size, disability, child births, Dependent variables:
|
CHE at ≥ 20% of CTP was 4.5%, it was higher among low income HH. The significant risk factors were absence of social security and having inpatient admission, and those with small family size and when the HH head was > 60 years old or have no work | Representative | Medium |
| 2011 | Effect of household and village characteristics on financial catastrophe and impoverishment due to health care spending in Western and Central Rural China: A multilevel analysis | Shi et al. (2011) | China | A cross-sectional community household survey 2008 | A total of 3,334 residents from 3,340 households |
Independent variables Age, sex, race, marital status, education, occupation and religion of HH head, no. of patient with chronic illness, hospitalisation, HH insurance status, HH income, adult illiteracy rate, availability of health clinic, distance from village to clinic, and to county hospital. Dependent variables:
|
CHE = 18.4% (based on ratio of health payment of 40% or more of CTP) Households with low per capita income, having elderly, hospitalised or chronically ill members, and whose head was unemployed were more likely to incur financial catastrophe and impoverishment due to health expenditure. Both catastrophic and impoverishing health payments increased with increased village deprivation. | Rural only response rate of 99.8% | Good |
| 2011 | Catastrophic spending on health care in Brazil: Private health insurance does not seem to be the solution | Barros et al. (2011) | Brazil | 2002–2003 Brazilian Household Budget Survey | 37,830 urban households only |
Independent variables:
|
CHE = 2% at 40% CTP, and 15.5% according to 10% of total income. Poorest had seven times greater risk of CHE than the rich, Socioeconomic position, sex of the head is insignificant, and presence of elderly increase the risk, HH with health insurance at greater risk of CHE. | Used only urban HH | Medium |
| 2012 | Unexpected impact of changes in OOP payments for health care on Czech household budgets | Krutilova, Yaya (2012) | Czech | Household budget survey, 2007, 2008 and 2009. | 3,000 households, 2007, 2008 and 2009. Before and after user fees |
Independent variables: Sociodemographic factors (head age, sex, educational status and work status, residence, economic status, no. of kids. Dependent variables:
|
CHE = 11.89% (based on 5% or above of total income). Most affected HH were those with pensioners, elderly and low income. |
Representative | Medium |
| 2012 | Factors affecting catastrophic health expenditure and impoverishment from medical expenses in China: Policy implications of universal health insurance | Li et al. (2012) | China | Fourth National Health Service Survey (NHSS, 2008). | 55,556 households |
Independent variables Sex, educational and employment status of HH head, health insurance status, HH economic status, HH size, having at least one member older than 60 or younger than 5 years or with tuberculosis or any chronic non-communicable condition or hospitalised member. Dependent variables:
|
CHE = 13% (based on ratio of health payment of 40% or more of CTP) Determinants of CHE were HH headed by a female, an unemployed person or having little education. Having at least one member who was elderly, ill from tuberculosis or chronic non-communicable illness, or hospitalised, without insurance and rural HH were at greater risk of CHE. |
Representative | Good |
| 2012 | Measuring incidence of catastrophic OOP health expenditure: With application to India | Pal (2012) | India | Household Consumer Expenditure Survey 2004–2005 | Not mentioned |
Independent variables: Economic variables: land, wealth index, regular salary, education status, plus the socioeconomic factors as HH size, No. of children and elderly, sex of HH head and his age. Dependent variables:
|
CHE = 14.68% among the poorest and 34.90% among the richest (using 10% threshold of total budget) CHE = 4.84% among poor −13.76% among rich (based on 40% threshold of CTP) Large HH, presence of children and elderly and aged HH head were the significant determinants of CHE. |
Representative | Medium |
| 2012 | Inequality in HH catastrophic health care expenditure in a low-income society of Iran | Kavosi et al. (2012) | Iran | WHO survey in 2003 and repeated again by research team in 2008 | 1123 households in 2003, 635 households in 2008 |
Independent variables: HH head sex, HH size, presence of a member of > 65 years old or less than 5 years old, HH insurance status, presence of a member with disability, HH economic status, using dentistry service or inpatient or out patient service. Dependent variables:
|
CHE = 12.6% in 2003, 11.8% in 2008 (based on ratio of health payment of 40% or more of CTP) The important determinants were presence of HH member over 65 years old or with disability, lower economic quintile, using of inpatient, outpatient and dentistry health services and lack of insurance. |
Representative | Medium |
| 2012 | Iranian household financial protection against catastrophic health care expenditures | Moghadam et al. (2012) | Iran | Iranian household survey 2008 | 39,088 households |
