Measuring unmet healthcare needs - findings from a tribal district in the Chhattisgarh State of India

doi:10.21203/rs.3.rs-1396831/v1

Download PDF

Research Article

Measuring unmet healthcare needs - findings from a tribal district in the Chhattisgarh State of India

https://doi.org/10.21203/rs.3.rs-1396831/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

There are many indicators and methods in use for measuring access to healthcare, much of which relate to physical access or utilization of services. This study measures unmet healthcare needs in a district with relatively good physical access and insurance coverage and discusses the relevance of our findings for the design and monitoring of progress towards achieving universal health coverage (UHC).

Methods

A multi-stage sampling survey identified and interviewed a sample of 3153 individuals from 598 households drawn from 13 villages and seven urban wards. In-depth qualitative interviews of five purposively chosen individuals from each village and urban ward, was also undertaken. The study examined access in terms of perceived healthcare needs and unperceived or latent healthcare needs and the proportion of such needs that utilized effective health care and the proportion that did not. Estimation of unperceived healthcare needs was limited to five tracer conditions- hypertension, diabetes, depression, rheumatic conditions and dependence in activities of daily living in the elderly. Bivariate, multivariate (logistic regression), and thematic analyses were used for this study.

Results

Out of the total population, 10.47% reported perceived healthcare needs for acute ailments in the last 15 days, and 10.62% of the population self-reported suffering from chronic conditions. 12.75% of those with acute ailment and 18.40% of those with chronic ailment received no treatment, while 27.83% and 9.07% respectively took care from informal care provider. Unperceived healthcare needs were high in all five-tracer conditions studied. Incomplete treatment was another form of unmet need. Inappropriate care conceals unmet healthcare needs. Unperceived healthcare needs were significantly higher for poor and marginalized population subgroup, whereas perceived healthcare needs were higher among upper socioeconomic and less marginalized population.

Conclusion

The measurement of unmet healthcare needs should include both perceived and latent needs, as well as incomplete care and care from informal providers. Further, inappropriate care conceals a lot of unmet healthcare needs, and its measurement is not easy. In a context of information asymmetry and prevailing health markets, the measurement of unmet healthcare needs in all its complexity, becomes essential for measuring progress towards UHC.

Trial registration: Not Applicable

Unmet healthcare needs

Universal Health Coverage (UHC)

access to healthcare

Universal Health Coverage (UHC) has been defined as the desired outcome of health system performance whereby all people who need health services (promotion, prevention, treatment, rehabilitation, and palliation) receive effective care without undue financial hardship [1].Access and financial protection are two core components of UHC [2].

Conceptualization of access has been done in different ways, but the precise meaning is unclear [3, 4]. In various research and healthcare policies, it is agreed that access should be measured in the context of the community's healthcare needs and its utilization [5]. In the context of measuring progress towards UHC, the proportion of those who need healthcare services and who receive effective healthcare without undue financial hardship becomes the measure of access [1].

Access has been studied from many angles. One type of studies conceptualize access as a purely supply-side concept. It deals with the availability of services. It relates, access to a measure of location[6]. The concept of provider population ratio is part of this approach. Another approach is to study the demand side: ‘who uses services?’ or utilization of services. It could also study differentials in utilization, especially when equity to access is the focus [7]. Another important determinant of access is affordability. Ability to pay for services, supplementing income for those who cannot pay, insurance coverage, eligibility criteria for the publicly funded program or other matters related to cost of care at the time of delivery comes under this [8]. In the utilization of services, there is also the dimension of the fit between people's healthcare needs and supply of healthcare to them. It conceptualizes access as not merely the availability of services but the quality and information dimension, which could be termed the acceptability of services [3, 9].

The conceptualization of access has implications for measuring progress towards universal health coverage [10]. One implication is that neither physical access nor insurance coverage nor even utilization rates for different services can be equated with “effective access” viz the proportion in need of healthcare who could access appropriate services of sufficient quality [11, 12]. Further there are many challenges with measuring healthcare needs [13]. This study categorizes all healthcare needs into perceived and unperceived healthcare needs. Perceived needs can be met (fulfilled) or unmet. Further unmet needs could take the form of a) no treatment b) treatment that is incomplete, c) treatment from an unqualified provider and d) treatment that is inappropriate viz. it does not qualify to be called or categorized as “health care that is effective”. In “incomplete care” as measured in this study, the individual was unable to continue medication for the prescribed period. Unqualified or informal providers are not a part of the government strategy of provision of essential healthcare services, and therefore irrespective of whether their prescription worked, they are for the purpose of this study categorised as part of the gap in services. In "inappropriate care”, the individual got the treatment, but it was not in accordance with what is normatively considered as required.

The aim of the study was to measure unmet healthcare needs of all the above types in a tribal district of Chhattisgarh and understand the implications of the large levels and types of unmet needs for designing road-maps and monitoring strategies towards the achievement of UHC.

The cross-sectional study was done in the Korba district of Chhattisgarh, India (population 12.06 lakhs) from June 2018 to October 2018. The rationale behind the selection of Korba for the study is that it is one of those districts where on one hand a significant proportion of the population belong to marginalized communities (scheduled tribe -40.9%, and scheduled caste 10.33%) and on the other hand the roll out of strategies that constitute the Indian road map towards UHC is in a relatively advanced stage. Every resident of the state is officially covered by one of the two government-funded health insurance programs¹ whose objectives are to provide access to almost the entire range of secondary and tertiary care services. There is also a public district health system with the required density of public health facilities and with a district hospital that is quality certified [16, 17].

A total of 3153 individuals (598 households) were selected through multistage random sampling. Selection of villages for rural areas and urban wards for the urban area was done through proportionate to population size sampling (PPS). In the first stage, 13 rural villages and seven urban wards were selected. All households in the village were listed, and from these, the required number of 30 households were selected by systematic random sampling (Additional file 1, Figure S1). The sampling method is the same as in India’s large demographic surveys [18]. Out of the total sample, 374 households (2012 individuals) were from rural areas, whereas 224 households (1141 individuals) were from urban areas viz- the sample was 64% rural and 36% urban households. -which is consistent with the 63% rural and 37% urban ratio for the whole district (Census 2011).

Data was collected via interview schedule from all the 598 households and 3153 individuals. The interview schedule included demographic details of the households, hospitalization episodes in the last 365 days, acute illness and outpatient visits in the last 15 days, chronic illness in the last one-year, choice of provider, medical and non-medical expenditure, health insurance coverage and benefits, and perceived unmet healthcare needs.

