Cost of screening, out-of-pocket expenditure & quality of life for diabetes & hypertension in India

Given the rising burden, early age of onset and the associated economic burden of non-communicable diseases (NCDs), the Government of India launched the National Programme for Prevention and Control of Cancer, Diabetes, Cardiovascular Diseases and Stroke (NPCDCS) aiming at prevention as well as early detection and treatment of diabetes, hypertension and common cancers. The objectives of the program are to prevent and control common NCDs through behaviour and lifestyle changes and provide early diagnosis and management of common NCDs1. The focus of NPCDCS was to enable opportunistic screening for common NCDs at secondary care level, through setting up of NCD clinics. Further, population based screening (PBS) for NCDs including diabetes, hypertension and the three common cancers (breast, cervical and oral) has been initiated to complement the NPCDCS at primary care level. PBS is envisaged to address issues of low levels of care-seeking, limited access to health services, reaching marginalized groups, in addition to increasing awareness in the community about NCDs and the need for periodic screening. The process of screening includes active population enumeration, risk scoring on the community based assessment checklist (CBAC) by an accredited social health activist (ASHA) and allocation of unique identification number, followed by screening of all individuals above the age of 30 yr at a facility based fixed day camp or through outreach2.

While screening of NCDs has been initiated at a national level, questions arise related to the most efficient or cost-effective screening methodology. To build evidence for these policy decisions, economic evaluations of different strategies need to be undertaken. The Government of India has established a health technology assessment agency called Health Technology Assessment India (HTAIn) to strengthen evidence based policy making3,4. A study was commissioned by the HTAIn to assess the cost-effectiveness of screening for diabetes and hypertension in India. A pre-requisite for such a study involving an economic evaluation for screening strategies for diabetes and hypertension is data on health system costs of screening, out-of-pocket expenditure (OOPE) of patients and health-related quality of life (HRQoL) of diabetes and hypertension patients. While a few studies have been conducted in different parts of the country to assess the implementation of the NPCDCS programme5,6, cost of implementing the PBS has not been assessed. Second, while the National Sample Survey Organization (NSSO) estimates the nationally representative OOPE for outpatient and inpatient care for different illnesses, it is not possible to classify diabetes and hypertension into those with and without complications. Third, the NSSO data do not specifically provide OOPE for those with comorbidity, i.e. diabetes and hypertension, as well as those visiting a tertiary-level hospital. Finally, there is no Indian study on HRQoL of diabetes and hypertension patients using the Euro Quality of Life Questionnaire (EuroQoL), with five dimensions and five-level (EQ-5D-5L) scale tool, which is recommended by the HTAIn for use in economic evaluations.

To address this evidence gap, the current study aimed at determining the health system cost of implementing PBS for diabetes and hypertension, disaggregated OOPE expenditure and HRQoL of patients with diabetes, hypertension, both diabetes and hypertension – both uncomplicated and complicated.

Material & Methods

The study was conducted by the department of Community Medicine & School of Public Health, Postgraduate Institute of Medical Education & Research (PGIMER), Chandigarh, a tertiary care hospital. The study was carried out after approval from the Institute’s Ethics Committee. The investigation comprised three parts namely; costing of population based screening (PBS) of diabetes and hypertension, out-of-pocket expenditure (OOPE) and health-related quality of life (HRQoL) of diabetes and hypertension patients. Assessment of OOPE and HRQoL was done by recruiting outpatients from the departments of Endocrinology and Internal Medicine, PGIMER, Chandigarh. An informed written consent was obtained from all participants.

Results

Health system cost of screening:’The per capita health system cost of PBS for diabetes, hypertension and both diabetes and hypertension at sub centre level in Haryana was ₹ 92 (US $ 1.32), 70 (US $ 1.00) and 130 (US $ 1.86), respectively, and in Tamil Nadu these were ₹ 22 (US $ 0.31), 13 (US $ 0.19) and 25 (US $ 0.36), respectively. The pooled unit cost of PBS for diabetes, hypertension and both diabetes and hypertension screening at 50 per cent screening coverage was ₹ 38.4 (US $ 0.55), 16.2 (US $ 0.23) and 45.2 (US $ 0.65), respectively (Figure).

Close

Figure

Unit cost of screening by level of population-wide coverage.

