https://orcid.org/0000-0001-7468-1249
Abstract
More than 50% of health expenditure is financed through out-of-pocket payments in India, imposing a colossal financial burden on households. Amidst the rising incidence of non-communicable diseases, injuries, and an unfinished agenda of infectious diseases, this study examines comprehensively the economic impact of out-of-pocket health expenditure (OOPE) across 17 disease categories in India. Data from the latest round of the National Sample Survey (2017–18), titled ‘Household Social Consumption: Health’, were employed. Outcomes, namely, catastrophic health expenditure (CHE), poverty headcount ratio, distressed financing, foregone care, and loss of household earnings, were estimated. Results showed that 49% of households that sought hospitalization and/or outpatient care experienced CHE and 15% of households fell below the poverty line due to OOPE. Notably, outpatient care was more burdensome (CHE: 47.8% and impoverishment: 15.0%) than hospitalization (CHE: 43.1% and impoverishment: 10.7%). Nearly 16% of households used distressed sources to finance hospitalization-related OOPE. Cancer, genitourinary disorders, psychiatric and neurological disorders, obstetric conditions, and injuries imposed a substantial economic burden on households. OOPE and associated financial burden were higher among households where members sought care in private healthcare facilities compared with those treated in public facilities across most disease categories. The high burden of OOPE necessitates the need to increase health insurance uptake and consider outpatient services under the purview of health insurance. Concerted efforts to strengthen the public health sector, improved regulation of private healthcare providers, and prioritizing health promotion and disease prevention strategies are crucial to augment financial risk protection.
Key messages
The economic burden of out-of-pocket health expenditure was higher for outpatient care (CHE: 47.8% and impoverishment: 15.0%) compared with hospitalization (CHE: 43.1% and impoverishment: 10.7%).
Nearly 1.8% of ailing individuals did not seek treatment and 10.1% of ailing individuals did not seek treatment on medical advice.
Cancer, genitourinary disorders, psychiatric and neurological disorders, obstetric conditions, and injuries imposed a substantial economic burden on Indian households.
Introduction
Universal Health Coverage (UHC), the centrepiece of the United Nations’ sustainable development goals on health (SDG-3), aims to ensure that everyone has access to quality healthcare without facing financial hardships (WHO, 2021a). SDG-3 focuses on a broad gamut of health-related issues pertinent to the global community as well as developing countries such as India (United Nations Development Programme (UNDP), 2022). India is experiencing a triple burden of diseases, i.e. increasing non-communicable diseases (NCDs), an unfinished agenda of infectious diseases, and a rising incidence of injuries (Bloom et al., 2014). Between 1990 and 2016, the proportion of all deaths in India due to NCDs increased from 37.9% to 61.8%, and the contribution of NCDs to total disability-adjusted life years increased from 30.5% to 55.4% (Indian Council of Medical Research, Public Health Foundation of India, and Institute for Health Metrics and Evaluation (ICMR, PFHI, and IHME), 2017). Communicable diseases too, such as diarrhoea, tuberculosis, lower respiratory infections, and vector-borne diseases (for instance, dengue, malaria, and chikungunya), continue to pose substantial challenges in India (Indian Council of Medical Research, Public Health Foundation of India, and Institute for Health Metrics and Evaluation (ICMR, PFHI, and IHME), 2017). Furthermore, in 2018, India accounted for ∼11% of accident-related deaths worldwide, ranking first among the 199 countries in terms of road accident mortality (Government of India (GOI), 2019). Studies estimate that NCDs and mental disorders will lead to ∼$4.58 trillion output loss in India during 2012–30 due to savings lost and foregone productivity (Bloom et al., 2014). Despite this overwhelming scenario in India, the government health expenditure is dismally low (1.15% of gross domestic product) (Government of India (GOI), 2017a). A combination of low health insurance coverage and a dominant presence of fee-for-service private health sector has forced Indian households to rely on out-of-pocket health expenditure (OOPE) as a means of financing healthcare (Shahrawat and Rao, 2012).
In India, OOPE accounts for 50.6% of health expenditure (World Health Organization (WHO), 2019a). High OOPE reduces access to healthcare services, decreases the consumption of food and basic necessities, and exposes households to financial catastrophe and impoverishment. Healthcare payments are a major cause of poverty in India, pushing ∼32–39 million individuals below the poverty line each year (Van Doorslaer et al., 2006; Bonu et al., 2007; Garg and Karan, 2009). Poor people not only lack the financial resources to access healthcare, but illness also reduces labour supply and limits their financial ability (World Bank, 2014), creating a vicious circle of poverty and poor health.
The rising disease burden in India, accompanied by abysmally low public health spending and insurance uptake, warrants analysis of the economic impact of OOPE across all types of diseases—communicable, non-communicable, and injuries. However, limited literature is available on the financial burden of OOPE across various ailments in India. Previous studies focussed mainly on OOPE due to hospitalization (Kastor et al., 2018) or specific ailments, such as maternal health (Bonu et al., 2009; Mohanty and Kastor, 2017), NCDs (Engelgau et al., 2012; Tripathy et al., 2016; Verma et al., 2021), cancer (Mahal et al., 2013; Rajpal et al., 2018), diabetes (Tripathy and Prasad, 2018), and tuberculosis (Yadav et al., 2021a). Other studies provided estimates for small geographic areas with non-representative data, thereby limiting the generalizability (Sneha et al., 2017; Swain et al., 2018). One study examined the economic burden of OOPE across all diseases but did not report it separately for inpatient and outpatient care and was based on data from the previous National Sample Survey (NSS) (Sangar et al., 2019a). Another study examined the OOPE burden separately for hospitalization and outpatient care across all ailments but was limited to evaluating catastrophic health expenditure (CHE) and impoverishment impact only (Yadav et al., 2021c). Moreover, to the best of our knowledge, no study in India has computed CHE usingcapacity-to-pay approach for all categories of diseases.
Against this backdrop, we provide a comprehensive examination of the economic burden of OOPE across 17 disease categories, disaggregated by the type of care sought (hospitalization, outpatient care, and either hospitalization or outpatient care or both) and the type of healthcare facility utilized (public or private). Specifically, our study was guided by the following objectives. First, we computed OOPE and the share of OOPE in total household consumption expenditure. Second, we estimated the incidence of CHE, percentage of households falling below the poverty line due to OOPE, and the incidence of using distressed sources to cope with the cost of illness. Third, we gauged the loss of household earnings resulting from hospitalization and outpatient care for various ailments. Fourth, we estimated the unmet needs (i.e. percentage of individuals who did not seek treatment) and percentage of individuals who did not seek treatment on medical advice and reasons for the same. This holistic assessment is expected to serve as a valuable resource for evidence-based policy decisions to improve the accessibility of healthcare services and augment financial risk protection for Indian households.
Data and methodology
Overview of data source
The study used data from the latest round of the NSS on health, titled ‘Household Social Consumption: Health’, which was conducted from July 2017 to June 2018. This is a nationally representative survey that covered 113 823 households and 555 115 individuals across the country. The data were collected using a stratified multi-stage sampling design, with village and urban blocks as the first unit and households as the second unit. The survey collected detailed information about the nature of the ailment, utilization of health facilities, cost of inpatient and outpatient services, and demographic and socio-economic characteristics of households and their members. It collected information about the prevalence of 61 types of diseases, which were further classified into 17 broad categories. The disease classification under the NSS health survey is provided in Supplementary Table 1.