Independent variables: HH economic status, family size, inpatient and outpatient health care utilisation, drug consumption, drug addiction cessation and insurance status. Dependent variables:
|
CHE = 2.8% (based on 40% of CTP) Important determinants were large family size, low economic status, inpatient and outpatient health care utilisation, drug consumption, drug addiction cessation |
Representative | Medium |
| 2012 | Catastrophic health care spending and impoverishment in Kenya | Chuma, Maina (2012) | Kenya | Health expenditure and utilisation survey, 2007 | 8,414 households |
Independent variables: HH economic status, Type of health care utilisation Dependent variables:
|
CHE = 15.5% (using 10% threshold of total budget) and 11.4% (based on ratio of health payment of 40% or more of CTP). Lower income HH was more likely to had CHE. The use of outpatient services leads to CHE more than the use of inpatient services. The poverty level = 54.9% and it increased 2.7% after health care payment. | Representative | Medium |
| 2012 | Measuring the catastrophic and impoverishing effect of household health care spending in Serbia | Arsenijevic et al. (2012) | Serbia | Serbian Living Standard Measurement Study (LSMS) | 5,557 households |
Independent variables: HH economic status, residence, HH size, HH head educational level, age, employment, marriage status, gender and presence of member with chronic illness Dependent variables:
|
CHE = 2%–2.4% (based on total income) and 0.8%–1.1% (base on CTP), significant determinants were rural residence, not married HH head, low education, low economic status, large family size, presence of member with chronic illness | Representative | Medium |
| 2013 | Financial burden of HH OOP health expenditure in Vietnam: Findings from the National Living Standard Survey 2002–2010 | Van Minh et al. (2013) | Vietnam | Vietnam Living Standard Survey 2002, 2004, 2006, 2008 and 2010 | 45,000, 37,200, 36,756, 36,756 and 46,995 households in 2002, 2004, 2006, 2008 and 2010, respectively |
Independent variables: HH head sex, HH size, presence of a member of > 65 years or less than six years, HH insurance status, HH economic status and residence Dependent variables:
|
CHE = 4.7% in 2002, 5.7% in 2004, 5.1% in 2006, 5.5% in 2008 and 3.9% in 2010 (based on ratio of health payment of 40% or more of CTP) The important determinants were presence of HH member over 65 years or less than six years, higher economic quintile and living in rural area. Those who pushed into poverty were 3.4%, 4.1%, 3.1%, 3.5%, 2.5% in 2002, 2004, 2006, 2008 and 2010, respectively. |
Representative | Good |
| 2013 | Catastrophic health expenditure and entitlement to health services in the occupied Palestinian territory: A retrospective analysis | Ashour et al. (2013) | West bank and Gaza (Palestine) | Palestinian Consumption and Expenditure Survey, 2010 | 3,754 households |
Independent variables: HH head sex, education and work status. HH income, residence and insurance status. Dependent variables:
|
CHE = 2.4% (based on ratio of health payment of 40% or more of CTP). The prevalence was less among insured HH in compare to uninsured ones. CHE significantly differed according to different factors considered (HH head sex, education and work status. HH income and residence) | Representative | Medium |
| 2013 | Health-Related financial catastrophe, inequality and chronic illness in Bangladesh | Rahman et al. (2013) | Bangladesh | Household survey of 1600 households in Rajshahi city August to November 2011 | 1,600 households |
Independent variables: HH head sex and educational level, presence of a member of > 65 years, HH economic status and type of health care utilised. Dependent variables:
|
CHE = 9% (based on ratio of health payment of 40% or more of CTP). The important determinants were presence of HH member hospitalised or had a chronic illness, number of illness, the economic status and the educational level of the HH head. | Represented only the urban household Response rate 99.6% |
Medium |
| 2013 | Assessing the magnitude, distribution and determinants of catastrophic health expenditure in urban Lucknow, North India | Misra et al. (2013) | India | Household survey in 2011–2012 in urban Lucknow | 400 households |
Independent variables: HH economic status, HH size and type of health care utilised. Dependent variables:
|
CHE = 11.5%, 4%, 3%, 2.75% at 10%, 20%, 30% and 40% of HH capacity to pay, respectively. Important determinants were outpatient and inpatient health care utilisation and the economic status of the HH. | Urban representation | Medium |
| 2013 | Catastrophic health expenditure in un urban city: Seven years after universal coverage policy in Thailand | Weraphong et al. (2013) | Thailand | A cross sectional survey in Nakhon Sawan Municipality in 2008 | 406 sampled households |
Independent variables: HH economic status and type of health care utilised the cost components of treatment and the insurance scheme. Dependent variables:
|
CHE = 7.1% in non-poor and 12.5% poor (based on 10% of total HH income). Important determinants were the use of public and private hospitals and clinics, transportation cost, loss of time cost and civil servants card holder. | Urban representation | Medium |