As a single question, individuals were asked if they were suffering from any acute illness in the last 15 days or any chronic condition (diseases duration more than three months) in the last 365 days. And then they were asked what the disease was? Whatever was stated- symptoms or disease was noted. If there was any documentation, this was used to verify further, but it was as perceived and self-reported for the most part. The response was later categorized into 78 different ailment categories by the principal investigator in consultation with two medical experts. To understand the chronic disease burden and aid recall, respondents were further asked if anyone in the family was suffering from any of the following ten diseases conditions: 1) Hypertension, 2) Diabetes, 3) Kidney or urinary disease, 4) Lung diseases (asthma, chronic obstructive pulmonary diseases and chronic bronchitis), 5) Heart diseases, 6) Stroke, 7) Arthritis or rheumatic conditions, 8) Neurological conditions 9) mental illness 10) Cancer. This is consistent with the methodology followed in Longitudinal Aging Study in India (LASI) and the WHO study of Global Aging and Adult Health (SAGE), two studies that also had a focus on chronic disease[19]. The aim was also to calculate recall for these ten major chronic conditions compared to the one-line question response. This latter method is used in the NSSO morbidity surveys [20].

Individuals were also asked if they were enrolled under the health insurance scheme (RSBY and MSBY).

For analysis, ailments were classified into six broader categories: 1) Acute illnesses, 2) Chronic communicable diseases, 3) chronic non-communicable diseases-, 4) Maternal health and family planning, 5) Injuries and 6) not classified in any of the above categories (Additional File 1, Table S1).

Choice of service providers during hospitalization and outpatient care was also recorded along with broader public and private categories as indicated in Additional file 1, Table S2. The background characteristics of each household were recorded and this included age, sex, residence, caste, education levels, occupation, and stated income (Table 1). In addition, the household's usual annual per capita consumption expenditure (UAPCE) was estimated, and this was used to categorize them into economic quintiles.

All household members who were 30 years or above and gave consent for participation were screened for hypertension and diabetes. In case of non-availability of the household member, they were followed up three times- once on the same day evening and twice on the following day. Blood pressure equal to or more than 140 (systolic)/90 (diastolic) persistent on repeating measurement thrice with time intervals in case of high blood pressure recording, and random blood glucose of 200mg/dl or above, were cut-off points for suspecting hypertension and diabetes [21]. Standard protocols as used in earlier National Family Health Surveys (NFHS) were followed for blood pressure and blood glucose measurements [18]. If the measurement indicated presence of the disease, the participant was referred to the nearest government health facility for follow-up. The local community health worker (also called ‘Mitanin’ in Chhattisgarh) was informed to facilitate follow-up care.

The standard “Patient Health Questionnaire-9 (PHQ-9)” was used to screen for depression. It was administered to one male and one female member of the household above 18 years of age. Selection of the respondent for the questionnaire was done randomly from among the available eligible members of the household and who gave consent, at the time of the survey. ² This PHQ-9 tool has been validated for making a diagnosis of depression according to DSM-IV criteria and used in many Indian studies [22–25].

Rheumatic patients were asked about their care seeking pathways, and elderly (60 years or above age) were asked about their dependence in basic activities of daily living (BADL), which included dressing, walking, bathing, eating, getting out of bed, and using the toilet. Elderly were also asked on instrumental activities of daily living (IADL), including preparing the meal, shopping for groceries, making telephone calls, taking medication, working around the house, managing money, and finding addresses at unfamiliar places.

Perceived healthcare needs (same as self-reported morbidity), and the proportion of these where health care needs were met or unmet were correlated with the background characteristics named earlier using multivariate logistic regression models.

The qualitative component of the study did in-depth interviews, with five purposively chosen survey participants who had met multiple providers, to understand the care pathway they have experienced, the reasons for following such a pathway, and the reasons why initial choice of providers were inadequate. Five persons who had been most ill or had required more providers and care were selected for the interview from each of our sample villages and urban wards. In this way, 102 such interviews were recorded, 68 were from rural and 34 from urban areas. The interview started with an open-ended question of what the patient did once he or she became ill and then followed up on their experiences in receiving healthcare and paying for it subsequently. Across the diseases category, 68 out of 102 interviews were related to non-communicable diseases (NCDs), 11 from chronic communicable, nine from injury, eight from acute illnesses, and six from maternal health and family planning. Thematic analysis was used for analysing the qualitative data.

The study was approved by the Institutional Ethical Committee of State Health Resource Centre (SHRC), Chhattisgarh.

^[1] Rashtriya Swasthya Bima Yojana (RSBY) and Mukhyamantri Swasthya Bima Yojana (MSBY)- the RSBY is paid for by the centre covers all below poverty line (BPL) households. The MSBY, which is paid for by the state, includes all above poverty line (APL) households. So every family has the entitlement, though, in practice across the state, 80% of families and 60% of individuals have had their health-card made [14]. According to state data, 1,97,895 families under RSBY and 66,588 under MSBY and a total 2,64,483 overall Korba were enrolled at the time of the survey in June 2018. Almost all private hospitals are empanelled as providers under RSBY and MSBY[15]. The up-gradation of sub-centres to health and wellness centres (HWCs) is also progressing well.

^[2] Randomization was done by the chit-pull system, where all the names of above the age of 18 years were written on the different pieces of paper for male members and female members separately. Out of them, one male and one female member names were picked up. In the case of those houses where only male or female members whose age was above the age of 18 years, randomization was done for that sex group only.

The findings of the study are organized into following four sections: Section A presents perceived healthcare needs of the individual, Section B presents unperceived healthcare needs, Section C presents perceived unmet healthcare needs, and Section D presents Inappropriate care.

A. Perceived healthcare needs

1. 10.47% of respondents reported having experienced an acute ailment, which they perceived required healthcare in the last 15 days (Table 1).

2. In addition to the above, 8.18% of respondents reported having a chronic (condition persisting more than three months). This was when there was no cue or prompt. Once given a prompt with ten common chronic conditions, this went up to 10.62% (Table 1).

3.The reported hospitalization was 5.07% in the district, and it was higher in urban areas (6.13%) than in rural areas (4.47%-Table 1).

4. The rate of perceived healthcare need/ self-reported morbidity for acute illness, for chronic illness, and for hospitalization, all changed with background characteristics of the place of residence (urban-rural), sex, age-group, caste, education level, occupation and income (Table 1 &2).

