Costs and Affordability of COVID-19 Testing and Treatment in India

The COVID-19 pandemic has triggered several underlying vulnerabilities with potentially far reaching consequences in low- and middle-income countries (LMICs) like India. Evidence of physical and socio-economic vulnerabilities caused by the pandemic are emerging rapidly, but one area that has received limited attention so far, is the nancial vulnerability COVID-19 causes for households and the government. This paper aims to assess the nancial burden imposed on governments and households and the ability of households to afford the required medical costs. and 232 days of work respectively. Thus, affordability of COVID-19 services is far worse among casual workers, wherein annual wage falls short of ICU hospitalization cost for 90% of workers and hospital isolation costs for 48% of workers. Among self-employed workers, the proportions whose annual wages could not afford ICU hospitalization and home isolation were 66% and 27% respectively. For regular employees, we found that for 51% and 15% of them, their annual salaries could not afford to pay for ICU admission or hospital isolation respectively.


Introduction
Since the outbreak of the COVID-19 pandemic, India has reported over 30 million total cases and about 281 deaths per million population, making it one of the most severely impacted countries in the world.
[i] The pandemic has resulted in both economic distress as well as loss of human lives. The lockdown mandates have cost millions of workers their jobs, along with large scale displacement and migration of laborers back to rural villages across the country. The International Monetary Fund (IMF) predicts that India's economy will contract by 11.2%, one of its worst performances in decades. [ii] The pandemic has also disrupted health services resulting in lower immunization rates and decreased access to treatment of various diseases including tuberculosis, malaria, and several non-communicable diseases.
[iii] [iv] In response, the government of India introduced policies to mitigate the economic and health impact of the pandemic.
[v] , [vi] To address COVID-19, the government set up a three-tiered COVID-19 healthcare delivery system that triages patients based on severity of illness (See Box 1). [vii] In response to the disproportionate effect of the lockdowns on private health facilities, the public healthcare system expanded both preventive and curative services for COVID-19. However, India's public health system is plagued by several challenges ranging from workforce shortages, to poor infrastructure and quality of services. [viii] , [ix] Moreover, the high reliance on the private sector for health care, coupled with the low levels of health insurance coverage put many households at great risk of nancial hardships due to COVID-related treatment costs. [x] Box 1: India's three-tiered health system response for COVID-19 treatment India has developed a three-tiered structure for pandemic preparedness to quarantine, isolate, and treat COVID-19 cases, through dedicated COVID care centres, COVID health centres and COVID care hospitals in the public and private sector. vi This three-tier strategy for managing COVID patients has been followed almost uniformly across the country. , far below the global average of 6% and the 2.5% recommended by the High-Level Expert Group for Universal Health Coverage in India. [xiii] As per World Bank estimates, even without the COVID pandemic, 17% of the total population faced catastrophic health expenditures, spending more than 10% of their household income on health.
[xiv] The expanded coverage of COVID-19 treatment under government sponsored programs is bound to put further scal strain on central and state governments to maintain health services while ensuring adequate nancial protection for the public. Given the severe impact of health care costs on households in India, which are further exacerbated by the COVID-19 pandemic, this paper investigates: (i) the COVID-19 related nancial costs to government and households; and ii) the affordability of COVID-19 testing and treatment costs for households.

Methods
Our analysis focused on four COVID-19 service packages including: (i) COVID-19 testing, (ii) Home isolation for COVID-19, (iii) Hospital isolation for COVID-19, and (iv) Intensive Care Unit (ICU) hospitalization for COVID-19, with and without ventilator. These services are performed in India in accordance with the national pandemic preparedness plan to quarantine, isolate, and treat COVID-19 cases, through dedicated COVID-19 care centers, COVID-19 health centers, and COVID-19 care hospitals in the public and private sector[i]. Table 1 and S1 Appendix section A1 describes details of the protocols, and levels of care. For each service package, we estimated the unit cost of providing the service, the affordability for Indian households by socioeconomic groups, and the total cost to the government of India of providing the service given the burden of COVID-19 cases. Our analysis included only nancial costs, not economic or opportunity costs as the former is a better representation of the per-episode nancial burden in the short-run.
[ii] We obtained data for this analysis from publicly accessible sources, therefore, this study quali ed as non-human subjects research and did not require ethics approval.

Estimating unit costs of COVID-19 intervention packages
We conducted a top-down cost estimation using publicly available data from national-and state-level sources including, the government-insurance package rates, the government capped package rates for private facilities, and the National Sample Survey The mean costs of providing each COVID-19 intervention outlined above was estimated in 2020 Indian Rupees (INR) and US Dollars. Unit cost for COVID-19 testing was obtained from average cost of testing as mandated by state governments (see methods section in S1 Appendix Section A2). For symptomatic patients requiring home isolation for COVID-19, we used the cost of home-treatment for fever and the common-cold as proxies. Unit-level per episode average expenditures were derived from NSS data that capture mean spending per person for common cold and fever -code '04' for fever and code '36' for respiratory conditions. Since NSS was conducted in 2018, the estimated cost was then adjusted for in ation to report expenditure in 2020 equivalents (See S1 Appendix section A2 for details). To estimate unit costs for hospital-based services, we analyzed the government-insurance package rates and the government-capped package rates from eighteen major states. These approved package rates represent how much the government-sponsored health insurance will reimburse health facilities for providing COVID-19 care while governmentcapped rates re ect nancial burden on households. The rates vary by state so we used the average cost estimate. Further details of unit cost estimation are highlighted in S1 Appendix Section A2.