Outcome variables
Out-of-pocket health expenditure
The NSS health survey recorded total health expenditure under three broad categories: medical, non-medical, and transportation expenditure. Medical expenditures included doctors’ fees, cost of medicines, diagnostic tests, bed charges, other medical expenses (attendant charges, physiotherapy, blood, etc.), and package component, and non-medical expenditures included expenses on registration, food, lodging, etc. To determine OOPE, any reimbursement amount received was deducted from the total health expenditure. The survey recorded hospitalization expenditure during the last 365 days, and this was divided by 12 to compute monthly hospitalization OOPE. Similarly, outpatient expenses were recorded for the last 15 days, and these were multiplied by 2 to compute monthly outpatient care OOPE.
Share of OOPE in total household consumption expenditure
The share of OOPE in total household consumption expenditure was calculated by the following formula, as done by previous studies (Selvaraj et al., 2018; Yadav et al., 2021c).
Catastrophic health expenditure
Literature suggests two alternative approaches for calculating CHE. The first method defines health spending as catastrophic if it exceeds some fraction of a household’s income or total expenditure (Berki, 1986). A potential problem in this approach is that it fails to account for the CHE of poor households (Mohanty and Kastor, 2017). This is because the health payment budget share among poor households is low since most resources are consumed by the items essential for sustenance, leaving little money to spend on healthcare (Bonu et al., 2009; Mohanty and Kastor, 2017). The second method partially overcomes this limitation by defining health spending as catastrophic if it exceeds 40% of the household’s capacity to pay (Xu et al., 2003). We adopted the capacity-to-pay approach in this study, while previous studies followed the former approach (Kastor et al., 2018; Sangar et al., 2019a; Yadav et al., 2021c).
For each household i, we defined CHE as follows:
In the above equation, OOPEi is the out-of-pocket health expenditure of ith household, HCEi is the consumption expenditure of ith household, and SEi is the subsistence expenditure of ith household.
The subsistence expenditure is calculated using either food expenditure or the official poverty line (Bonu et al., 2009; Mohanty and Kastor, 2017). However, since the NSS health survey does not provide detailed information on individual expenditure items (for instance, food and non-food items), the food expenditure was unavailable. Thus, we used the inflation-adjusted official poverty line as defined by the Tendulkar Committee (Planning Commission, 2014) as a proxy for subsistence expenditure (Bonu et al., 2009; Mohanty and Kastor, 2017). The poverty line was multiplied by the household size to obtain the household-level subsistence expenditure, which was then subtracted from the total household consumption expenditure to derive a household’s capacity to pay (Bonu et al., 2009; Engelgau et al., 2012).
The proportion of households experiencing CHE was estimated as follows:
where N is the total number of households.
Poverty headcount ratio
The poverty headcount ratio estimates the proportion of households falling below the poverty line due to OOPE (Yadav et al., 2021d).
In the above equation, PL is the inflation-adjusted official poverty line given by the Tendulkar Committee (Planning Commission, 2014).
where N is the total number of households.
Incidence of using distressed sources to finance OOPE
The NSS health survey collected information about various sources of finance (household income/savings, borrowings, sale of physical assets, contributions from friends and relatives, and other sources) used as coping mechanisms. We categorized a household as incurring distressed financing if it used any of these sources except household income or savings (Sangar et al., 2020).
The proportion of households employing various sources of finance to cope with OOPE was calculated as follows.
In the above formula, I is the incidence of using a particular source of finance, n is the number of households using a particular source of finance, and N is the total number of households.
In case of hospitalization, NSS classified the various sources of finance as major and second major sources because households might have used more than one source in varying proportions. We have shown the percentage of households using distressed sources to finance hospitalization-related OOPE separately for major and second major sources.
Statistical analysis
Descriptive statistics, two-part model, and multivariable logistic regression were employed in the study.
Two-part model
A two-part model was employed to assess the socio-economic determinants of OOPE. This model is suitable when the outcome variable (i.e. OOPE) is skewed and contains a large number of zero values (Belotti et al., 2015). The first part describes the probability of a household to incur OOPE using a logit model:
where |${y_i} = 0$| indicates that a household has not incurred OOPE on healthcare.
The second part of the model predicts the level of OOPE, conditional on non-zero value. OOPE is estimated using ordinary least square regression, and the dependent variable is the log of OOPE.
Multivariable logistic regression
Multivariable logistic regression was used to estimate the likelihood of households incurring CHE, impoverishment, and distressed financing due to various disease conditions:
In the above equation, |${\pi _i}$| is the probability of occurrence of the binary outcome variable (i.e. incurring CHE, impoverishment, and distressed financing), |${X_1}$| denotes the disease category, and |${X_2} \ldots {X_n}$| represent covariates (economic quintiles, household’s major source of earnings, social group (scheduled castes (SCs), scheduled tribes (STs), other backward classes (OBCs), and others), sector (rural or urban areas), religion (Hinduism, Islam, and others), household size, gender, age, educational status of household head, state, health insurance status, and type of healthcare facility (public or private) from where treatment was sought). We clustered standard errors by the primary sampling unit.
Sample weights provided by the NSS were applied as applicable. Statistical analysis was performed using STATA version 14.1. All amounts reported in Indian rupees (INR) were converted into US dollars (USD) using the average 2018 exchange rate (i.e. USD 1 = INR 68.30).
Results
Childbirth was the most common cause of hospitalization, causing 35.9% of households to seek inpatient care, followed by infections (20.6%), and injuries (7.7%). For outpatient care, households where any member was suffering from infections (31.7%), cardiovascular conditions (16.3%), and endocrine, metabolic, and nutritional conditions (14.9%) sought the highest outpatient care (Supplementary Table 2). Health insurance uptake was dismally low, covering only 15.5% of individuals (Supplementary Table 3).
Out-of-pocket health expenditure
Supplementary Table 4 shows the median monthly OOPE of households by the disease type. Cancer caused the highest OOPE in the case of hospitalization (USD 35.6), followed by genitourinary disorders (USD 18.6), and psychiatric and neurological disorders (USD 16.5), whereas childbirth led to the lowest OOPE (USD 3.3). In the case of outpatient care, obstetric conditions (USD 51.2), cancer (USD 35.1), and genitourinary disorders (USD 26.9) were the leading ailments in terms of OOPE. Compared with public healthcare facilities, the median monthly OOPE for all diseases was invariably higher in private healthcare facilities, regardless of the type of care sought (Supplementary Table 5–8). Supplementary Figure 1 shows the incidence of utilization of public and private healthcare facilities. Nearly 51.0% of hospitalization episodes were sought at public healthcare facilities, whereas only 30.2% of outpatient cases were sought at public facilities. The primary reasons for seeking care at private facilities instead of public ones were the non-availability of doctors or quality of public health facilities not satisfactory, preference for a trusted doctor or hospital, and long waiting times at public health facilities in case of both hospitalization and outpatient care (Supplementary Figure 2).