Table 1: Proportion of population who had perceived needs for ailment in the last 15 days, chronic condition and hospitalization in the last 365 days

	Total no in sample (n)	Self-reporting of ailment in last 15 days (%)	Self-reporting of Chronic Condition (%)	Self-reporting of Chronic Condition with prompts* (%)	Hospitalization rate (%)
Total	3153	10.47	8.18	10.62	5.07
Rural-Urban divide
Rural	2012	10.04	6.86	8.55	4.47
Urban	1141	11.22	10.52	14.29	6.13
Sex
Male	1549	8.46	8.13	10.01	3.81
Female	1604	12.41	8.23	11.22	6.30
Age group
0-4	309	19.74	0.94	0.97	3.24
5-14	581	11.57	2.25	2.42	1.21
15-29	920	7.73	3.20	3.31	6.63
30-44	642	9.56	7.89	9.71	5.01
45-59	458	9.83	20.31	29.26	6.11
60+	243	9.88	26.75	37.04	9.05
Social groups
Scheduled Tribe (ST)	667	9.71	5.85	7.43	5.65
Schedule Caste (SC)	1052	10.44	9.19	12.53	4.59
Other backward caste(OBC)	933	11.33	8.42	10.60	4.72
Others/ General (GEN)	501	9.03	13.54	18.75	5.56
Education
Illiterate	1009	12.89	10.04	13.79	4.35
Up to primary	479	11.88	8.75	11.22	5.04
Up to secondary	1377	9.22	7.07	8.79	5.25
Above secondary	288	6.59	6.59	8.58	5.79
Occupation
Own Agriculture	646	8.05	6.66	8.82	2.79
Daily wages worker	1621	11.29	6.42	8.57	5.74
Self employed	407	12.53	11.30	12.29	5.16
Government Employee	110	4.55	19.09	25.45	6.36
Private employee	319	10.66	10.97	16.30	5.96
Coal Industry Employee	50	10.00	18.18	18.18	4.00
Wealth Index- Rural
Poorest	406	9.36	6.90	8.87	4.93
Poor	408	10.29	4.41	4.90	5.39
Middle	395	10.38	7.34	7.59	4.56
Rich	410	10.73	5.61	8.05	3.17
Richest	393	9.41	10.18	13.49	4.33
Wealth Index-Urban
Poorest	235	15.74	7.66	10.64	5.96
Poor	220	13.18	8.18	10.00	7.27
Middle	230	5.22	10.00	13.48	6.52
Rich	226	16.81	15.93	20.35	6.64
Richest	230	5.22	10.87	16.96	4.35

(*): Ten prompts for NCD: Hypertension, diabetes, kidney or urinary disease, diseases of lung, coronary heart disease, stroke, arthritis, neurological problem, psychiatric condition, cancer, others. (#): Proportion of ailing population is the proportion of those people who had chronic ailment or acute ailment at the time of survey.

Table 2: Factors affecting perceived healthcare needs for ailments in the last 15 days, hospitalization, and chronic conditions (n=3153)

	Ailment in last 15 days		Hospitalization		Chronic condition
	Odds ratio	p-value	Odds ratio	p-value	Odds ratio	p-value
Rural-Urban divide
Rural	1.00		1.00		1.00
Urban	1.23 (0.90-1.66)	0.18	1.61 (1.06-2.45)	0.02	1.32 (0.95-1.83)	0.08
Sex
Male	1.00		1.00		1.00
Female	1.56 (1.23-1.99)	0.00	1.88 (1.33-2.67)	0.00	1.18 (0.90-1.55)	0.20
Age group
0-4	1.00		1.00		1.00
5-14	0.45 (0.29-0.70)	0.00	0.27 (0.09-0.76)	0.01	2.29 (0.63-8.24)	0.20
15-29	0.33 (0.20-0.53)	0.00	1.73 (0.77-3.90)	0.18	4.56 (1.32-15.74)	0.01
30-44	0.41 (0.26-0.65)	0.00	1.33 (0.59-2.99)	0.48	11.71 (3.52-38.96)	0.00
45-59	0.42 (0.26-0.67)	0.00	1.87 (0.85-4.12)	0.11	43.10 (13.27-140.01)	0.00
60+	0.45 (0.27-0.77)	0.00	3.26 (1.46-7.27)	0.00	61.37 (18.77-200.56)	0.00
Social groups
Scheduled Tribe (ST)	1.00		1.00		1.00
Schedule Caste (SC)	1.05 (0.72-1.52 )	0.80	0.71 (0.41-1.20)	0.20	1.60 (1.06-2.42)	0.02
Other backward caste (OBC)	1.08 (0.81-1.44)	0.59	0.70 (0.47-1.05)	0.09	1.44 (1.02-2.01)	0.03
Others/ General (GEN)	0.80 (0.48-1.36)	0.43	0.65 (0.33-1.26)	0.21	1.92 (1.17-3.15)	0.00
Education
Illiterate	1.00		1.00		1.00
Up to primary	1.19 (0.85-1.68)	0.30	1.75 (1.04-2.94)	0.03	1.01 (0.71-1.44)	0.93
Up to secondary	1.10 (0.72-1.67)	0.65	1.52 (0.82-2.81)	0.18	0.83 (0.54-1.28)	0.42
Above secondary	0.89 (0.52-1.51)	0.66	1.54 (0.76-3.13)	0.23	0.42 (0.25-0.73)	0.00
Occupation
Own Agriculture	1.00		1.00		1.00
Daily wages labourer	1.42 (1.01-2.00)	0.04	2.11 (1.23-3.61)	0.01	0.99 (0.68-1.43)	0.95
Self employed	1.80 (1.15-2.81)	0.01	1.99 (1.00-3.99)	0.05	0.96 (0.60-1.55)	0.89
Government Employee	0.68 (0.25-1.79)	0.44	2.84 (1.14-7.08)	0.03	2.45 (1.31-4.57)	0.00
Pvt. Salaried employee	1.42 (0.85-2.37)	0.17	2.32 (1.11-4.83)	0.02	1.36 (0.82-2.26)	0.23
Coal Industry Employee	1.24 (0.45-3.41)	0.67	1.27 (0.27-5.91)	0.76	1.36 (0.54-3.34)	0.50
Wealth Index
Poorest	1.00		1.00		1.00
Poor	1.23 (0.84-1.78)	0.28	1.22 (0.71-2.09)	0.45	0.93 (0.58-1.49)	0.78
Middle	1.16 (0.79-1.72)	0.44	1.04 (0.59-1.82)	0.88	1.18 (0.75-1.85)	0.47
Rich	1.10 (0.72-1.70)	0.63	1.00 (0.54-1.84)	0.98	1.58 (0.99-2.52)	0.05
Richest	0.90 (0.54-1.48)	0.67	0.64 (0.31-1.31)	0.23	2.08 (1.25-3.46)	0.00
Insurance coverage
No insurance	1.00		1.00		1.00
RSBY/MSBY	1.04 (0.79-1.37)	0.763	0.77 (0.53-1.14)	1.19	1.14 (0.81-1.59)	0.42
Constant	0.11(0.06-0.18)	0.00	0.01 (0.004-0.02)	0.00	0.003 (0.001-0.013)	0.00

We expect that the more marginalized the social circumstance, the greater the morbidity, but as per this table which reports on perceived morbidity, we see that the greater the marginalization, the lesser the perceived healthcare need or reported morbidity. For instance, self-reporting of chronic ailment was 6.90% in the poorest rural quintile whereas it was 10.87% in the richest urban quintile (Table 1 & 2). This was the general pattern. Gender was the only exception, with women reporting more morbidity viz higher perceived healthcare need for both acute and chronic illness in women.