Export to PPT

In the scenario analysis using a national average wage for healthcare workers in Tamil Nadu, it was observed that the unit cost for screening of diabetes, hypertension and both diseases in Tamil Nadu changed to ₹ 65.7 (US $ 0.94), 56.3 (US $ 0.81) and 80.65 (US $ 1.15) respectively. The adjusted pooled estimates for unit cost of screening of diabetes, hypertension and both diseases were ₹ 53.5 (US $ 0.77), 31.1 (US $ 0.44) and 65.0 (US $ 0.93), respectively.

OOPE: Table II outlines the OOPE per outpatient consultation by socioeconomic and clinical factors. The mean OOPE of a patient with diabetes visiting a tertiary care facility was ₹ 4381 (95% CI: 3786-4976). The mean OOPE for a patient with diabetes with and without complications was ₹ 6007 (95% CI: 3832-8181) and ₹ 3990 (95% CI: 3484-4496), respectively. Similarly, for hypertension, the mean overall OOPE on outpatient consultation was ₹ 1427 (95% CI: 1278-1576), which varied from ₹ 1710 (95% CI: 1198-2223) to ₹ 1392 (95% CI: 1238-1547) among patients with and without complications. It was observed that OOPE was higher among older patients, those belonging to higher richer wealth quintile, those with complications and those with uncontrolled disease status.

The CHE was estimated to have been incurred among 20, 1.3 and 14.8 per cent of the patients with diabetes, hypertension and both diabetes and hypertension, respectively. Further, the percentage of patients experiencing CHE was higher in patients suffering from complications and among those with uncontrolled disease, across the three conditions. An increasing pattern was observed with an increase in the number of complications suffered by patients in both OOPE and CHE (Supplementary Table I).

Wealth quartile, disease condition and presence of complications were found to be associated with higher OOPE (P<0.05; Table III). OOPE was significantly lower in patients with hypertension as compared to diabetes patients and patients with both (beta coefficient –0.286, P<0.001).

The odds of incurring CHE was 1.8 times higher in females and those with complications as compared to males and those without complications, respectively (Table IV). The odds of incurring CHE was highest among patients with diabetes and least in patients with hypertension. The odds of incurring CHE was 36.5 times higher in the poorest quintile as compared with the richest quartile (P<0.001).

HRQoL: The mean HRQoL score for diabetes patients with and without complications was 0.61 (95% CI: 0.53-0.69) and 0.8 (95% CI: 0.76-0.84), respectively. Similarly, the mean HRQoL score for patients of hypertension with and without complications was 0.78 (95% CI: 0.76-0.8) and 0.9 (95% CI: 0.88-0.92), respectively, and for those with both diabetes and hypertension with and without complications was 0.62 (95% CI: 0.58-0.66) and 0.72 (95% CI: 0.68-0.76), respectively. The quality of life of patients was lower in patients with complications, patients with uncontrolled disease and patients who were on insulin-based treatment (Table V). Further, the distribution of levels reported of different domains in the tool was also assessed (Supplementary Table II).

HRQoL scores and mean OOPE of patients with different complications are provided in Supplementary Table I.

Discussion

NCDs present a significant public health challenge with adverse health effects and economic burden. They not only burden the health system but also push households into poverty owing to the requirement of long term management and treatment. To build evidence to inform policies on screening of diabetes and hypertension, information on cost borne by the health system and patients is required. The first year target of the PBS was 50 per cent coverage of target population2. The pooled screening cost for diabetes and hypertension at 50 per cent coverage was ₹ 45.2 (US $ 0.65). The unit cost of screening declined with rise in coverage and plateaued at 70 per cent population coverage. A wide variation was observed in screening cost in the two States (Haryana and Tamil Nadu), attributable to the mode of implementation. First, the primary screening in Tamil Nadu was done at household level compared to camp or facility-based mode in Haryana. Second, human resource cost contributed to 75-90 per cent cost in Haryana as compared to 27-35 per cent in Tamil Nadu (