Table 1 shows the results of the two-part model. The first part (logit regression) revealed that the likelihood of incurring OOPE was higher among households belonging to higher economic quintiles, those headed by the elderly, having insurance coverage, and larger family sizes in the case of both hospitalization and outpatient care (P < 0.05). Households belonging to SCs, OBCs, and other categories were also statistically significantly more likely to incur OOPE compared to STs, regardless of the type of care sought (P < 0.05). By contrast, households primarily earning from casual work were statistically significantly less likely to incur OOPE compared with those earning from self-employment for both hospitalization and outpatient care (P < 0.05). The results for the second part indicated that among the households incurring OOPE, those belonging to higher economic quintiles, SCs, OBCs and other categories, consisting of larger family size, headed by a member aged ≥60 years and having higher education levels were incurring higher OOPE (P < 0.05), irrespective of the type of care sought. Conversely, in the case of both hospitalization and outpatient care, rural households, female-headed households, households earning from casual work, and those with insurance coverage were associated with lower OOPE (P < 0.05).
Results of two-part model
| Hospitalization | Outpatient care | |||
|---|---|---|---|---|
| Background characteristics | Coefficient (Logit) | Coefficient (OLS) | Coefficient (Logit) | Coefficient (OLS) |
| Sector | ||||
| Urban areas® | ||||
| Rural areas | 0.01 [−0.01–0.04] | −0.32* [−0.35 to −0.28] | −0.21* [−0.26 to −0.16] | −0.17* [−0.21 to −0.12] |
| Economic quintiles | ||||
| Quintile 1® | ||||
| Quintile 2 | 0.07* [0.02–0.11] | 0.19* [0.16–0.23] | 0.12* [0.07–0.18] | 0.08* [0.03–0.14] |
| Quintile 3 | 0.14* [0.09–0.18] | 0.32* [0.28–0.36] | 0.21* [0.16–0.27] | 0.15* [0.09–0.20] |
| Quintile 4 | 0.10* [0.05–0.14] | 0.46* [0.41–0.50] | 0.24* [0.18–0.30] | 0.24* [0.18–0.30] |
| Quintile 5 | 0.11* [0.07–0.16] | 0.68* [0.63–0.72] | 0.36* [0.29–0.42] | 0.42* [0.36–0.48] |
| Major sources of household income | ||||
| Self-employment® | ||||
| Regular wage or salary | 0.06* [0.02–0.09] | −0.03* [−0.06 to –0.001] | 0.03 [−0.01–0.07] | −0.04 [−0.08–0.003] |
| Casual labour | −0.10* [−0.14 to −0.07] | −0.22* [−0.25 to −0.19] | −0.14* [−0.18 to −0.09] | −0.09* [−0.13 to −0.04] |
| Others | −0.42* [−0.48 to −0.36] | −0.01 [−0.06–0.05] | −0.06 [−0.12–0.005] | −0.02 [−0.08–0.05] |
| Social group | ||||
| STs® | ||||
| SCs | 0.21* [0.16–0.26] | 0.23* [0.18–0.28] | 0.20* [0.12–0.28] | 0.15* [0.07–0.24] |
| OBCs | 0.16* [0.12–0.21] | 0.40* [0.35–0.45] | 0.20* [0.12–0.28] | 0.24* [0.16–0.32] |
| Others | 0.15* [0.10–0.20] | 0.51* [0.46–0.56] | 0.21* [0.13–0.29] | 0.29* [0.21–0.38] |
| Religion | ||||
| Hinduism® | ||||
| Islam | 0.02 [−0.02–0.05] | −0.10* [−0.14 to −0.06] | 0.17* [0.11–0.23] | −0.01 [−0.06–0.04] |
| Others | 0.02 [−0.03–0.08] | 0.09* [0.04–0.15] | 0.14* [0.06–0.22] | 0.04 [−0.03–0.12] |
| Education level of household head | ||||
| Not literate/Literate without no formal schooling® | ||||
| Up to primary | 0.09* [0.05–0.13] | 0.08* [0.04–0.11] | 0.11* [0.06–0.15] | 0.08* [0.03–0.12] |
| Up to secondary | 0.15* [0.12–0.19] | 0.17* [0.14–0.20] | 0.01 [−0.03–0.06] | 0.16* [0.11–0.21] |
| Up to higher secondary | 0.12* [0.07–0.17] | 0.27* [0.23–0.32] | −0.12* [−0.18 to −0.05] | 0.22* [0.15–0.29] |
| Graduation and above | 0.05 [−0.01–0.10] | 0.45* [0.41–0.50] | −0.16* [−0.23 to −0.10] | 0.32* [0.26–0.39] |
| Age of household head | ||||
| <60 years® | ||||
| ≥60 years | 0.32* [0.28–0.36] | 0.41* [0.38–0.44] | 0.70* [0.66–0.74] | 0.22* [0.18–0.26] |
| Gender of household head | ||||
| Male® | ||||
| Female | −0.22* [−0.27 to −0.18] | −0.08* [−0.12 to −0.04] | 0.02 [−0.03–0.07] | −0.15* [−0.20 to −0.10] |
| Household size | ||||
| Up to five members® | ||||
| >5 members | 0.62* [0.59–0.65] | 0.19* [0.17–0.22] | 0.36* [0.32–0.39] | 0.19* [0.15–0.22] |
| Health insurance status | ||||
| Not covered ® | ||||
| Covered | 0.19* [0.15–0.23] | −0.20* [−0.24 to −0.17] | 0.37* [0.33–0.42] | −0.05* [−0.09 to −0.01] |
| Hospitalization | Outpatient care | |||
|---|---|---|---|---|
| Background characteristics | Coefficient (Logit) | Coefficient (OLS) | Coefficient (Logit) | Coefficient (OLS) |
| Sector | ||||
| Urban areas® | ||||
| Rural areas | 0.01 [−0.01–0.04] | −0.32* [−0.35 to −0.28] | −0.21* [−0.26 to −0.16] | −0.17* [−0.21 to −0.12] |
| Economic quintiles | ||||
| Quintile 1® | ||||
| Quintile 2 | 0.07* [0.02–0.11] | 0.19* [0.16–0.23] | 0.12* [0.07–0.18] | 0.08* [0.03–0.14] |
| Quintile 3 | 0.14* [0.09–0.18] | 0.32* [0.28–0.36] | 0.21* [0.16–0.27] | 0.15* [0.09–0.20] |
| Quintile 4 | 0.10* [0.05–0.14] | 0.46* [0.41–0.50] | 0.24* [0.18–0.30] | 0.24* [0.18–0.30] |
| Quintile 5 | 0.11* [0.07–0.16] | 0.68* [0.63–0.72] | 0.36* [0.29–0.42] | 0.42* [0.36–0.48] |
| Major sources of household income | ||||
| Self-employment® | ||||
| Regular wage or salary | 0.06* [0.02–0.09] | −0.03* [−0.06 to –0.001] | 0.03 [−0.01–0.07] | −0.04 [−0.08–0.003] |
| Casual labour | −0.10* [−0.14 to −0.07] | −0.22* [−0.25 to −0.19] | −0.14* [−0.18 to −0.09] | −0.09* [−0.13 to −0.04] |
| Others | −0.42* [−0.48 to −0.36] | −0.01 [−0.06–0.05] | −0.06 [−0.12–0.005] | −0.02 [−0.08–0.05] |
| Social group | ||||
| STs® | ||||
| SCs | 0.21* [0.16–0.26] | 0.23* [0.18–0.28] | 0.20* [0.12–0.28] | 0.15* [0.07–0.24] |
| OBCs | 0.16* [0.12–0.21] | 0.40* [0.35–0.45] | 0.20* [0.12–0.28] | 0.24* [0.16–0.32] |
