Hospitalization & health expenditure in Odisha: Evidence from National Sample Survey (1995-2014)

Every year, millions of households are pushed to poverty on account of high spending in healthcare services, especially in low- and middle-income countries1-3. Financing healthcare services through out-of-pocket (OOP) payments is the general norm in many Asian countries including India4. Nearly eight per cent of population in India are pushed to below the poverty line due to OOP payments for healthcare services5. Evidently, public health spending in India is abysmally low as compared to many developing nations given that the health budget of India is only one per cent of its gross domestic product6. More importantly, about 67 per cent of the health expenses are managed by the households, making India one of the highly privatized healthcare systems in the world7. The ever-rising cost of healthcare services remains a grave concern in India more so in the poor states.

Although India witnessed a rapid increase in non-communicable diseases (NCDs) indicating a pattern of an epidemiological transition but an overwhelming burden of communicable diseases is suggestive of a double burden of diseases8. With the changing morbidity profile, hospitalization and health expenditure patterns are also expected to alter and pose a serious threat, especially to the socio-economically weaker sections of the society in the absence of adequate health financing mechanism. It was shown that proportion of OOP expenditure for hospitalization was significantly higher among the economically weaker sections of the society9,10.

Odisha is an economically poor State and catastrophic health spending in Odisha is one of the highest in India11. Despite the majority of the healthcare institutions in Odisha are in the public sector, OOP expenditure remained high12. Odisha has a high percentage of older population in India (10%) and an elevated rate of hospitalization and health expenditure among them is a growing concern13. In India, health being a State subject, the majority of healthcare planning and interventions are state specific. Therefore, State-level analysis is important to inform health policy.

Population-based large scale surveys are an important source of evidence to identify the emerging health concerns and thereby effectively design and implement health programmes. While there are several studies on catastrophic health expenditure and its effect on household impoverishment at the national level3,5,9, but studies at the State level are sparse. Considering Odisha has a high poverty rate and high catastrophic health spending11, examining the interplay between the socio-economic and demographic characteristics of population with hospitalization and health spending over a period of time may provide new insights. This study was undertaken to investigate hospitalization and patient’s health expenditure in Odisha over the two time periods (1995-2014) using the National Sample Survey (NSS) data.

Material & Methods

The data from the 52^nd and 71^st round of NSS were used, which were available in public domain (http://www.icssrdataservice.in, accessible on request). The total sample for Odisha comprised 21,723 individuals in 52^nd round and 11576 in 71^st round of NSS. The survey design including the methods of data collection and other survey-related information can be found elsewhere14,15. As the objective of the study was to assess the variations of socio-economic and demographic factors on hospitalization and health spending in the long run (over 2 decades), the 52^nd and 71^st round of NSS were chosen.

Andersen’s behavioural model of healthcare utilization was used to examine the predisposing, enabling and need factors16. Andersen’s model postulates that the utilization of healthcare services by the households is a function of their predisposition to use services that emerge by naturally possessed conditions such as geographical accessibility, gender, caste and any other affiliations17. The enabling factors are further dependent on the economic positions, educational attainment, place of residence and other basic infrastructure-related factors that facilitate access. At the end, the utilization of healthcare services is determined by the need of the households based on health condition and disease burden.

The rate of hospitalization and health expenditure were examined by the socio-economic and demographic characteristics as the predisposing and enabling factors for five distinct morbidity categories: infectious diseases, cardiovascular diseases (CVDs), NCDs, disability and other diseases. The morbidity classifications are based on the 10^th revision of the international classification of disease18 (Supplementary Table). The references period for hospitalization cases was 365 days preceding the survey. Health expenditure is provided in Indian rupees ₹ and for comparison purpose, the expenditure value of the year 1995 was adjusted for inflation as per the Reserve Bank of India by taking base year as 201419.

The socio-economic and demographic characteristics used in this study were gender (male, female), place of residence (rural, urban), age groups (0-14, 15-34, 35-59, 60 yr and above), level of education (illiterate, primary, higher secondary, graduate and above), marital status (never married, currently married, widowed/separated), household size (1-5 members, 6-7 members, 8 and above members), caste (ST/SC, OBC, other), religion (Hindu, others), monthly per capital expenditure, i.e. MPCE (poorest, poor, medium, rich, richest), regions (northern, coastal, southern), whether health service was sought in private or public facility, and whether had insurance coverage or not. Hospitalization rate was calculated using the following formula:

Where, hi= Number of persons hospitalized in the past 365 days; pi= Total number of persons alive in the sample households.

Bivariate analysis was conducted to examine hospitalization rate by socio-economic covariates and health expenditure was estimated for two different time period i.e. year 1995 and 2014. Data analysis was conducted using STATA 12 (StataCorp LP, College Station, Texas, USA).

Results

Hospitalization by socio-economic and demographic characteristics: Hospitalization rate in Odisha nearly increased by three times between 1995 and 2014 from 12 per thousand to 30 per thousand population. Hospitalization for CVDs followed by disability and NCDs showed a steep increase between 1995 and 2014 (Table I). Further, hospitalization rates for CVDs and NCDs, especially among the older population (aged 60 yr and above) and the widowed or separated were the highest. For instance, seven of 1000 older persons and widowed or separated reported hospitalization due to CVDs and 16 of 1000 for NCDs in the year 2014. On the other hand, hospitalization for other diseases remained consistently higher in both the time periods. Other disease category comprised the diseases other than the remaining four categories, i.e. infectious, CVDs, NCDs and disability. In 2014, hospitalization for other diseases was higher among widowed (17/1000), other religion (16/1000), and in Southern region of Odisha (13 in 1000). Overall, hospitalization was higher among urban resident, older population, widowed, those who were economically well-off and lived in the coastal region of Odisha as opposed to their counterparts.