The estimates of HRQoL for patients with diabetes were found to be consistent with other studies in Canada, Norway, Japan and the Netherlands, which reported utility scores of 0.76 (95% CI: 0.58-0.94), 0.85 (95% CI: 0.82-0.87), 0.88 (95% CI: 0.85-0.91) and 0.81 (95% CI: 0.8-0.82) among uncomplicated patients and 0.7 (95% CI: 0.5-0.9), 0.73 (95% CI: 0.69-0.78), 0.85 (95% CI: 0.82-0.97) and 0.61 (95% CI: 0.59-0.63) among complicated patients20-23. The HRQoL estimates for patients with hypertension in our study were in keeping with a study in Nepal, reporting HRQoL of 0.87 (95% CI: 0.86-0.89) and 0.64 (95% CI: 0.51-0.77) on EQ-5D-3L and VAS, respectively24. The quality of life of patients with complications and those with uncontrolled disease was observed to be lower across all disease groups, as reported by a number of other studies20,22,25, highlighting the need to focus on treatment compliance and prevention of development of complications amongst diabetes and hypertension patients.

The present study also assessed OOPE per outpatient consultation for patients with diabetes and hypertension. Patients with diabetes had the highest OOPE followed by patients with both diabetes and hypertension. While the difference between OOPE amongst diabetes patients and those with both diabetes and hypertension was not statistically significant (P=0.176), possible reasons for this difference were explored. The samples of both disease groups were similar in age distribution, socioeconomic status, education, insurance status and other demographic variables. It was noted that the average duration of disease amongst the diabetes group was 21.3 yr as compared with 12.4 yr among the group with both diseases. It has been reported by multiple studies that a longer duration of illness amongst diabetes patients is associated with difficulty in achieving glycaemic control, requirement of more and higher doses of medication26,27. Such medication is associated with higher costs. This is reflected by the average monthly OOPE on medicines of patients with diabetes (₹ 2936) and those with diabetes and hypertension (₹ 2500). This is consistent with another study from India which reported that expenditure increased with duration of diabetes26. An increasing trend in OOPE was observed with the presence of complications (Table II) and the number of complications (Supplementary Table I); this was in keeping with other studies that reported expenditure proportionately increasing with the number of complications28,29. Furthermore, the input-wise distribution of OOPE was assessed (Supplementary Table III) and majority (63%) of the expenditure was found to be incurred on medicines, which is consistent with a number of other studies in India28,29.

Policy implications: This study highlights a few operational and programmatic considerations with regard to PBS of diabetes and hypertension. The Tamil Nadu model was observed to have more effective implementation in limited resources. The recently announced health and wellness centres (HWCs) under the Ayushman Bharat programme30 could help in smooth implementation of PBS with better coverage and resources. The adverse effects of comorbidity of diabetes and hypertension in terms of poor HRQoL and economic burden was clearly established in the present study, highlighting the importance of screening for both conditions together and continued treatment to prevent progression to complications.

Our investigation, however, had certain limitations. First, the OOPE and HRQoL estimates were generated from a cross-sectional sample drawn from one tertiary level public healthcare facility in north India; thus, the study results could have limited generalizability. However, this facility has patient footfall from more than six Indian States and as a result represents the heterogeneity in terms of geography, rural and urban distribution, severity of disease and socioeconomic status. As a result, there is little possibility of any selection bias resulting from a hospital based sample. Further, the findings of this study do not comment upon prevalence, incidence and long term management of NCDs since it was a one time survey. Long term consequences of the financial hardship should be assessed in future studies using a cohort study design to understand the implications pertaining to individuals and households. Second, the EQ-5D-5L health states were covered into utility scores using the Thailand tariff value set due to the absence of an Indian value set. Third, health system cost data on resources such as training, IEC and supervision were collected using a top-down approach from the state level and apportioned to the facility at which the costing was being performed. Finally, the overall cost of management for diabetes and hypertension would require an assessment of health system costs, which is not the objective of the present study. A more comprehensive assessment of health system costs and OOPE would help in determining the overall economic burden of diabetes and hypertension.

To conclude, the current study findings can be used to further undertake cost-effectiveness analysis to determine the ideal interval of screening, mode of screening and diagnostic test. The cost estimates after incorporation of estimates of health system cost as well may be used for determining the reimbursement package rates under various publicly financed health insurance schemes in India. The present study also highlights the rising economic burden of NCDs, largely borne by the patients, which calls for steps in the direction of health system strengthening such as establishment of HWCs.