| Others | 0.15* [0.10–0.20] | 0.51* [0.46–0.56] | 0.21* [0.13–0.29] | 0.29* [0.21–0.38] |
| Religion | ||||
| Hinduism® | ||||
| Islam | 0.02 [−0.02–0.05] | −0.10* [−0.14 to −0.06] | 0.17* [0.11–0.23] | −0.01 [−0.06–0.04] |
| Others | 0.02 [−0.03–0.08] | 0.09* [0.04–0.15] | 0.14* [0.06–0.22] | 0.04 [−0.03–0.12] |
| Education level of household head | ||||
| Not literate/Literate without no formal schooling® | ||||
| Up to primary | 0.09* [0.05–0.13] | 0.08* [0.04–0.11] | 0.11* [0.06–0.15] | 0.08* [0.03–0.12] |
| Up to secondary | 0.15* [0.12–0.19] | 0.17* [0.14–0.20] | 0.01 [−0.03–0.06] | 0.16* [0.11–0.21] |
| Up to higher secondary | 0.12* [0.07–0.17] | 0.27* [0.23–0.32] | −0.12* [−0.18 to −0.05] | 0.22* [0.15–0.29] |
| Graduation and above | 0.05 [−0.01–0.10] | 0.45* [0.41–0.50] | −0.16* [−0.23 to −0.10] | 0.32* [0.26–0.39] |
| Age of household head | ||||
| <60 years® | ||||
| ≥60 years | 0.32* [0.28–0.36] | 0.41* [0.38–0.44] | 0.70* [0.66–0.74] | 0.22* [0.18–0.26] |
| Gender of household head | ||||
| Male® | ||||
| Female | −0.22* [−0.27 to −0.18] | −0.08* [−0.12 to −0.04] | 0.02 [−0.03–0.07] | −0.15* [−0.20 to −0.10] |
| Household size | ||||
| Up to five members® | ||||
| >5 members | 0.62* [0.59–0.65] | 0.19* [0.17–0.22] | 0.36* [0.32–0.39] | 0.19* [0.15–0.22] |
| Health insurance status | ||||
| Not covered ® | ||||
| Covered | 0.19* [0.15–0.23] | −0.20* [−0.24 to −0.17] | 0.37* [0.33–0.42] | −0.05* [−0.09 to −0.01] |
® denotes Reference category,
P < 0.05 and 95% confidence interval are given in brackets. The state was also taken as one of the variables (result not shown in table).
Results of two-part model
| Hospitalization | Outpatient care | |||
|---|---|---|---|---|
| Background characteristics | Coefficient (Logit) | Coefficient (OLS) | Coefficient (Logit) | Coefficient (OLS) |
| Sector | ||||
| Urban areas® | ||||
| Rural areas | 0.01 [−0.01–0.04] | −0.32* [−0.35 to −0.28] | −0.21* [−0.26 to −0.16] | −0.17* [−0.21 to −0.12] |
| Economic quintiles | ||||
| Quintile 1® | ||||
| Quintile 2 | 0.07* [0.02–0.11] | 0.19* [0.16–0.23] | 0.12* [0.07–0.18] | 0.08* [0.03–0.14] |
| Quintile 3 | 0.14* [0.09–0.18] | 0.32* [0.28–0.36] | 0.21* [0.16–0.27] | 0.15* [0.09–0.20] |
| Quintile 4 | 0.10* [0.05–0.14] | 0.46* [0.41–0.50] | 0.24* [0.18–0.30] | 0.24* [0.18–0.30] |
| Quintile 5 | 0.11* [0.07–0.16] | 0.68* [0.63–0.72] | 0.36* [0.29–0.42] | 0.42* [0.36–0.48] |
| Major sources of household income | ||||
| Self-employment® | ||||
| Regular wage or salary | 0.06* [0.02–0.09] | −0.03* [−0.06 to –0.001] | 0.03 [−0.01–0.07] | −0.04 [−0.08–0.003] |
| Casual labour | −0.10* [−0.14 to −0.07] | −0.22* [−0.25 to −0.19] | −0.14* [−0.18 to −0.09] | −0.09* [−0.13 to −0.04] |
| Others | −0.42* [−0.48 to −0.36] | −0.01 [−0.06–0.05] | −0.06 [−0.12–0.005] | −0.02 [−0.08–0.05] |
| Social group | ||||
| STs® | ||||
| SCs | 0.21* [0.16–0.26] | 0.23* [0.18–0.28] | 0.20* [0.12–0.28] | 0.15* [0.07–0.24] |
| OBCs | 0.16* [0.12–0.21] | 0.40* [0.35–0.45] | 0.20* [0.12–0.28] | 0.24* [0.16–0.32] |
| Others | 0.15* [0.10–0.20] | 0.51* [0.46–0.56] | 0.21* [0.13–0.29] | 0.29* [0.21–0.38] |
| Religion | ||||
| Hinduism® | ||||
| Islam | 0.02 [−0.02–0.05] | −0.10* [−0.14 to −0.06] | 0.17* [0.11–0.23] | −0.01 [−0.06–0.04] |
| Others | 0.02 [−0.03–0.08] | 0.09* [0.04–0.15] | 0.14* [0.06–0.22] | 0.04 [−0.03–0.12] |
| Education level of household head | ||||
| Not literate/Literate without no formal schooling® | ||||
| Up to primary | 0.09* [0.05–0.13] | 0.08* [0.04–0.11] | 0.11* [0.06–0.15] | 0.08* [0.03–0.12] |
| Up to secondary | 0.15* [0.12–0.19] | 0.17* [0.14–0.20] | 0.01 [−0.03–0.06] | 0.16* [0.11–0.21] |
| Up to higher secondary | 0.12* [0.07–0.17] | 0.27* [0.23–0.32] | −0.12* [−0.18 to −0.05] | 0.22* [0.15–0.29] |
| Graduation and above | 0.05 [−0.01–0.10] | 0.45* [0.41–0.50] | −0.16* [−0.23 to −0.10] | 0.32* [0.26–0.39] |
| Age of household head | ||||
| <60 years® | ||||
| ≥60 years | 0.32* [0.28–0.36] | 0.41* [0.38–0.44] | 0.70* [0.66–0.74] | 0.22* [0.18–0.26] |
| Gender of household head | ||||
| Male® | ||||
| Female | −0.22* [−0.27 to −0.18] | −0.08* [−0.12 to −0.04] | 0.02 [−0.03–0.07] | −0.15* [−0.20 to −0.10] |
| Household size | ||||
| Up to five members® | ||||
| >5 members | 0.62* [0.59–0.65] | 0.19* [0.17–0.22] | 0.36* [0.32–0.39] | 0.19* [0.15–0.22] |
| Health insurance status | ||||
| Not covered ® | ||||
| Covered | 0.19* [0.15–0.23] | −0.20* [−0.24 to −0.17] | 0.37* [0.33–0.42] | −0.05* [−0.09 to −0.01] |
| Hospitalization | Outpatient care | |||
|---|---|---|---|---|
| Background characteristics | Coefficient (Logit) | Coefficient (OLS) | Coefficient (Logit) | Coefficient (OLS) |
| Sector | ||||
| Urban areas® | ||||
| Rural areas | 0.01 [−0.01–0.04] | −0.32* [−0.35 to −0.28] | −0.21* [−0.26 to −0.16] | −0.17* [−0.21 to −0.12] |
| Economic quintiles | ||||
| Quintile 1® | ||||
| Quintile 2 | 0.07* [0.02–0.11] | 0.19* [0.16–0.23] | 0.12* [0.07–0.18] | 0.08* [0.03–0.14] |
| Quintile 3 | 0.14* [0.09–0.18] | 0.32* [0.28–0.36] | 0.21* [0.16–0.27] | 0.15* [0.09–0.20] |
| Quintile 4 | 0.10* [0.05–0.14] | 0.46* [0.41–0.50] | 0.24* [0.18–0.30] | 0.24* [0.18–0.30] |
| Quintile 5 | 0.11* [0.07–0.16] | 0.68* [0.63–0.72] | 0.36* [0.29–0.42] | 0.42* [0.36–0.48] |
| Major sources of household income | ||||
| Self-employment® | ||||
| Regular wage or salary | 0.06* [0.02–0.09] | −0.03* [−0.06 to –0.001] | 0.03 [−0.01–0.07] | −0.04 [−0.08–0.003] |
| Casual labour | −0.10* [−0.14 to −0.07] | −0.22* [−0.25 to −0.19] | −0.14* [−0.18 to −0.09] | −0.09* [−0.13 to −0.04] |