Assessing nancial burden of COVID-19 intervention packages
We measured nancial burden by estimating affordability of treatment and testing for households by comparing unit costs with the equivalent of individual daily wages, thus the cost of using each service was transformed to the equivalent of number of days needed to work in order to afford the service. Separate estimates were derived for each of the three occupational classi cations de ned by the Government of India, namely: casual worker, regular employees and self-employed.
[vi] We used proxy measures for affordability by calculating the number of days required to pay for the cost of testing and treatment. Work-days and wages/salaries for three types of workers were measured based on Periodic Labor Force Surveys (PLFS, 2017-18). PLFS is a national level survey, with a sample size of 102,113 households and 433,339 persons. The PLFS utilized a strati ed multi-stage random sampling method. Details of our estimation methods are described in S1 Appendix Section A3.
Estimating total cost of COVID-19 services borne by households and the Government The nancial resources required to meet the cost of providing testing and treatment was estimated by multiplying the total number of COVID-19 cases by the average cost of the COVID-19 packages. We estimated total costs per annum as well as one quarter total costs -rst wave (April 2020 to March 2021) and (April 2021 to June 2021). Total number of COVID-19 cases and tests administered for the period under consideration were obtained from o cial estimates [vii]. Using data from the US Centers for Disease Control and Prevention, we assumed that 14% of COVID-19 cases will require hospitalization for severe disease conditions while 2% will require ICU hospitalization (with or without mechanical ventilation) for very severe disease.
[viii] We further assumed that pre-COVID patterns of using public vs. private health facilities did not change and will therefore mirror service-use patterns as in the latest national health survey (NSSO) 2017-18. By contrast, the number of cases tested in public and private facilities depends on the distribution of testing facilities.

Results
Unit costs of COVID-19 services   Figure 2 shows the proportion of employees whose annual salary/wage is lower than the cost of seeking each COVID-19 service. Among regular salaried employees, about 50% have annual incomes that are less than the per episode cost for ICU hospitalization for COVID-19 and while 14% have annual incomes less than per episode cost for hospital isolation. As far as self-employed workers are concerned, the respective proportions that have incomes lower than cost of ICU hospitalization and hospital isolation are 66% and 27% respectively, and for casual workers, the proportions were larger at 86% and 43%. Hospitalizations and ICU support (with and without ventilator use) account for approximately 3%. 3. The annual number of COVID-19 cases tested is based on share of private and public laboratories, which are currently 53% and 47% respectively, with households and government bearing costs.

Total costs borne by households and the Government for COVID-19 testing and treatment
4. Disaggregation in respect to government and households' treatment for hospitalization is based on NSS distribution of inpatient in the proportion of 42% and 58% respectively.

Discussion
We estimated the unit costs incurred for COVID-19 testing and treatment interventions, the affordability and nancial burden to households, and the total costs borne by governments and households. Our ndings showed that on average the unit cost of COVID-19 interventions ranged from a low of Rs. 2,229 (US$ 30) for testing to a high of Rs. 140,000 (US$ 1,907) per episode for ICU admission. While COVID-19 testing combined with home isolation were relatively more affordable ( ve days, seven days, and 11 days of work needed by regular employees, self-employed and casual workers respectively), hospital isolation and ICU admission were not affordable (232 days, 318 days, and 481 days respectively for regular employees, self-employed and casual workers). Casual workers therefore are the most impacted implying that their annual wage fell short of per episode cost for 90% of workers when seeking treatment for ICU hospitalization and 48% of workers while receiving treatment for hospital isolation. However, since fewer people require ICU hospitalization or hospital isolation, the nancial impact is concentrated in the households affected and not generalized to the entire population. By contrast, a signi cant portion of the population are required to home isolate whenever there is suspicion of an infection and this creates a problem for casual workers in particular, some of which earn annual incomes that fall below the cost of home isolation.
The estimated total costs to households (i.e., out-of-pocket payments) for COVID-19 testing and treatment over the period April 2020 This study was devoted to measuring medical costs of testing and treatment. We did not seek to generate comprehensive economic costs of the pandemic, lockdowns, productivity loss and wage loss associated with the disease conditions. Future research must address and demonstrate comprehensive economic costs of the pandemic and the imperative for pandemic preparedness besides the need to prevent lockdowns and other stringent economic measures that can cripple vulnerable households. In the absence of a national household survey, future research must look for measurements around catastrophe and impoverishment that can help to further quantify the nancial vulnerabilities on households. Quantifying foregone resource allocation for other basic needs within a household must also be assessed in order to understand the extent of nancial vulnerabilities.

Conclusion
Use of COVID-19 prevention and treatment services is exerting enormous nancial burdens on household and governments in India.
In addition to dealing with job losses, household members have to spend a high number of daily wages equivalents in order to afford testing, home isolation, hospital isolation and ICU care. To mitigate this effect, it is imperative that the historically low budget allocation and utilization of funds in India receives urgent attention. The pandemic driven contraction and the projected decline in economic growth is likely to put enormous strain on exchequer. Therefore, the need to accelerate public spending by both central and state governments assumes importance due to nancial vulnerabilities sustained by households. Job and wage loss suffered during the pandemic by households implies declining household income resulting in poor nancial risk protection, worsening health status, and exacerbating health inequities. The government of India at all levels should improve the effectiveness of public spending by removing bottlenecks that prevent access to and utilization of health interventions.