Costs incurred by patients with oral potentially malignant disorders: is there a public health need for financial protection in India?

Objectives Financial protection mechanisms are in place to overcome the costs of a few diseases in India. Our objective was to estimate the costs incurred due to Oral Potentially Malignant Disorders (OPMD) and to determine predictors of such costs. Results We found that the median (Interquartile range IQR) total costs of OPMD was Indian Rupees (INR) 500 (350–750), direct medical costs was INR 0 (0–50), direct non-medical costs was INR 150 (40–200) and indirect costs was INR 350 (250–500). The travel cost to attend the health facilities [INR 100 (40–150)] and the patient’s loss of wages [INR 200 (150–400)] mainly accounted for the direct non-medical and indirect costs respectively. The median expenditure on smokeless and smoking forms of tobacco was INR 6000 (5400–7200) and INR 2400 (1800–3600) respectively. On multiple linear regression analysis, rural residence, belonging to below poverty line family, being a sole earner in the family, number of months since diagnosis and first visit at a private provider were found to be the significant predictors of total costs of OPMD. Financial protection mechanisms are needed for covering the direct non-medical and indirect costs. Early management of OPMD might mitigate the costs of OPMD. Supplementary Information The online version contains supplementary material available at 10.1186/s13104-021-05814-2.

Existing literature on costs of OPMD, none from India, focused on the costs borne by the healthcare facilities in treating these patients [8,9]. The average annual cost borne by a hospital for a patient with oral lichen planus was reported as £400 (~ 36 000 Indian Rupees-INR) and the cost of laser surgery for OPMD was reported as £998 (~ INR 89 000) during 2015-16 [8,9]. Evidence is scarce on costs incurred by patients with OPMD and on the predictors associated with such costs in India. Evidence suggests that financial protection mechanisms are beneficial in reducing morbidity as well as mortality among cancer patients [10,11]. We conducted this study to estimate the costs incurred by patients for the care of OPMD and to determine the predictors of such costs in the Bhavnagar district of Gujarat (western India).

Study design, setting and participants
We conducted a cross-sectional study in the outpatient department (OPD) of the Otorhinolaryngology Department of Sir Takhtsinhji Hospital of Bhavnagar district of Gujarat from July-September 2019. The population of Bhavnagar district and Bhavnagar city as per census 2011 is around 2.8 million and 0.6 million respectively (a sparsely-populated semi-urban and rural setting) [12]. Patients ≥ 18 years and diagnosed with OPMD by an Otorhinolaryngology specialist at our hospital were included in the study.

Sample size
A pilot study was conducted among 40 OPMD patients in the hospital, which reported the standard deviation of total costs of OPMD as 1133. A sample size of 219 was calculated using the formula N = 4Z α 2 S 2 /W 2 ,[13] considering 95% confidence level (Z α = 1.96), the desired width of the confidence interval (W) as 300, and a standard deviation (S) of 1133.

Recruitment and sampling
The principal investigator (PI-first author) enrolled patients in the study through the register maintained by the staff nurse of Otorhinolaryngology OPD. On an average day in the Otorhinolaryngology OPD of our hospital manages approximately 10-15 patients with OPMD. Every second patient was enrolled consecutively in the study. After the written informed consent procedure, the questionnaire was administered by the PI.

Study tool
The questionnaire used for the study was a validated tool of the World Health Organization (WHO) for estimating direct medical, direct non-medical and indirect costs among tuberculosis patients, which was adapted for estimating costs among OPMD patients [14].

Cost survey
The costs incurred by patients were summed up for the visits they made starting from the symptoms of OPMD till they were diagnosed or treated at our hospital (that is, till the time of interview). The costs were categorized as direct medical costs (costs for consultation, hospital admission, medicine, radiology, laboratory, and any diagnostic procedures), direct non-medical costs (costs for travel, food, and accommodation) and indirect costs. Indirect costs included patient's and accompanying member's loss of wages, and monthly family income before symptoms of OPMD minus monthly family income at the time of interview. The total costs included the sum of direct medical, direct non-medical, and indirect costs incurred by the patients for the care of OPMD. The expenditure on consuming tobacco (both smokeless as well as smoking forms) was calculated from the time they first started their consumption until the time of the interview (it was not added to any category of costs of OPMD). During the study period, on an average, one US$ equalled 70 INR.

Variables
The primary outcome variable was the total healthcare costs incurred by OPMD patients as a continuous variable. The predictor variables were age in years, male gender, residence (rural vs. urban), years of education, occupation requiring labour, being a sole earner in the family, belonging to below poverty line family, having OSMF, time since diagnosis, time since consuming smokeless tobacco and having the first visit at a private provider.

Statistical analysis
The healthcare costs (being a continuous variable) was expressed in terms of median and interquartile range (IQR) as the skewness (4.7) and kurtosis (33.3) values suggested a lack of normality. Uni-variable linear regression analysis was performed to find out predictors to be included in the multi-variable analysis. The predictor variables with p-value < 0.2 on uni-variable regression were included in the multivariable analysis. Multiple linear regression analysis was applied using the "Enter" method for finding out the significant predictors of total healthcare costs of OPMD. The difference was said to be significant when the p-value was < 0.05.