| Others | −0.42* [−0.48 to −0.36] | −0.01 [−0.06–0.05] | −0.06 [−0.12–0.005] | −0.02 [−0.08–0.05] |
| Social group | ||||
| STs® | ||||
| SCs | 0.21* [0.16–0.26] | 0.23* [0.18–0.28] | 0.20* [0.12–0.28] | 0.15* [0.07–0.24] |
| OBCs | 0.16* [0.12–0.21] | 0.40* [0.35–0.45] | 0.20* [0.12–0.28] | 0.24* [0.16–0.32] |
| Others | 0.15* [0.10–0.20] | 0.51* [0.46–0.56] | 0.21* [0.13–0.29] | 0.29* [0.21–0.38] |
| Religion | ||||
| Hinduism® | ||||
| Islam | 0.02 [−0.02–0.05] | −0.10* [−0.14 to −0.06] | 0.17* [0.11–0.23] | −0.01 [−0.06–0.04] |
| Others | 0.02 [−0.03–0.08] | 0.09* [0.04–0.15] | 0.14* [0.06–0.22] | 0.04 [−0.03–0.12] |
| Education level of household head | ||||
| Not literate/Literate without no formal schooling® | ||||
| Up to primary | 0.09* [0.05–0.13] | 0.08* [0.04–0.11] | 0.11* [0.06–0.15] | 0.08* [0.03–0.12] |
| Up to secondary | 0.15* [0.12–0.19] | 0.17* [0.14–0.20] | 0.01 [−0.03–0.06] | 0.16* [0.11–0.21] |
| Up to higher secondary | 0.12* [0.07–0.17] | 0.27* [0.23–0.32] | −0.12* [−0.18 to −0.05] | 0.22* [0.15–0.29] |
| Graduation and above | 0.05 [−0.01–0.10] | 0.45* [0.41–0.50] | −0.16* [−0.23 to −0.10] | 0.32* [0.26–0.39] |
| Age of household head | ||||
| <60 years® | ||||
| ≥60 years | 0.32* [0.28–0.36] | 0.41* [0.38–0.44] | 0.70* [0.66–0.74] | 0.22* [0.18–0.26] |
| Gender of household head | ||||
| Male® | ||||
| Female | −0.22* [−0.27 to −0.18] | −0.08* [−0.12 to −0.04] | 0.02 [−0.03–0.07] | −0.15* [−0.20 to −0.10] |
| Household size | ||||
| Up to five members® | ||||
| >5 members | 0.62* [0.59–0.65] | 0.19* [0.17–0.22] | 0.36* [0.32–0.39] | 0.19* [0.15–0.22] |
| Health insurance status | ||||
| Not covered ® | ||||
| Covered | 0.19* [0.15–0.23] | −0.20* [−0.24 to −0.17] | 0.37* [0.33–0.42] | −0.05* [−0.09 to −0.01] |
® denotes Reference category,
P < 0.05 and 95% confidence interval are given in brackets. The state was also taken as one of the variables (result not shown in table).
Share of OOPE in total household consumption expenditure
The share of OOPE in household consumption expenditure was the highest for households in which any member sought cancer treatment (hospitalization: 51.9%, outpatient care: 47.0%, and hospitalization and/or outpatient care: 64.1%) (Supplementary Table 9). Cardiovascular conditions (24.4%), psychiatric and neurological disorders (23.6%), musculoskeletal conditions (23.6%), and injuries (22.3%) also contributed to a significant OOPE share in total household consumption expenditure in the case of hospitalization. For outpatient care, obstetric (36.7%) and genitourinary conditions (35.5%) were the top causes of high OOPE share in household consumption expenditure after cancer, whereas respiratory conditions (10.5%) resulted in the lowest OOPE share. Households in which members underwent hospitalization and outpatient care in private healthcare facilities experienced 5.2 times and 1.7 times higher OOPE share in household consumption expenditure, respectively, than those who availed care in public healthcare facilities (Supplementary Tables 5–8).
Catastrophic health expenditure
Out of all the households who sought hospitalization, outpatient care, and hospitalization and/or outpatient care, 43.1%, 47.8%, and 49.0% of households, respectively, experienced CHE (Figure 1). In the case of hospitalization, households with cancer-afflicted members reported the highest CHE incidence (61.4%), followed by obstetric (54.1%), and psychiatric and neurological disorders (51.6%). In the case of outpatient care, obstetric conditions, genitourinary disorders, injuries, and blood diseases caused CHE among >55% of the respective disease-afflicted households. Notably, among all households reporting CHE, those in which any member was hospitalized for childbirth (36.3%) accounted for the largest share of the total CHE burden, followed by infections (16.0%) and injuries (9.1%), whereas in the case of outpatient care, infections (38.9%) and cardiovascular conditions (11.1%) constituted the largest share of the total CHE burden (Figure 2). The incidence of CHE was higher among households who sought care in private healthcare facilities compared with those treated in public healthcare facilities in the case of both hospitalization (49.7% vs 36.0%) and outpatient care (50.4% vs 39.1%), and the pattern was similar across most disease categories (Figure 3; Supplementary Tables 5–8).
The percentage of households incurring catastrophic health expenditure, impoverishment, and distressed financing
The share of each disease in the total financial burden
The percentage of households incurring catastrophic health expenditure by the type of care (hospitalization and outpatient care) and healthcare facility (public and private)
Supplementary Table 10 shows the results of logistic regression to reveal the impact of different ailments on the likelihood of incurring CHE. Compared with infections, the likelihood of incurring CHE was statistically significantly higher for households in which any member sought hospitalization for cancer [odds ratio (OR): 21.73 (13.30–35.51); P < 0.05] and sought outpatient care for cancer [OR: 6.40 (3.57–11.46); P < 0.05]. In the event of hospitalization, all the diseases (except ear-related ailments) resulted in statistically significantly higher odds of incurring CHE compared with infections (P < 0.05). For outpatient care, the likelihood of incurring CHE was statistically significantly higher for households suffering from cancer, blood diseases, genitourinary disorders, obstetric conditions, injuries, and other disease categories compared to infection-afflicted households (P < 0.05).
