Out-of-pocket payments & catastrophic healthcare expenditure for non-communicable diseases: Results of a State-wide STEPS survey in north India

Study design

A descriptive, cross-sectional research design was adopted for the present study conducted in the North Indian State of Punjab. Data from 576 individuals were collected using a convenience sampling approach through a survey questionnaire from healthcare facilities like public hospitals, private hospitals, etc., from March 6 to September 21, 2023. The out-of-pocket expenses were quantified for both outpatient and inpatient care, as shown in figure. The occurrence of outpatient care was assessed using a two-month recall at the time of the survey. In contrast, the occurrence of inpatient care was evaluated within a one-yr reference period at the time of the survey.

Close

Figure.

Framework on the healthcare utilisation and economic enquiry of NCDs. HCU, healthcare utilisation; NCDs, non-communicable diseases. ^*Multiple responses are possible.

Export to PPT

The rationale of the study

According to the Indian Council of Medical Research⁹, Punjab’s affluence and consuming habits make it the State with the highest rates of obesity and hypertension. The proportion of NCDs to the total disease burden is 52.1 per cent in Punjab. Moreover, Punjab witnessed a drastic shift in Disability-Adjusted life years (DALYs) from 1990 to 2016. NCDs such as ischaemic heart disease, chronic obstructive pulmonary disease (COPD), and diabetes are among the top three leading causes of fatality and debility in Punjab⁹, with a mean percentage change of 17.1, 4, and 3.9 per cent, respectively, in 2016.

Study variables

The method adopted by us identified direct and indirect costs and their total expenditure as essential components for assessing the economic burden of NCDs. Direct costs consist of expenses related to resource usage because they originate from diagnosis and treatment followed by healthcare procedures²⁴. Direct costs are derived from medical and non-medical costs; their sum calculation produces the total direct cost. Direct medical expenses consist of payment for physician appointments, testing fees, drug prescriptions, and their required supplies. Hospitalisation costs, costs of patient transport to medical facilities, and expenditures on food items and materials make up direct non-medical expenses^16,25.

The loss of worker productivity affects patients with illness and their household members who function as caregivers, resulting in indirect costs. Wage loss data was documented by combining outpatient and inpatient appointment attendance periods through interview responses from patients and their acquaintances¹⁷. The human capital approach^26,27, served as the methodology followed throughout this investigation. Health expenses become catastrophic whenever individuals spend more on out-of-pocket medical costs than a predetermined share of their annual spending. This study investigated threshold levels of 5, 10, 20, and 30 per cent relating to the incidence and intensity of total household expenditures.

Developing research instrument

This study used a self-administered research instrument adapted from the ‘WHO STEP wise Surveillance’ manual. Before the primary survey, a four-stage procedure was followed to develop and adapt measures regarding scale development²¹; (i) designing the instrument, (ii) modifications in research items, (iii) psychometric trait testing, and (iv) reliability test. Demographic characteristics are presented in the first section, followed by section two, which discusses the type of NCD the respondents were diagnosed with. Lastly, section 3 focuses on both inpatient and outpatient costs associated with NCDs.

Statistical analysis

Descriptive analyses were performed in the present study to estimate the socio-demographic characteristics and out-of-pocket payments. Mean, median, and range were calculated for costs by the components of direct and indirect costs. Moreover, catastrophic healthcare expenditure (CHE) was measured to estimate the incidence and intensity of healthcare expenditures²⁸. The headcount and overshoot method was employed to assess the incidence of NCDs^29,30, and mean positive overshoot was used to assess the intensity of healthcare expenditures²⁹. The headcount ratio and overshoot technique collectively offer significant insights into the cost of illness (COI) research, facilitating a thorough comprehension of both the prevalence and severity of the economic burden resulting from disease. The headcount ratio offers breadth by indicating who and how many individuals are affected, whereas the overshoot method provides depth by assessing the severity of their impact. Together, these methods create a comprehensive framework for understanding and addressing the economic burdens of illness. The mean positive overshoot technique enhances the study by quantifying the average economic distress of individuals over a cost threshold, thereby guiding targeted interventions, reducing disparities, and offering a more thorough assessment of illness-related financial burdens. Finally, a logistic regression model was used to assess the influence of several independent factors provided concurrently to predict membership in one of the two dependent variable categories³¹.

Where, E = Indicator equal to one, if Oi/Yi>z or else Zero; Pi = Out-of-Pocket expenditure of household; Yi = Household income; Z = Catastrophic threshold level; N = Sample size.