Characteristics of patients
A total of 223 eligible patients were enrolled in the study (4 patients did not agree to participate in the study). The median age of the patients was 35 (IQR 28-45) years, 78% were male, 24% had no schooling, 75% belonged to other backward classes (OBC), 47% were living in the rural area and 71% had an occupation requiring labour ( Table 1). The median family income was Indian Rupees (INR) 15 000 (IQR 10 000-20 000), 41% were belonging to below poverty line family and 27% were sole earners in their family. The median expenditure of consuming smokeless and smoking forms of tobacco was INR 6000 (IQR 5400-7200) and INR 2400 (IQR 1800-3600) respectively. Seventy-five percent of patients were newly diagnosed, 28% were on treatment for OPMD and 8% had first visited a private provider.

Cost burden of OPMD
The direct medical costs incurred by patients with OPMD were low and were contributed mainly by costs for the purchase of medicines [median INR 0 (IQR 0-50)] (

Predictors of costs of OPMD
Age in years, male gender, rural residence, being a sole earner in the family, belonging to below poverty line family, number of months since diagnosis, number of years since consuming smokeless tobacco, and having the first visit at a private provider were the variables found to have a p-value < 0.2 on unadjusted uni-variable linear regression analysis [see Additional file 1]. On performing multiple linear regression, rural residence, belonging to below poverty line family, being a sole earner in the family, number of months since diagnosis, and first visit at a private provider were found to be significantly predicting total costs of OPMD (Table 3). Rural residence, increase in one month since diagnosis and having the first visit at a private provider increased the total costs of OPMD by INR 1.2, 1.5, and 1.7 respectively. Belonging to below the poverty line family and being a sole earner in family decreased the total costs of OPMD by INR 1.2 and 1.3 respectively. The results of the regression indicated that the predictors explained 45.6% of the variance [adjusted R 2 = 0.456, F(8,205) = 23, p < 0.001]. Age in years, male gender, and the number of years since consuming smokeless tobacco were not significantly predicting costs of OPMD.

Discussion
Patients with OPMD incurred lower direct medical and total healthcare costs when compared with their monthly family incomes in this study setting. As might be the case with patients managed at a government hospital, patients with OPMD in the present study incurred lower direct medical costs due to free diagnostic and treatment services. Over 90% of patients first visited our facility (government-run) for the management of OPMD. The low percentage of patients visiting private providers indicates their healthcare-seeking behaviour towards the public health system. Almost half of the patients resided in rural   areas where specialist services of the private sector are scarce. Also, since three-fourth of the patients were diagnosed on the first visit when the cost survey took place, the costs for pre-diagnostic visits elsewhere were not incurred. None of the patients required hospitalization and were managed on an OPD basis, cancelling out costs of hospitalization. The loss of wages of patients with OPMD and the travel cost for attending health facilities were the prominent costs reported in this study. Early identification and appropriate case management of OPMD would be pertinent in reducing the costs. The current study found substantially high expenditure on tobacco consumption (especially for SLT) among OPMD patients. Studies in India have documented similar expenditures on tobacco consumption in other populations as well [15][16][17]. Tobacco cessation activities would arrest the initiation and progression of OPMD into oral cancer, and would eventually allay current and future healthcare costs.
The present study found that rural residence, duration since diagnosis of OPMD, and having the first visit at a private practitioner increased the costs of OPMD. Patients staying in rural areas have to travel to the tertiary care centres, which are usually found in towns or cities, for the care of their OPMD complaints. Also, due to the long travel, there is a possibility of delayed careseeking and a longer duration of diagnosis since the onset of symptoms. Also, the poor accessibility to specialists in rural areas makes patients seek care at local traditional healers or quacks or private practitioners, thereby adding to the costs incurred for OPMD. These findings were supported by research on the determinants of catastrophic expenditures in patients with cancer [18][19][20].
We conclude that the costs of OPMD in our setting are low, however, the need for financial protection mechanisms such as universal cash transfers should be explored to overcome the direct non-medical and indirect costs. This is the first study estimating the healthcare costs of patients with OPMD in a sparsely-populated district of western India with a high caseload. Further studies are required in densely-populated urban metropolitan settings of India for estimating costs, and exploring the effects of early management, tobacco cessation, and financial protection on cost mitigation among patients with OPMD.

Limitations
There are a few inherent limitations of studies estimating costs. Since the costs are estimated based on a survey, recall bias is inherent. However, we interviewed more than 90% of patients within a month of diagnosis to avoid recall bias. We could not use any scale for assessing the severity of OPMD among the patients, which might have differentially affected the costs incurred. Also, costs incurred till the resolution of symptoms or cure or progression could not be included as we did not follow up with the patients. Our study did not employ the 'cost of illness' perspective, that is, we did not estimate the financial burden incurred by the hospital for the care of the patients. Since patients taking treatment at a private facility were not included, the study findings can be generalized only to patients being treated at government health facilities in similar semi-urban settings of India.