5. Out of the total population who reported perceived healthcare needs in the last 15 days, 80.59% were related with acute ailments, 2.61% for chronic communicable diseases, 10.44% for NCDs, 2.32% for maternal health and family planning, 2.03% for injury, and 2.03% could not be classified (Table 3).

Table 3: Disease pattern for perceived health care needs which was reported in the last 15 days, chronic condition at present and hospitalization in last one year

Sl. No	Ailment name	In last 15 days (n=345)	Chronic Condition (n=375)	Hosp. In last 365 days (n=177)
1.	*Acute illnesses*: All types of fever, diarrhoeas/ dysentery, the discomfort or pain in the eye with redness, acute upper respiratory infection, diseases of mouth/teeth/gums, pain in the abdomen, dog bite, cough with sputum with or without fever, and snakebite	80.59	0	17.48
2.	*Chronic communicable diseases:* Tuberculosis, HIV/AIDS, leprosy, chronic filariasis, and skin ailments.	2.61	7.21	1.13
3.	*Non-communicable diseases:* Diabetes, mental disorders, seizure, weakness in limbs/difficulty in movement, hypertension, heart diseases, bronchial asthma, joint or bone diseases, back or body ache, headache, cancer, thyroid, stroke, cataract, decreased hearing, lump or fluid in abdomen or scrotum, gastrointestinal bleeding, difficulty in urination, sickle cell, cancer, mental retardation, decreased vision, cerebral palsy, disability, liver problem, piles, hypertension and diabetes together.	10.44	90.67	30.47
4.	*Maternal health and family planning:* Normal delivery, Cesarean section delivery, complications in pregnancy, antenatal care, illness in new-born – also pain in pelvic region, change or irregularity in the menstrual cycle, abortion, female sterilization, anemia, under-nutrition.	2.32	1.60	38.96
5.	*Injury:* Road traffic Accidents, falls or any other injuries	2.03	0.53	11.86
6.	Could not be classified	2.03	0	0
	Total	100.00	100.00	100.00

6. Out of the total population who reported perceived healthcare needs due to chronic conditions in the last 365 days, 90.67% for non-communicable diseases, 7.21% for chronic communicable diseases, 1.60% for maternal health and family planning, and 0.53% for injury (Table 3).

7. Out of the total population who reported perceived healthcare for hospitalization in the last 365 days, 17.48% were related to acute ailments, 1.13% for chronic communicable diseases, 30.47% for NCDs, 38.96% for maternal health and family planning, and 11.86% for injury (Table 3).

A.1 Perceived healthcare need- that was Unmet:

Acute ailments in the last 15 days. Out of the total population who reported an acute requiring healthcare, 12.75% did not take treatment (rural: 8.41%, urban: 19.85%-Table 4). Apart from not taking treatment at all, another 27.83% took treatment from the informal provider in the district. This was significantly higher in rural areas (32.71%) than urban areas (19.85%). In other words, 40.58% of persons with acute ailments in the last 15 days did not receive effective health care as defined in this study (Table 4). Of the 59.42 per-cent who went to a formal provider, 28.41% availed services from public provider (CHW: 9.57%, HSC/ANM: 3.77%, PHC: 10.72%, CHC: 2.61%, and DH: 1.74%), whereas 31.01% went to private provider (private doctor clinic: 21.74%, private hospital: 9.28%). Services provisioning by the private hospital was mainly limited in the urban areas (19.85%- Table 4).

Hospitalization in the last 365 days: Out of total hospitalization cases, 44.32% utilized public providers, whereas 55.68% went to the private provider (Table 4). District hospitals (20.34%) and PHC (14.69%) were the major service providers under the public sector.

Chronic conditions in the last 365 days: An estimated 27.47 % of those with perceived chronic illness had unmet needs-. 18.40% took no treatment, 9.07% went to an informal provider (see Table 4).

Table 4: Utilization by type of provider (in percent) for perceived needs in ailments of the last 15 days (n=345), chronic conditions (n=375) and hospitalization in last 365 days (n=177)

To seek care for a chronic condition patient had to travel on an average 44 KM for the public provider, 40 KM for the private provider, and 5.48 KM for the informal provider (Table 5). In terms of median distance, it was 4.0 KM for the public, 7.0 KM for private, and 1.0 KM for the informal provider. The average distance for public providers was the same (4.0KM) in rural and urban areas, whereas for the private provider, the patient had to travel 22.5 KM in rural areas compared to 3.0 KM in urban areas.

Table 5: Access to medicine for perceived health needs in context of patient with chronic conditions and who are “on treatment” (n=319)