Impoverishment impact
Figure 1 shows that 10.7%, 15.0%, and 15.0% of the households who sought hospitalization, outpatient care, and hospitalization and/or outpatient care, respectively, were pushed below the poverty line as a result of OOPE. The percentage of households falling below the poverty line due to hospitalization-related OOPE was the highest for cancer (31.8%), followed by psychiatric and neurological disorders (21.0%), musculoskeletal conditions (17.9%), and injuries (17.8%). Among households where any member sought outpatient care, obstetric conditions, genitourinary disorders, and cancer were among the top three conditions that led to the highest poverty headcount ratio. Of all the households that fell below the poverty line due to OOPE, childbirth (19.6%), injuries (13.7%), and infection-afflicted households (13.5%) accounted for the largest share in the case of hospitalization, while infections (31.3%) and cardiovascular conditions (12.5%) constituted the largest share of households falling below the poverty line in the case of outpatient care (Figure 2). The poverty headcount ratio was higher for households that sought care at private healthcare facilities than public healthcare facilities in the case of both hospitalization (17.5% vs 3.7%) and outpatient care (17.3% vs 10.9%) (Supplementary Tables 5–8).
Logistic regression showed that the likelihood of falling below the poverty line was statistically significantly higher for cancer-affected households for both hospitalization [OR: 9.82 (6.99–13.79); P < 0.05] and outpatient care [OR: 4.13 (2.44–6.98); P < 0.05] compared with households with any infection-afflicted member. In the case of hospitalization, households affected by any ailment (except for eye-, ear-, and skin-related ailments and respiratory issues) demonstrated statistically significantly higher odds of falling below the poverty line compared to infection-afflicted households (P < 0.05). For outpatient care, the odds of falling below the poverty line were statistically significantly higher for households affected by cancer, psychiatric and neurological disorders, gastrointestinal conditions, genitourinary disorders, obstetric conditions, and other disease categories (P < 0.05) (Supplementary Table 10).
Distressed financing
In India, the majority of households primarily relied on income/savings to finance hospitalization-related OOPE (83.9%) and outpatient care (94.7%) (Figure 4), and the incidence of using distressed sources as the major source of finance was relatively low (hospitalization: 16.1% and outpatient care: 5.3%) (Figure 1). However, among 27 769 households who reported using a second major source to finance hospitalization-related OOPE, 78.0% relied on distressed sources (Supplementary Figure 3). Borrowings (43.9%) and contributions from friends and relatives (27.7%) were the most common second major distressed sources used in the event of hospitalization (Supplementary Figure 3). Cancer, psychiatric and neurological disorders, and injuries were the top three disease conditions causing households to rely on distressed financing, irrespective of the type of care sought (Figure 1). Among all households that relied on distressed financing, infection-afflicted households accounted for the largest (28.1%) and second-largest share (17.4%) of the total burden of distressed financing for outpatient care and hospitalization, respectively (Figure 2). Notably, the incidence of distressed financing was higher for all disease categories (except childbirth and others category) for households where members sought outpatient care at public healthcare facilities than private ones. By contrast, in the case of hospitalization, a higher incidence of distressed financing was reported at private healthcare facilities across all disease categories (except ear-related ailments) compared to public healthcare facilities (Supplementary Tables 5–8).
The share of various sources of finance used as coping strategies
Compared with infections, the odds of using distressed sources were statistically significantly higher for households who sought hospitalization [OR: 2.88 (2.42–3.44); P < 0.05] and outpatient care [OR: 3.37 (2.38–4.78); P < 0.05] for cancer. All the ailments (except eye-, ear-, and skin-related ailments) showed statistically significantly higher odds of incurring distressed financing than infections in the case of hospitalization (P < 0.05). For outpatient care, the likelihood of using distressed sources was statistically significantly higher for households where members sought care for cancer, psychiatric and neurological disorders, gastrointestinal conditions, genitourinary disorders, injuries, and other categories compared with infections-afflicted households (P < 0.05) (Supplementary Table 10).
Loss of household income
In addition to the economic burden of financing healthcare, households also experienced indirect costs associated with the loss of earnings due to the inability of the patient or the caregiver to attend work. Supplementary Table 11 shows that hospitalization for cancer caused the highest loss of average household income (USD 94.0), followed by hospitalization for psychiatric and neurological disorders (USD 65.4), and injuries (USD 57.4). For outpatient care, injuries (USD 9.6), obstetric conditions (USD 9.3), and cancer (USD 8.2) resulted in the highest loss of average household earnings.
Proportion of individuals not seeking treatment
Out of all the individuals who reported having an ailment during the last 15 days prior to the survey date, 1.8% of ailing individuals did not seek treatment (Supplementary Table 12). Individuals suffering from ear-related ailments reported the highest incidence of not seeking treatment (19.3%), followed by individuals suffering from eye-related ailments (10.7%), and psychiatric and neurological disorders (5.4%). We also determined the proportion of ailing individuals who did not seek treatment on medical advice and found that 10.1% of all the individuals suffering from an ailment during the last 15 days sought treatment not administered on medical advice (Supplementary Table 12). Although the primary reason for not seeking treatment on medical advice was that the ailment was not considered severe (73.5% of cases), substantial variations were observed across disease categories (Supplementary Figure 4). For instance, 47.0% and 20.8% of individuals suffering from blood diseases and injuries, respectively, reported the non-availability of medical facilities in their neighbourhood as one of the reasons for not seeking treatment on medical advice.
Discussion
This study provides a comprehensive examination of the economic burden of OOPE across 17 disease categories to convey the magnitude of hardships experienced by Indian households incurring OOPE. We found that OOPE is an alarming predicament in India, causing 49% of households to incur CHE and plunging 15% of households below the poverty line due to hospitalization and/or outpatient care. Notably, outpatient care was more burdensome (CHE: 47.8% and impoverishment: 15.0%) than hospitalization (CHE: 43.1% and impoverishment: 10.7%). Nearly 16% of households relied on distressed sources to finance hospitalization-related OOPE. Cancer, genitourinary disorders, psychiatric and neurological disorders, obstetric conditions, and injuries pose a substantial economic burden on households.