Where,

Socio-demographic profile

Table I highlights the socio-demographic profile of the study respondents from Punjab. Of the 576 respondents, a sizable proportion diagnosed with NCDs was male. The majority of the NCDs were identified as lifestyle diseases, and most of the respondents were from urban regions. The analysis revealed that a sizable proportion of respondents diagnosed with any NCD were from the age groups of 41-60 yr and above 60 yr. Moreover, a history of NCDs in the family was reported by 54 per cent of the respondents. Among diverse types of NCDs, a large proportion of respondents were diagnosed with diabetes (31%), followed by hypertension (24%), CVD (19%), stroke (15%), and COPD (11%).

Healthcare utilisation of non-communicable disease in Punjab

Table II elucidates the healthcare utilisation of NCDs in Punjab. It was found that a large proportion of respondents used allopathic treatment (76%), followed by home remedies (18%), ayurveda (5%), and homoeopathy (1%). The study results revealed that 48 per cent of respondents utilised private healthcare facilities, 26 per cent utilised public healthcare facilities, 19 per cent preferred chemists, and seven per cent utilised homoeopathy care. Lastly, of all the 576 respondents, a substantial proportion of 259 individuals own private insurance, followed by 34 per cent with governmental (AB-MMSBY) insurance, and 21 per cent do not know about the health insurance they possess. The mean distance covered to access healthcare facilities was 5.49 km.

Annual outpatient and inpatient costs of NCDs by component of direct and indirect costs

The results of the annual outpatient and inpatient cost of NCDs by components have been shown in table III. Respondents with cardiovascular disease incurred the highest mean direct outpatient cost of ₹48,000 (3800 - 90,000) and mean indirect cost of ₹6000 (500 - 7500). It was found that respondents with stroke experienced the second-highest mean direct outpatient cost of ₹40,300 (4000 - 85,000). In contrast, respondents with COPD incurred the second-highest mean indirect cost (₹4909). The results revealed that patients with hypertension incurred the least direct and indirect outpatient costs. Expenditure on medicines constituted the most significant cost element among all NCDs, succeeded by expenses related to diagnostics, consultation fees, and transportation costs.

The analysis revealed that of the total 576 study respondents, only 209 utilised inpatient healthcare. The mean direct inpatient cost incurred by cardiovascular disease respondents was the highest at ₹90,000 (2500 - 1,50,000), followed by COPD, hypertension, stroke, and diabetes, as exhibited in table III. Under indirect inpatient cost, respondents with COPD incurred the highest mean cost of ₹5000 (500 - 10000), followed by cardiovascular disease, diabetes, stroke, and hypertension.

Incidence and intensity of catastrophic healthcare expenditure (CHE) across various NCDs

The incidence and intensity of CHE across various NCDs in Punjab have been presented in table IV. At any threshold level, CHE incidence (headcount) was the highest among diabetic patients. The respondents with COPD experienced a minor incidence of CHE at 20 and 30 per cent threshold levels. The intensity (overshoot) of CHE with diabetes and hypertension was above eight per cent at the threshold level of five per cent, whereas respondents diagnosed with stroke experienced the least intensity (overshoot) of CHE at any given threshold level (i.e., from 5 to 30%). The analysis revealed that respondents diagnosed with hypertension suffered the second highest ‘overshoot’ and highest ‘mean positive overshoot’ intensity at five per cent and 10 per cent threshold levels, respectively. Similarly, respondents diagnosed with stroke experience the second-highest intensity (mean positive overshoot) of 11.43 per cent at a threshold level of five per cent.

Determinants of catastrophic healthcare expenditure of NCDs in Punjab

The critical determinants of CHEs of NCDs were identified with logistic regression, as shown in table V. Based on the ‘Omnibus Test of Model’, the value of 0.000 exhibits the models as statistically significant. Hosmer and Lemeshow’s goodness of fit test shows that the model is an excellent fit for the data as the P value is more significant than 0.05. At a 30 per cent threshold level, respondents diagnosed with NCDs observed a statistically significant association with CHE and major socio-demographic variables such as gender, age, income, work status, education, and complications. As shown in table V, the probability of CHE was higher among male respondents diagnosed with hypertension, diabetes, and cardiovascular disease.

On the contrary, the odds of incurring CHE were lower among male respondents diagnosed with COPD and stroke. The odds of incurring CHE were two times higher among the respondents diagnosed with hypertension and COPD and having a primary level of education in Punjab. It was found that the odds of incurring CHE were thirteen times higher among diabetic homemakers as compared to other respondents diagnosed with hypertension, stroke, CVD, and COPD. The analysis revealed that respondents with low-income levels were likelier to experience CHE at any given type of NCD than respondents with high income. The logistic regression results revealed that the probability of incurring CHE was higher among respondents diagnosed with diabetes, COPD, and stroke as the major NCD complications. Lastly, the odds of incurring CHE were high among diabetic and CVD respondents with a household size of 4-6 members.