| 2013 | Household catastrophic medical expenses in Eastern China: Determinants and policy implications | Li et al. (2013) | China | Health care utilisation and expense survey, 2008 | 11,577 households |
Independent variables: HH economic status, residence, HH size, presence of children or elderly. HH head educational level, presence of a member with chronic illness or being hospitalised and the insurance scheme. Dependent variables:
|
CHE = 9.24% to 24.79% (based on ratio of health payment of 40% or more of CTP). Important determinants were low economic status, rural residence, hospitalisation, member with chronic illness, presence of elderly or children, large HH size, no or low education of HH head and type of insurance scheme. | Representative | Medium |
| 2014 | Catastrophic health expenditure and rural household impoverishment in China: What role does the new cooperative health insurance scheme play? | Li et al. (2014) | China | Fourth National Health Service Survey (NHSS, 2008) | 56,400 households |
Independent variables: HH economic status, presence of elderly or children, HH size, HH head sex, work status and educational level, hospitalisation and presence of a member with chronic illness, insurance status. Dependent variables:
|
CHE = 14.4% (based on ratio of health payment of 40% or more of CTP), poverty = 9.2%. Important determinants were hospitalisation, member with chronic illness, presence of elderly or children, HH head female, no or low education and unemployment of the HH head and type of insurance scheme. | Representative | Medium |
| 2014 | Correlates of out-of-pocket and catastrophic health expenditures in Tanzania: Results from a national household survey | Brinda et al. (2014) | Tanzania | National Panel Survey (TZNPS) in 2008–2009 | 3,265 households |
Independent variables: HH size, HH head age, sex, work status and educational level, presence of a member with chronic illness or disability. Dependent variables:
|
CHE = 18% (based on ratio of health payment of 40% or more of CTP). Significant determinants were large HH size, unemployment or manual labourer HH head, presence of a member with chronic illness or disability. | Representative | Medium |
| 2014 | Out-of-pocket health care expenditure in Turkey: Analysis of the 2003–2008 household budget surveys | Brown et al. (2014) | Turkey | Turkish Household Budget Surveys (2003–2008) | 800 household surveyed per month in all the years except 2003, where 2,200 household surveyed in that year. |
Independent variables: HH size, economic status, presence of elderly or less than 5 years children, residence, HH head sex, work status and educational level, presence of a member with illness or disability and insurance status. Dependent variables:
|
CHE = 1.2%–17.6% at different years (2003–2008) at different cut off points (2.5%, 5%, 10%, 15% and 20%) of total HH expenditure. Significant determinants were presence of elderly or less than 5 years children, or presence of a member with illness or disability, no insurance and low education of HH head. | Representative | Medium |
| 2014 | Financial catastrophe and poverty impacts of out-of-pocket health payments in Turkey | Narci et al. (2014) | Turkey | Turkish Household Budget Surveys (2004–2010) | 62,886 households in study years |
Independent variables: HH size, economic status, presence of elderly or less than 5 years children, residence, HH head sex, work status and educational level, presence of a member with disability, inpatient care and insurance status. Dependent variables:
|
CHE varied according to different thresholds used and at different years using both methods (capacity to pay and total income method). All the determinants studies had a positive relationship to CHE except the work status of household head. The prevalence of impoverishment was less than 1 in all the studied years. | Representative | Medium |
| 2014 | Catastrophic healthcare expenditure – drivers and protection: The Portuguese case | Kronenberg, Barros (2014) | Portugal | Portuguese Household Budget Survey (2000 and 2005) | 10, 020 households (2000) 10,403 household (2005 ) |
Independent variables: HH size, economic status, presence of elderly or less than 5 years children, residence, HH head sex, age, work status and educational level, presence of a member with disability Dependent variables:
|
CHE = 5.03%–32.76% at different thresholds in 2000 and 2005 year analysis (based on the CTP calculation). Important determinants were age of HH head, presence of member with disability, economic status and rural residence in 2005. | Representative | Medium |
| 2014 | Socioeconomic inequality in catastrophic health expenditure in Brazil | Boing (2014) | Brazil | National Household Budget 2002–2003 and 2008–2009 | 48,470 HH in 2002–2003 and 55,970 HH in 2008–2009 |
Independent variables: HH economic status, HH head level of education Dependent variables:
|
CHE = 0.7% and 21.0%. CHE prevalence and socioeconomic inequality increased from 2002–2003 to 2008–2009. Determinants: The low economic status and low educational level. | Representative | Medium |