	Average duration of medicine discontinued in last 12 months against medical advice (in months)	Median distance of provider (in Km) (n=305)			Proportion of them on treatment received medicine for free (n=305)	Average distance of private medical shop from home (in Km)
		Public(n=76)	Pvt. (n=208)	Inf.(n=35)
Total	6.06	4.0 (mean=44)	7.0 (mean=40)	1.0 (mean5.48)	23.49	11.79
Rural-Urban divide
Rural	6.93	4.0	22.5	2.0	22.50	15.35
Urban	5.07	4.0	3.0	1.0	24.52	8.11
Sex
Male	6.10	5.0	16.0	2.0	22.22	15.30
Female	6.03	2.0	5.0	1.0	24.56	8.83
Age group
0-4	10.00	-	35.00	-	0.00	35.00
5-14	5.76	50.0	27.5	1.0	16.67	21.91
15-29	7.69	6.0	15.0	1.0	28.57	16.44
30-44	7.04	5.0	5.0	0.0	20.00	13.14
45-59	5.42	5.0	5.0	1.5	21.01	12.25
60+	5.50	1.0	7.0	1.0	28.42	7.47
Social groups
Scheduled Tribe (ST)	7.54	5.0	7.0	1.0	37.88	11.78
Schedule Caste (SC)	5.85	3.5	20.0	1.0	21.88	17.42
OBC	5.71	2.0	5.0	1.0	20.42	8.70
Others/ General (GEN)	5.00	11.5	5.0	20.5	13.95	13.60
Education
Illiterate	6.52	2.0	10.0	1.0	27.50	8.45
Up to primary	5.58	2.0	5.0	2.0	25.44	8.92
Up to secondary	6.66	5.0	15.0	1.0	21.52	13.36
Above secondary	5.15	6.0	15.0	-	14.29	22.98
Occupation
Own Agriculture	6.83	3.5	20	1.0	20.37	14.35
Daily wages worker	6.95	5.0	10.0	1.0	18.11	10.93
Self employed	4.83	5.0	4.5	5.0	18.37	15.52
Govt. employee	5.28	28.0	35.0	1.0	25.93	13.90
Private employee	4.69	2.0	1.0	1.0	37.50	8.15
SECL Employee	4.33	3.5	2.5	-	60.00	2.4
Wealth Index- Rural
Poorest	8.19	1.0	20.0	5.0	53.57	12.03
Poor	8.52	8.5	22.5	7.0	27.78	13.69
Middle	8.16	2.0	5.0	1.5	22.22	9.05
Rich	7.02	7.5	20.0	1.0	9.68	14.70
Richest	4.55	6.5	35.0	3.0	12.50	20.94
Wealth Index-Urban
Poorest	6.14	2.0	2.0	0.5	23.53	2.70
Poor	6.06	20.0	5.0	1.0	14.81	10.40
Middle	5.02	1.0	4.0	0.0	22.58	14.11
Rich	4.50	7.5	2.0	1.0	34.78	5.25
Richest	4.50	2.0	3.0	1.0	20.59	7.41

The higher the social or economic status indicators, the longer the distance they were willing to travel (Table 5).

Continuity of care- medication compliance for chronic illness. On an average, patients suffering from chronic conditions discontinued their medicine against healthcare professional advice for 6.06 months (or continued for 5.94 months) out of 12 months. This was highest (10 months) in 0-4 months’ age group, where the chronic condition was a disability (cerebral palsy, autism, mental retardation, developmental delay in achieving milestone) and parents lost hope with treatment. Discontinuation of medicine was relatively higher in populations belonging to lower socioeconomic status. Out of the total patients on treatment, 23.49 % got free medicines, which was marginally higher in urban areas (24.52%) than rural areas (22.50%). Availability of free drug was higher in the poorest economic quintile (53.57%) in rural areas (Table 5).

B. Unperceived unmet healthcare needs (latent) in 5 tracer conditions- hypertension, diabetes, rheumatic disease, depression and disability:

1. Hypertension. Out of the total population of the age of 30 years or above, 47.42% never had blood pressure measured before in their lifetime. Measurement of blood pressure once a year or at least once in three years is a preventive healthcare need. When screened for high blood pressure as per protocol 39.89% had high blood pressure but only 11.39% had been diagnosed as hypertensive earlier, and 9.27% were on the medication (Table 6). In other words, of all latent and known hypertensive patients, 30.62 % were not on treatment, and of known hypertensive, less than one in five were not on medication.

2. Diabetes. Similarly, in the 30 years or above age group in our sample 72.17% never had blood glucose measured before in their lifetime, and it was higher in rural areas (73.86%) compared to urban areas (69.34%-Table 7). Disaggregating by background characteristics it is usually higher where the socioeconomic disadvantage was more. For example, the chances of being screened before for blood glucose was 3.66 times higher (95% CI: 1.71-7.81) in the richest quintile compared to the poorest quintile (Table 8).

Table 6: Healthcare needs and access to care in the context of hypertension in 30 or more-year age group (n=755)

	Blood pressure never measured before	Screening showed high blood pressure*	Had been earlier diagnosed as hypertensive	Medicine being taken for hypertension
Total	47.42	39.89	11.39	9.27
Rural-Urban divide
Rural	59.10	37.72	6.42	5.57
Urban	28.47	43.40	19.44	15.28
Sex
Male	50.00	41.67	10.71	8.93
Female	45.35	38.46	11.93	9.55
Age group
30-44	51.53	27.47	3.99	2.45
45-59	46.07	44.74	14.61	13.11
60+	41.36	56.79	20.99	16.67
Social groups
Scheduled Tribe (ST)	59.92	38.17	6.20	4.96
Schedule Caste (SC)	42.57	38.00	13.86	10.89
Other backward caste (OBC)	44.19	39.94	12.79	10.17
Others/ General (GEN)	26.47	48.53	19.12	17.65
Education
Illiterate	59.31	42.79	10.82	8.66
Up to primary	50.40	40.73	11.69	10.08
Up to secondary	40.00	33.51	9.19	7.57
Above secondary	24.18	43.33	16.48	12.09
Occupation
Own Agriculture	69.72	42.96	8.45	7.04
Daily wages labour	49.88	36.79	8.64	6.67
Self employed	39.77	40.00	9.09	6.82
Government Employee	16.00	52.00	36.00	24.00
Retired govt. employee	17.65	64.71	17.65	17.65
Private employee	15.87	39.68	22.22	22.22
Coal Industry Employee	33.33	46.67	33.33	26.67
Wealth Index- Rural
Poorest	73.04	42.61	5.22	4.35
Poor	78.35	31.25	3.09	2.06
Middle	56.12	32.65	5.10	4.08
Rich	56.52	42.03	4.35	4.35
Richest	25.00	40.70	14.77	13.64
Wealth Index-Urban
Poorest	46.55	44.83	15.52	15.52
Poor	27.87	27.87	13.11	6.56
Middle	34.55	49.09	14.55	12.73
Rich	25.45	43.64	27.27	18.18
Richest	8.47	52.54	27.12	23.73

(*): Includes normotensive people who are on medication

Table 7: Health care needs in the context of diabetes in 30 or more-year age group (n=733)