India is experiencing a sizeable cancer burden, with 1.39 million new cancer cases registered in the country every year (Indian Council of Medical Research (ICMR), 2022), and as per the World Health Organization (WHO), one in 10 Indians is expected to develop cancer during their lifetime (World Health Organization (WHO), 2020). In tandem with previous studies (Kastor et al., 2018; Rajpal et al., 2018; Boby et al., 2021), we found that cancer led to the highest incidence of CHE, impoverishment, distressed financing, and loss of household earnings in the event of hospitalization. The deleterious effects of high cancer costs are associated with poor quality of life, non-compliance with treatment, debt accumulation and premature entry of younger family members into labour market (Boby et al., 2021), along with physical, psychological, and emotional ramifications. In India, the rural healthcare system is blighted by the paucity of personnel, especially specialists (Government of India (GOI), 2021), and cancer care facilities are largely limited to big cities, causing many patients to travel long distances to seek treatment—a situation that has two negative repercussions: (1) substantial travelling and lodging expenses, coupled with the loss of earnings due to travel, and (2) overloading and long waiting time at major cancer centres (Pramesh et al., 2014). In addition, studies have found that increased travel requirements are associated with more advanced stages of disease at diagnosis, inappropriate treatment, and poor prognosis (Ambroggi et al., 2015). Given the rising cancer burden and the grave economic consequences of cancer care, there is an urgent need for multifaceted policy measures, such as improving diagnostic and imaging equipment, ensuring optimum surgical and radiotherapy infrastructure and palliative care facilities in all publicly funded cancer centres, and promoting cost-effective therapies (Pramesh et al., 2014; Boby et al., 2021). Furthermore, telemedicine must be scaled as it can bring quality healthcare, including specialists to a large proportion of population, decrease the burden of the healthcare system, and increase access to cost-efficient medical services (Chellaiyan et al., 2019; Aashima and Sharma, 2021; Dash et al., 2021).
In tandem with previous studies, we found that cardiovascular conditions not only pose a substantial economic burden on households (Engelgau et al., 2012; Tripathy et al., 2016; Kastor et al., 2018; Yadav et al., 2021c) but also constitute a sizeable share of total financial hardships. For the management of coronary artery disease, percutaneous coronary intervention with coronary stent placement is an important treatment modality (Heart & Stroke, 2022; Johns Hopkins Medicine, 2023). However, in India, substantial unethical price mark-ups (varying from 270% to 1000%) are applied in the stent supply chain, which makes the stent prices exorbitantly high, irrational and restrictive (Government of India (GOI), 2017b; Medical Dialogues, 2017; Pattnaik, 2019). Despite the National Pharmaceutical and Pricing Authority of India introducing a ceiling on stent prices (lowering it by up to 85%) in 2017 (National Pharmaceutical Pricing Authority (NPPA), 2018; The Times Of India, 2018; Pattnaik, 2019), several hospitals have not passed on the full benefits to the patients (The Times Of India, 2018; Pattnaik, 2019). Therefore, it is essential to cap the prices of other accessories (such as guiding catheters, balloons, and guide wires) (National Pharmaceutical Pricing Authority (NPPA), 2018), so that procedures like angioplasty can become more affordable and accessible to patients.
Injuries were among the top-5 ailments across all parameters (i.e. share of OOPE in total household consumption expenditure, CHE, impoverishment, and distressed financing) for hospitalization and/or outpatient care and accounted for one of the largest shares in the total financial hardships in the event of hospitalization. Previous studies have also found that injuries lead to CHE and substantial productivity losses (Prinja et al., 2016; 2019; Yadav et al., 2021b). Notably, in India, road injuries are the leading cause of mortality in the economically active younger age group of 15–39 years (Dandona et al., 2020), highlighting that the burden of injuries far exceeds their immediate medical costs (Government of India (GOI), 2019). Consequently, the Indian government has recently devised a scheme to provide cashless treatment to road accident victims during the golden hour, which is the first hour after an injury when timely medical care can significantly reduce the risk of death (Government of India (GOI), 2019). There is a huge potential for cost-savings if prevention strategies such as mandatory motorcycle helmets and seat belts, speed limits and speed bump installations, and breath testing are effectively implemented in India (Pal et al., 2019; UNICEF, 2022).
In line with previous studies (Singh and Kumar, 2017; Kastor et al., 2018; Yadav et al., 2021c), we found that the brunt of OOPE was lower among households where members sought care in public facilities compared with those treated in private healthcare facilities across most disease categories. However, the lacunae in the public health system, including insufficient healthcare infrastructure, perceived low-quality care, unavailable services, and long waiting times, compel individuals to seek care from private health facilities, resulting in a significant financial burden. As per the latest report published by NITI Aayog (2021), the capacity and quality of healthcare services in India’s public health sector have been constrained due to low public health expenditure, mandating the need for significant and sustained investment to strengthen the public health system. Furthermore, although private hospitals mainly cater to tertiary care services and employ advanced technologies and sub-specialities, these are inadequately monitored by the government, resulting in a plethora of cases of overpricing, unnecessary tests and treatments, and malpractices (Phadke, 2016; The Times of India, 2016; Dehury et al., 2019). Consequently, a substantial portion of financial catastrophes, impoverishment, and indebtedness due to OOPE occurs within the private health system in India (Kastor et al., 2018; Yadav et al., 2021c; Behera and Pradhan, 2021). Conversely, we also found that the incidence of distressed financing was nearly three times higher among households who sought outpatient care in public healthcare facilities compared with those treated in private healthcare facilities. This can be attributed to the higher utilization of public facilities by households belonging to lower economic quintiles (Supplementary Figure 1), who rely on distressed sources to finance even relatively small amounts of outpatient expenses at the subsidized public facilities (Joe, 2015). Previous studies have also found that the incidence of distressed financing is concentrated among the poor households (Joe, 2015; Sangar et al., 2019b; 2020). The overwhelming majority of evidence highlights the need to regulate private healthcare facilities as improved regulation is one of the potential drivers to reduce healthcare costs and improve the quality of care (Selvaraj et al., 2022).
Furthermore, we found a dismally low uptake of health insurance in India, covering only 15.5% of the Indian population. According to the two-part model, insured households showed a higher likelihood of incurring OOPE and a lower OOPE (conditional on having health spendings) compared to the uninsured. This reflects an increase in the utilization of healthcare services as well as lower OOPE among the insured than the uninsured. In India, a majority of the insured population is covered under government-sponsored health insurance (GSHI) schemes, and previous studies have reported that GSHI schemes have improved the utilization of healthcare services (Prinja et al., 2017; Reshmi et al., 2021). Plausible reasons for the increase in utilization can be a genuine reduction in financial barriers, supplier-induced demand, or moral hazard (Prinja et al., 2017; Sengupta and Rooj, 2019). Furthermore, studies have shown mixed results regarding the financial protection provided by the GSHI schemes (Prinja et al., 2017; Reshmi et al., 2021). A recent study found that GSHI schemes provide marginal financial protection to insured households against OOPE and CHE (Ranjan et al., 2018). In India, there is a need to increase awareness about health insurance schemes and its related aspects, such as eligibility criteria, enrolment procedures, stipulated benefits, details about empanelled hospitals, and information on how to avail the benefits, in order to improve the uptake and outcomes of health insurance (Hooda, 2020). Moreover, GSHI schemes are primarily meant to cover the poor and vulnerable population (Hooda, 2020), and social health insurance schemes cover only organized sector employees, and thus, a substantial portion of the Indian population is left with the choice to either arrange private health insurance (constrained by the ability-to-pay premium) or to remain uninsured (NITI Aayog, 2021). Therefore, it is important to increase the affordability of health insurance products, particularly for the missing-middle population (NITI Aayog, 2021).