| 2015 | Measurement and explanation of socioeconomic inequality in catastrophic health care expenditure: Evidence from the rural areas of Shaanxi Province | Xu (2015) | China | National Household Health Service Surveys of Shaanxi Province, 2008 and 2013 | 3,217 HH in 2008 and 13,085 HH in 2013 |
Independent variables: HH head gender and educational level. HH characteristics (presence of a member of 65 years old, presence of a member of less than 5 years old, economic status, HH size and insurance status. Presence of a member with chronic illness, or receiving inpatient or outpatient care Dependent variables:
|
CHE = 17.19% in 2008 and 15.83% in 2013, the inequality in facing CHE strongly increased. The determinants of CHE were HH economic status and HH size in 2013, the absence of commercial health insurance and having elderly members | Representative for rural area | Good |
| 2015 | Catastrophic health expenditure and its determinants in Kenya slum communities | Buigut (2015) | Kenya | Data from Indicator Development for Surveillance of Urban Emergencies (IDSUE) project, 2011–2013 | 9447HH |
Independent variables: HH head gender, age and work status, HH characteristics (presence of a member of less than 5 years old, economic status and insurance status. Type of illness, seeking care in case of illness and the type of health care facility Dependent variables:
|
CHE = 1.52%–28.38%. The CHE determinants were the number of working adults in a HH and membership in a social safety net appear to reduce the risk of catastrophic expenditure. Seeking care in a public or private hospital increases the risk of CHE. | Representative for slums | Medium |
| 2015 | Health care expenditure of households in Magway, Myanmar | Khaing (2015) | Myanmar | Cross-sectional Household survey, 2012 | 700 HH |
Independent variables: HH head gender, age and education level, HH characteristics (family size, residence). Seeking outpatient or inpatient health care. Dependent variables:
|
CHE = 25.2% in urban area and 22.7% in rural area. The CHE determinants were HH medium educational level, large family and hospitalisation. |
Representative | Medium |
| 2015 | Financial risks from ill health in Myanmar-Evidence and policy implications | Htet (2015) | Myanmar | World health survey, 2002–2003 | 6,045 HH |
Independent variables: HH head gender, female education level, self-rated health. HH characteristics (family size, residence, presence of a member of less than 5 years old or > 60 years old, economic status, presence of pregnant woman, ethnicity, use of insecticide treated bed net). Dependent variables:
|
CHE = 41%. CHE determinants were presence of a member of less than 5 years or > 60 years old, large HH size, poor self-rated health, poor HH, presence of member with chronic illness and being of ethnic minority, female head | Representative | Medium |
| 2016 | Catastrophic health expenditure according to employment status in South Korea: A population-based panel study | Choi (2016) | South Korea | Korean Welfare Panel Study Survey (KOWEPS), 2009–2012 | 5,335 HH |
Independent variables: HH head gender, age, education level, change in employment status, marital status, self-rated health), HH characteristics (family size, HH income, insurance status, presence of a member of > 65 years, with chronic disease or depression or disability). Dependent variables:
|
CHE = 4.1%, The CHE determinants were female HH head, married, change job status, family size of two persons, negative self-rated health, having a member of > 65 years old, or a member with chronic illness, disability or depression | Representative | Good |
| 2016 | Catastrophic health expenditure after the implementation of health sector evolution plan: A case study in the West of Iran | Piroozi (2016) | Iran | A cross sectional survey in Sanandaj city, 2015 | 663 households |
Independent variables: Gender of HH, HH characteristics (presence of a member of 65 years old, presence of a member of less than 5 years old, economic status, HH size, insurance status, receiving dental care, rehabilitation, impatient and outpatient spending). Dependent variables:
|
4.8% of all HH had a CHE. The determinants of CHE were household economic status, presence of elderly or disabled members in the household and utilisation of inpatient or rehabilitation services. | Representative for West of Iran | Good |
| 2016 | Does user fee removal policy provide financial protection from catastrophic health care payments? Evidence from Zambia | Masiye (2016) | Zambia | Zambia Household Health Expenditure and Utilisation Survey (ZHHEUS) in 2014 | 12,000 households |
Independent variables: Gender of patient, HH head age, work status and educational level, HH characteristics (economic status, residence, distance to health care facility). Facility type of health care and type of illness). Dependent variables:
|
CHE = 10%, the CHE prevalence reduced after implementation of user fees removal policy. The determinants of CHE were age of patients, distance, facility type, HH economic status and type of illness. | Representative Response rate 99.4% |
Good |
*
HH (household)
**
CHE (catastrophic health expenditure)
**
OOP (out of pocket payment)
****
CTP (capacity to pay)