	Blood glucose never measured before	Screening showed high blood sugar*	Diagnosed as diabetes by a healthcare provider before	Medicine being taken for diabetes
Total	72.17	9.22	6.55	5.05
Rural-Urban divide
Rural	73.86	6.39	4.58	3.70
Urban	69.34	13.92	9.85	7.30
Sex
Male	73.39	6.79	5.50	3.98
Female	71.18	11.17	7.39	5.91
Age group
30-44	77.46	5.10	1.90	1.59
45-59	69.38	12.20	9.30	7.36
60+	66.25	12.58	11.25	8.13
Social groups
Scheduled Tribe (ST)	80.83	5.91	3.33	2.92
Schedule Caste (SC)	71.58	10.87	8.42	6.32
OBC	65.97	10.75	7.76	5.97
Others/ General (GEN)	73.02	11.11	9.52	6.35
Education
Illiterate	75.88	10.22	7.46	6.14
Up to primary	76.76	9.62	7.05	5.39
Up to secondary	68.72	8.99	5.59	3.91
Above secondary	56.47	5.88	4.71	3.53
Occupation
Own Agriculture	75.00	3.60	3.57	2.14
Daily wages worker	79.54	8.76	5.63	4.86
Self employed	66.67	12.20	8.33	5.95
Government Employee	54.17	8.33	4.17	4.17
Retired govt. employee	23.53	29.41	29.41	17.65
Private employee	51.61	12.90	9.68	6.45
Coal Industry Employee	53.33	20.00	13.33	13.33
Wealth Index- Rural
Poorest	84.21	3.57	0.88	0.88
Poor	93.75	4.21	2.08	0.00
Middle	68.75	4.26	3.13	2.08
Rich	60.29	8.82	7.35	7.35
Richest	54.12	12.94	11.76	9.41
Wealth Index-Urban
Poorest	89.09	9.09	9.09	5.45
Poor	70.49	13.11	3.28	3.28
Middle	65.38	21.15	15.38	13.46
Rich	52.73	12.73	9.09	5.45
Richest	68.63	14.00	13.73	9.80

(*) Includes normal blood sugar in known diabetic on treatment

Table 8: Chances of having screened before for blood pressure (hypertension) and blood glucose (diabetes) in the community for age 30 years or above

	Hypertension (n=755)		Diabetes (n=733)
	Odds ration	p-value	Odds ration	p-value
Rural-Urban divide
Rural	1.00		100
Urban	1.51 (0.98-2.32)	0.06	0.51 (0.32-0.80)	0.00
Sex
Male	1.00		1.00
Female	1.35 (0.94-1.94)	0.09	1.18 (0.80-174)	0.40
Age group
0-4	NA
5-14	NA
15-29	NA
30-44	1.00		1.00
45-59	1.59 (1.06-2.38)	0.02	1.67 (1.07-2.60)	0.02
60+	2.6 (1.59 – 4.28)	0.0	2.22 (1.32-3.74)	0.03
Social groups
Scheduled Tribe (ST)	1.00		1.00
Schedule Caste (SC)	0.96 (0.55-1.68)	0.91	1.25 (0.68-2.31)	0.45
Other backward caste (OBC)	0.95 (0.63-1.42)	0.80	1.83 (1.16-2.89)	0.01
Others/ General (GEN)	0.68 (0.32-1.45)	0.32	0.75 (0.33-1.68)	0.49
Education
Illiterate	1.00		1.00
Up to primary	1.48 (0.96-2.28)	0.07	0.88 (0.54-1.43)	0.61
Up to secondary	2.25 (1.33-3.81)	0.00	1.44 (0.82-2.54)	0.20
Above secondary	2.50 (1.22-5.11)	0.01	1.74 (0.85-3.57)	0.13
Occupation
Own Agriculture	1.00		1.00
Daily wages labourer	2.07 (1.32-3.25)	0.00	0.86 (0.52-1.42)	0.57
Self employed	1.45 (0.77-2.71)	0.24	1.25 (0.64-2.44)	0.51
Government Employee	4.27 (1.62-11.23)	0.00	3.85 (1.69-8.77)	0.00
Pvt. Salaried employee	4.25 (1.84-9.80)	0.00	2.32 (1.11-4.84)	0.02
Coal Industry Employee	1.62 (0.48-5.39)	0.43	2.54 (0.79-8.23)	0.12
Wealth Index
Poorest	1.00		1.00
Poor	2.41 (1.45-4.02)	0.00	2.13 (1.13-4.02)	0.02
Middle	2.43 (1.40-4.20)	0.00	2.85 (1.48-5.46)	0.00
Rich	4.30 (2.37-7.78)	0.00	4.27 (2.18-8.35)	0.00
Richest	6.28 (3.09-12.77)	0.00	3.66 (1.71-7.81)	0.00
Insurance coverage
No insurance	1.00		1.00
RSBY/MSBY	0.80 (0.49-1.32)	0.40	0.69 (0.39-1.21)	0.20

Table 9: Proportion of population screened positive for depression and difficulty in their day today life based on their response to PHQ-9 score, in 18 or more-year age group (n=839)

	PHQ-9 score more or equal to 8	PHQ-9 score more or equal to 10	PHQ-9 score more or equal to 11	Difficulty in the day to day life if PHQ>10 (n=81)
Total	18.47	11.20	8.10	86.17
Rural-Urban divide
Rural	20.62	12.33	8.29	89.06
Urban	15.00	9.38	7.81	80.00
Sex
Male	18.58	11.20	7.65	85.37
Female	18.39	11.21	8.46	86.79
Age group
18-29	10.63	6.28	5.31	69.23
30-44	15.72	10.06	5.97	84.38
45-59	22.01	11.48	7.66	91.67
60+	35.24	23.81	20.95	92.00
Social groups
Scheduled Tribe (ST)	20.38	11.70	8.68	93.55
Schedule Caste (SC)	24.62	15.38	11.54	85.00
Other backward caste (OBC)	15.86	10.48	7.26	79.49
Others/ General (GEN)	13.89	5.56	4.17	100.00
Education
Illiterate	31.64	16.95	12.99	90.00
Up to primary	16.53	11.57	9.50	89.29
Up to secondary	15.79	9.77	5.64	88.46
Above secondary	11.04	6.49	4.55	60.00
Occupation
Own Agriculture	23.31	12.88	9.20	95.24
Labourer/ Work on daily wages	17.75	11.01	7.87	83.67
Self employed	20.39	13.59	11.65	85.71
Government Employee	34.78	21.74	4.35	100.0
Retired Government Employee	45.45	27.27	27.27	33.33
Private employee	3.70	2.47	2.47	100.00
Coal Industry Employee	7.69	0.00	0.00	-
Wealth Index- Rural
Poorest	22.66	11.72	9.38	100.00
Poor	24.07	13.89	8.33	100.00
Middle	23.30	12.62	6.80	76.92
Rich	16.67	10.71	5.95	77.78
Richest	14.58	12.50	10.42	83.33
Wealth Index-Urban
Poorest	21.25	15.00	13.75	100.00
Poor	16.92	12.31	10.77	62.50
Middle	18.64	8.47	5.08	60.00
Rich	6.45	1.61	1.61	100.00
Richest	9.26	7.41	5.56	75.00

In this sample of persons screened for high blood sugar, 9.22% showed a positive test for high blood sugar but only 6.55% were known diabetics and 5.05% were on diabetes medication (Table 7). If we take all of the latent need as an unmet need for treatment, the unmet needs including known and latent diabetes works out to almost 45.2% of all suspected diabetes. If we take only known diabetics unmet needs are 22.90 % (Table 7).