Notably, we found that financial burden was more pronounced in the case of outpatient care (CHE: 47.8% and impoverishment: 15.0%) compared with hospitalization (CHE: 43.1% and impoverishment: 10.7%). Previous studies have stated that although the cost of hospitalization is higher for any given event (Chatterjee et al., 2013), the overall financial burden and impoverishment are much larger due to outpatient care (Berman et al., 2010), which involves relatively small but more frequent payments. Outpatient care demands policy attention due to a multitude of reasons. First, outpatient care in India is overwhelmingly private, with private healthcare facilities providing ∼70% of outpatient care, thereby causing a substantial economic burden on households (Gupta et al., 2016). Second, medicines and drugs alone constitute a substantial portion of OOPE (>60%) in the case of outpatient care (Sangar et al., 2022). Unfortunately, the limited availability of free or subsidized essential medicines and drugs at public healthcare facilities forces households to buy them from private pharmacies, resulting in higher OOPE or treatment abstention (Maiti et al., 2015). Third, the increasing prevalence of NCDs leads to an increased use of outpatient clinics because chronic illnesses require multiple consultations, regular doctor visits, diagnostic tests, and long-term medication support (Selvaraj et al., 2018; Mukherjee and Chaudhuri, 2020).
Despite all this, most of the GSHI schemes in India mainly cover hospitalization, excluding outpatient care from the ambit of insurance coverage (Selvaraj and Karan, 2012; Hooda, 2020). A study found that removing OOPE for drugs and outpatient visits had the greatest impact on poverty reduction (Shahrawat and Rao, 2012). Thus, to safeguard against financial hardships, it is crucial to expand insurance coverage to include medicines and outpatient care (Shahrawat and Rao, 2012). In India, the new flagship scheme ‘Pradhan Mantri Jan Arogya Yojana (PM-JAY)’ was launched in 2018, with the aim to provide health insurance coverage to 100 million poor and vulnerable households, with a cover of up to USD 7320.6 (INR 500 000) per family per year for secondary and tertiary care hospitalization (National Health Authority, 2022). Although PM-JAY scheme has removed two major limitations of the previous national-level health insurance scheme (Rashtriya Swasthya Bima Yojana), i.e. insurance coverage of a mere USD 438.2 (INR 30 000) per annum and a cap on family size (covering five members only), still it does not cover outpatient care, which can help improve financial protection.
We also found that among all individuals who reported having an ailment during the last 15 days, 1.8% did not seek treatment and 10.1% did not seek treatment on medical advice. Nearly 47.0% of individuals suffering from blood diseases and 20.8% of individuals suffering from injuries reported the non-availability of medical facilities in their neighbourhood as one of the reasons for not seeking treatment on medical advice. Arora et al. (2020) found that >70% of surveyed Indian patients revealed a lack of required facilities in their home state and out-of-state referral (20%) as prominent reasons for seeking cross-border care, a situation resulting in higher travelling costs, loss of labour days, and treatment deferral or abandonment of follow-up care in certain cases. A recent systematic review found that the key reasons for unmet healthcare needs were affordability (20.6%), availability (17.0%), and accessibility (12.2%) (WHO, 2021b). Forgoing care may exacerbate health problems and put the concerned families in a downward spiral of ill health and poverty (Wagstaff, 2002; Petrovic et al., 2021; Rahman et al., 2022). Moreover, even if OOPE is avoided by not seeking care, a household may still experience indirect costs such as loss of earnings if the sick individual or caregiver is unable to attend work. The loss of earnings may be limited if the patient or caregiver works in a formal sector (Alam and Mahal, 2014). However, in a country like India, where 86.8% of the workforce is employed in the informal sector (Oxfam India, 2022), suffering from an ailment can lead to a considerable loss of earnings. We found that cancer, psychiatric and neurological disorders, and injuries were the top three ailments leading to significant losses in household earnings. A study estimated that ∼7% and 23% of middle-aged Indian adults had ever stopped working and had limited paid work, respectively, due to health-related issues (Akhtar et al., 2022). Furthermore, those with chronic diseases were 4% and 11% more likely to stop and limit their work, respectively (Akhtar et al., 2022).
As with all other studies, this study also has a few limitations. First, expenditure data used in the study are subject to potential recall bias, especially for hospitalization incidence where the recall period involves a longer time span of 365 days. Second, the NSS data collects information on self-reported ailments, and a reported diagnosis by a qualified healthcare professional was required only in the case of a few ailments. Surveys that rely on self-reported ailments are likely to underestimate the prevalence of various health conditions and are susceptible to potential recall biases and misclassification of diseases (Engelgau et al., 2012; Patra and Bhise, 2016). For instance, as per the NSS survey, fever was classified under the infection category (constituting 68% and 89% share in infections in the case of hospitalization and outpatient care, respectively) although there can be non-infectious causes of fever as well (Steele et al., 2018), highlighting potential misdiagnosis, particularly in the case of infection category. Third, disregarding coping mechanisms such as borrowings and sale of physical assets, to finance OOPE leads to overestimation of poverty impact (Berman et al., 2010). However, since the NSS health survey does not collect information on how much money is financed through distressed sources, we could not correct this. Overestimation of poverty and CHE may also be caused by our reliance on the NSS health survey, which tends to underestimate the total household consumption expenditure. Lastly, the magnitude of economic burden and disruption of living standards due to health expenditures should be ascertained using longitudinal data; however, in the absence of such data, cross-sectional studies such as ours can provide potential estimates of the financial impact of OOPE.
Conclusion
The study highlights the colossal economic impact of OOPE and the associated financial catastrophe and impoverishment faced by Indian households suffering from any type of illness. Concerted efforts, such as strengthening public healthcare facilities, increasing the uptake of health insurance, designing broader insurance packages to cover outpatient care, and ensuring affordability and availability of essential medicines, are imperative to augment financial risk protection. Moreover, even though cancer causes copious financial burdens among those who have the disease and seek treatment for it, policymakers should get to the bottom of what comprises infections (given the prospect of substantial misclassification and/or misdiagnosis in this category) and address the high spending on cardiovascular diseases and injuries, which constitute a sizeable share of the total financial hardships. Comprehensive disease-specific insurance packages (The Economic Times, 2020) should be designed for high-cost ailments such as cancer, psychiatric and neurological disorders, and genitourinary disorders. Furthermore, there is a need for improved regulation of the private health sector and to put standard treatment guidelines in place. For long-term sustainability, policymakers must prioritize health promotion and disease prevention strategies, as increasing life expectancy, ageing population, westernization, and motorization will further aggravate the burden of NCDs and injuries in India. Implementation of robust and effective evidence-based health promotion programmes holds the potential to significantly improve people’s health and reduce the financial burden they face. To achieve UHC and SDG goals, the epidemic of chronic diseases and injuries should be a political priority and central to the national consciousness.
Supplementary data
Supplementary data are available at Health Policy and Planning online.
Data availability
The research was conducted using the 75th round of the National Sample Survey on Health, which is available in the public domain.
Funding
This study has received no specific grant from any funding agency in public, commercial or not-for-profit sectors.
Author contributions
M.N. has contributed to the conception of the work, data analysis and interpretation, drafting the article, critical revision of the article, and final approval of the version to be submitted.
R.S. has contributed to the conception of the work, critical revision of the article, and final approval of the version to be submitted.
Reflexivity statement
The authors of the study include a female and a male researcher and span multiple levels of seniority. One author has research interest in the area of health expenditure and the other author has extensive experience and specialization in health economics, public health and epidemiology in Asia and at a global level.
Ethical approval
The research was conducted using data available in the public domain and did not include any human participants or animals. Therefore, no ethical approvals were required.
Conflict of interest statement
The authors declare no competing interests.