Patient’s health expenditure by types of morbidities: The health expenditure for hospitalization in Odisha doubled in the period between 1995 and 2014 (Table II), of which, expenditure for disability-related illness increased three folds and other morbidity by over two times. Although health expenditure for CVDs was the highest in 1995, in 2014, health expenditure for disability was the highest followed by CVDs and NCDs. Remarkably, health expenditure almost reduced by half for infectious disease and CVDs between 1995 and 2014. Health expenditure for most of the diseases was higher among male population, urban resident, in the age group 15-59, among never married, persons with higher level of education, affluent households, and those who were hospitalized in private health facilities. The health expenditure for CVDs among those who were hospitalized in private health facility, richest MPCE households, and with education level of graduate and above crossed rupees 0.1 million in the year 1995 and 2014 . Further, irrespective of the type of morbidities, health expenditure was generally higher among wealthy households, in the coastal region of Odisha and those who were hospitalized in the private health facility in 1995 and 2014 with few exceptions.

A comparison of hospitalization trend between Odisha and at the all India level indicated that total hospitalization rate in Odisha was lower than the national average in 1995 and 2014. However, the 2018 data suggested that total hospitalization in Odisha surpassed the national average (43/1000 vs. 39/1000 population) (Fig. 1). On the other hand, total health expenditure in Odisha remained consistently lower than the national average over a period of time (Fig. 2).

Close

Fig. 1

Hospitalization per 1000 population (during last 365 days) by type of diseases in Odisha and India, 1995-2018). Source: Refs 14,15.

Export to PPT

Close

Fig. 2

Total health expenditure in Odisha and India, 1995-2018. Source: Refs 14,15.

Export to PPT

Discussion

This study revealed that hospitalization in Odisha increased nearly three folds and health expenditure by more than two times between 1995 to 2014. Hospitalization for other diseases remained consistently higher over the period of time. However, the increase in hospitalization for CVDs, disability and NCDs was considerably higher between the last two decades. Similarly, health expenditure for disability was the highest in 2014 and the increase was three times higher than 1995. A decreasing pattern of health expenditure for infectious disease and CVDs during the same period was encouraging.

Hospitalization in Odisha was generally lower than the national average with a few exceptions, but health expenditure was noticeably lower than the national average. Further, the socio-economic and demographic divides in the hospitalization rate and health expenditure were evident. Predisposing factors such as age and marital status played an important role in hospitalization whereas enabling factors likely determined the health expenditure.

The results showed a clear distinction in hospitalization between older population and others as shown earlier9,10. The hospitalization rate among the older population was two times more than others. Similarly, widowed population were exposed to alleviated risk of hospitalization20. Irrespective of the burden of diseases, the health expenditure among the richest MPCE was consistently higher than others21. People with higher level of education as compared to others spent the highest in hospitalization perhaps because they perceived health to be important. Similarly, higher health spending among male, working age population, wealthy households were indicative of ability to pay as well as gender-biased attitude.

Of the three regions in Odisha, the coastal region is considered as a developed region22. The study findings indicated that CVDs and NCDs-related hospitalization were higher in the Coastal region. This trend may indicate availability of healthcare services in the region23. Moreover, higher health expenditure in the coastal region may also indicate the ability to pay as a possible determinant of health expenditure24,25. The data indicated that fever comprised a major part in the other disease categories (data not shown) [In the 71^st round of NSS (2014)], morbidity schedule introduced ‘all other fevers’ (includes malaria, typhoid and fevers of unknown origin) in the other disease category; for comparison purpose, fever was classified in other disease in 52^nd round of NSS. Tropical diseases such as malaria and other vector-borne diseases are quite prevalent in this region26-28. Although increased rate of hospitalization is a positive sign for curative healthcare point of view, but from a public health perspective, it calls for better intervention strategies to control vector-borne diseases29.

The increasing rate of hospitalization for disability-related illness as well as the highest spending on it may be due to road traffic accidents and injuries30. On the other hand, greater financial burden for accidents and injuries is particularly high in poorer households31. Trauma care facilities and a better financial protection mechanism for disability-related illness in Odisha are essential. The reduction in health expenditure for infectious disease and CVDs over the years is encouraging. However, considering treatment cost for CVDs being commonly high, more studies are needed to understand the reason for this.

The findings needs to be interpreted in light of a few limitations. Health seeking behaviour including accessibility and affordability of health services considerably influence health service utilization. However, these factors were not taken into account in this study because this study was aimed to understand the changing role of socio-economic and demographic determinants (predisposing and enabling factors) to use health services. Second, in India, religion and caste are closely linked to class divisions, however, due to lack of information on religion and other backward class, the influence of these could not be determined in the 52^nd round. We clubbed the religion variable considering low sample size for other religions.

The recently launched Biju Swasthya Kalyan Yojana (BSKY) in Odisha is a welcome step for universal health coverage. The experiences from the past have shown that such health insurance schemes have not yielded the desired result. Nevertheless, the new schemes in the State must recognize the unique vulnerabilities of older and widowed population for a better health finance mechanism for them. However, given that percentage share of older population in Odisha is high and expected to increase further, health system needs to prepare for gerontological care by equipping health facilities and training specialized human resources to cater to the needs of the older populations. Hospitalisation depends on epidemiological pattern, availability and access of services at different levels and therefore, it is necessary to take in to account such factors apart from the socio-economic determinants to intervene policy. The health policy of the State needs to prioritise health financing mechanism in reducing patient’s expenditure and optimising hospitalisation rates by ensuring a responsive health system.