Known diabetics and hypertensive (Perceived healthcare needs) were greater in higher socioeconomic groups but latent diabetes and hypertensive was greater in lower socioeconomic groups (Additional file 1, Table S 3). For example, for the poorest rural economic quintile, the total unmet healthcare needs in hypertension was 89.79% whereas in the richest urban quintile it was relatively less than 54.83%( Additional file 1, Table S 3).

3. Depression. As revealed by the screening protocol, prevalence of depression in our sample (PHQ-9 score> 10) was 11.9 percent in the18+ age-group and of these. Of those who screened positive less than 1 per cent were on normative treatment (see Table 9) despite over 80% having significant difficulties in c activities of daily living. In effect all cases of depression were cases were unmet needs for healthcare.

4. Rheumatic conditions. With rheumatic conditions, diagnosis mostly symptomatic, was universal and so was some form of treatment- usually discontinuous medication for pain relief and increased mobility. About 1.84 percent of the sample population had a rheumatic condition, but only 51.7% of them did not receive treatment at all or received treatment was inappropriate and incomplete (Additional file 1 Table S 6).

We illustrate with a case study:

A 58 years old SECL employee has had osteoarthritis for last ten years. She also has had diabetes for the last 15 years. She complains about knee pain when she walks and finds it difficult to sit on the ground while working on household's chores. She saw herself to a private practitioner in the city who gave drugs and asked her to follow up after three months. She felt much better after taking those medicines. Once medicine got over (after a month) she did not go to meet the doctor or continue the treatment. She was better next one month, but again knee pain started. She says she is already taking medicine for diabetes and she cannot take more drugs regularly for knee pain. She can neither afford it nor the regular consultations. She feels knee pain is a common problem with age, and she must accept it.

… Interview No. 20.08.2018.04, A 58 years old, Government employee

These should also be considered unmet healthcare needs, which will manifest with improved availability of services, increased awareness in patients and facilitation by community processes. But for now, these are difficult to quantify.

5. Disability in the elderly. Among the elderly, 10.70% of the elderly reported having any difficulty in basic activities of daily living (BADL- Additional 1, Table S7). On the other hand, 44.65% of the elderly had difficulty in any one of the IADL (Additional file 1, Table S 8). However most did not report this aa a healthcare need. This lack of perception of such healthcare needs could be highest in the elderly, where many treatable health conditions would be dismissed as a 'natural’ part of ageing and therefore denied care. Following are two case studies in the context of the elderly where family members perceived the health needs of the elderly member differently and had a different outcome.

A 63 years old woman from a village had a fall while carrying water for household work. After the fall, she was not able to walk and was bedridden. The family member did treatment by local remedies and applied turmeric and leaves on the hip areas, which was painful. She was not able to move for many days, and presently also she is bedridden, requiring help to even go to the toilet. Her mobility is restricted inside the house. The underlying pathology could be anything from a ligament tear to a fracture in hip areas. Upon asking a family member for not taking her to the formal provider, they replied that it happens in old age and is natural. Her son felt it is a common phenomenon in old age, and he has seen so many people like this in his village.

… Interview no: 01.08.2018.05, A 63 years old female of rural areas, disease condition: fall

A second case study in elderly due to fall:

A 70-year-old woman had a bull-gore injury while returning from the farm. She was bedridden for months. She was totally dependent on others. Even for the activities of daily living. She did not feel good and used to cry in despair. Her son would call some informal provider who would give her pain killer injection and the assumption was that no further care was possible. A month before the interview, her grandson, a primary school teacher in the nearby city, who learnt of this, facilitated her visit to a professional orthopaedic. With a plaster cast, physiotherapy, and a walking aid and with family member strained to continue the care, she recovered in four weeks and in 8 weeks could manage all her activities of daily living on her own. and is now independent. She was happy with the treatment provided in the Korba and surprised that they had not known of this earlier.

… Interview no: 09.08.2018.03, A 70-year-old female with hip fracture

The current methods of the survey if it had arrived before she was taken to the city would probably record the above episode as one of met healthcare needs, where a traditional healer or informal healthcare provider provided the care though it was completely inappropriate. Only the arrival of an educated, aware facilitator changes the perception and creates a healthcare need where none was there earlier. This is a larger problem for the elderly, where a large number of treatable conditions causing considerable suffering is perceived as ‘natural’ to ageing or due to financial barriers failing to seek care.

C. Perceived unmet healthcare need.

Individuals were also asked a direct question about whether any of their perceived healthcare needs, had remained unmet in the last 365 days, and 9.61% of respondents reported this positive (Additional File-1, Table S10), and it was significantly higher in upper socioeconomic compared to lower socioeconomic population group (Additional File-1, Table S11). This direct question can however lead to very misleading under-estimates. Since, 40.58% of individuals in acute ailment and 27.47% in the chronic ailments either did not had treatment at all or received from the informal care provider, not to speak about incomplete and inappropriate care (Table 4). We further illustrate this with an individual case study below:

"One of the qualitative interviews was of a 22-year-old male, daily wage labourer, collecting tendu leaves for his employment. This is a forest-based primary producer no one in the family was literate. It was a nuclear family where the husband, wife and one-year-old child lived. Wifes's age was 20 years. Their first child died due to malaria, the second child died due to malnutrition, the third child was a stillbirth, and the present child, which is fourth in number and is alive, has malnutrition. Six months back, husband had a snake bite where he was brought to PHC and then by private van to district hospital which was 95 Km away. Even in such a situation, when he was asked if they had any unmet healthcare needs in the last year, his reply was ‘no'".

…Interview no: 08.08.2018.07, A 22-year-old male, daily wage labourer (collection of 'Tend' leaves)

D. Inappropriate care: Overlap of perceived and unperceived unmet healthcare needs:

Health needs are also unmet if the health care provided was inappropriate. The Tanahashi framework refers to this as care that is not effective [11], and so does the UHC definition. Unmet healthcare needs in terms of the inappropriateness of care can be perceived and non-perceived. In perceived inappropriate healthcare, the patient understood that the nature of healthcare provided to them was inappropriate- often because in the care pathway, another provider informed and managed them better. On the other hand, in non-perceived inappropriateness of care, the patient did not know the inappropriate service provided to them. However, as compared to normative care, this was inappropriate. Only obvious cases are included.

We have depended mainly on our qualitative data to understand forms and determinants of inappropriateness care. There were 90 such in-depth interviews conducted, which focused on the care-seeking pathways, tracing the journey from first contact care to when normative care for that condition was accessed (Additional file 1, Table S9). Of these 90 patients, in 27, there was an improvement and/or satisfaction, whereas in 63, care remained inappropriate in some respects. In certain situations, patients themselves recognized the inappropriateness of care and in other situations, the researcher categorized it non-perceived healthcare needs based on patient case history. These qualitative in-depth interviews cannot estimate the prevalence of the problem, but they provide insight into the magnitude and nature.

One recurrent theme is that the healthcare seeker had a RSBY/MSBY card, and though the private provider could not provide the required care, they still admitted the patient and “swiped the smart card” (earned money from). After the sum assured was depleted or exhausted they communicated their inability to treat the patient and referred it to a highercentre. In many case studies, patients' financial resources were depleted by that time. In some, their health condition had also deteriorated.

One typical example is presented below.:

A 40-year-oldmale daily wage labourer., suffered a knife injury at night-time eight months before. The familyborrowed money, hired a vehicle, andtook the patient to the government district hospital. However,required services were not available and hehad to proceed to anearby privatehospital.He was admitted for two days there, andhis smart card (of the government-funded insurance program) was swiped and Rs. 25000 deducted from it. A further Rs. 10,000 was also taken in cash from the patient at the time of admission. After two days, the private hospital declared their inability to treat the patient, and the patient condition also deteriorated. The patient had tohire again a private vehicle, which charged Rs.5000 and shift the patient to the Government Medical College, Bilaspur.He got free treatment there though he had to purchase some drugs outside. The patient reported that the private hospital knew they could not manage his case, but had kept him for two days, charged him, and he had got cheated as well as a delay in treatment.

… Interview no: 11.07.2018.03, A 40 years old daily wage labourer

Another recurrent theme in the case studies was bizarrely inappropriate care from informal care providers. The diagnosis is completely missed in such instances, but the treatment is aggressively continued. Over the period, the condition worsens. Given below is a sample of such a form of inappropriate care:

A 14-year-old boy met with an injury, where a tree branch pierced his lung while travelling on the rooftop of the bus eight months before. He had difficulty in breathing. He first took treatment from a local informal care provider who treated him for malaria since he had developed fever. The informal provider treated him for ten days and charged over Rs. 400 to the family but there was no improvement in the boy's health. Then his parents took him to the city to a private hospital where a doctor diagnosed him with emphysema, and he was admitted.

… Interview no: 04.9.2018.02, 12.07.2018.01, A 14-year-old boy with emphysema following lung injury

Our study shows a relatively high level and variety of unmet healthcare needs. Indicators of physical availability of services or insurance coverage were not able to reflect these access problems, all of which happened in a context where there was physical access and a good insurance cover. This has been discussed as the difference between measuring access and effective access [11]. Community sample surveys that collect information on self-reported morbidity, services utilization, and out-of-pocket expenditures are the primary data sources for measuring effective access. However, this study shows a range of unmet needs that such surveys will fail to pick up. These are tabulated in figure 1.

Challenges in measuring healthcare needs. Studies on access like the health surveys of NSSO use response to a single question – “have you any illness now or in last 15 days?” to record morbidity and, therefore healthcare needs[20]. Other surveys like LASI and SAGE follow this single question with a list of leading questions asking specifically whether for the presence of the more common chronic illness[19]. The single question under-estimates chronic illness. In this study with a single question, the morbidity rate is 8.18% and with the additional set, the prevalence rate of chronic illness increases to10.62%. Even with perceived needs, using a set of questions instead of only one probe significantly increases the reporting for chronic conditions. Studies have shown that special interviewing techniques –respondents instruction, feedback, long question, probing – used singly or in combination increases reporting rates of chronic conditions [26].

But either way, such studies miss latent illness- and as shown in this study, a major part of unmet healthcare needs take this form. More recent demographic surveys[27] are adding blood pressure and blood sugar to the survey, but this only uncovers latent illness in a modest proportion of NCDs. Similar latency could extend to many more NCDs, mental illness and even chronic communicable diseases like tuberculosis, HIV/AIDS and leprosy.

Past studies, mostly done in developed countries, measure unmet healthcare needs via asking one-liner questions:" have you had unmet healthcare needs in the last 365 days?”. This study finds that such a question seriously under-estimates even perceived healthcare needs. Earlier studies too have reported this and attributed it to characteristics of the patient and the limitations of a one-year recall. [28, 29].

Measuring Inequities in healthcare needs and access. A higher level of unmet healthcare needs in the lower socioeconomic population measures unmet healthcare needs essential to understand inequity in access to care in society. In addition to missing latent needs, NSSO like health surveys, would miss incomplete or inappropriately met needs. In terms of self-reporting of ailment there is debate over subjective reporting versus objective reality [30].

This study notes that perceived health care needs(all) are higher in upper socioeconomic and less marginalized groups. If we include latent healthcare needs – the higher incidence of (perceived) healthcare needs in higher economic quintiles (mainly due to chronic illness and higher hospitalizations) reduces significantly. Amartya Sen had explained this phenomenon of states with higher socioeconomic states showing greater morbidity in morbidity surveys than lower socioeconomic states in terms of “positional objectivity’ [31]. Thus, Kerala, which has the highest literacy and density of health care facilities, and relatively high income per capita, has the highest morbidity in self-reporting based on NSSO surveys[20]. On the other hand, poorer states with lower literacy and educational attainment and more poverty are more at risk for disease but paradoxically report lesser morbidity than Kerala in the self-reporting based NSSO survey[20]. However, disease burden studies show that the poorer states have a higher disease burden. For instance, the relative age-standardized DALY rate in Chhattisgarh (1.71) was third highest in the country after Assam (1.81) and Uttar Pradesh (1.77), whereas it was minimum in Kerala (1.00) [32, 33]. Also, existing indicators to measure progress towards UHC do not factor in unmet healthcare needs.

This study showed that in contrast to perceived healthcare needs higher in higher socioeconomic groups within Chhattisgarh, unmet healthcare needs are higher in lower socioeconomic groups in the state- and this includes latent health needs and perceived health needs incomplete treatments and treatment from unqualified providers. Similar findings on hypertension were also reported from recent studies [27, 34]. Though we do not know its distribution across different socioeconomic groups, inappropriate care also contributes. From our qualitative studies we recognise that a fair proportion of what our survey recorded as a met need were instances of inappropriate care concealing what was ineffective care and therefore an unmet healthcare need.

Our major learning from this study is that uncovering unmet healthcare needs is a major task for equitable design of health care systems and measuring progress towards UHC. Since the prevalence of unmet healthcare needs is more in the poorer and more marginalized sections, measuring differentials in unmet healthcare needs across sub-groups with lower social and economic status is essential. Similar findings were also reported in other studies from LASI [34].

Not all researchers categorize care from informal providers as unmet health care needs. Many have made a case for acknowledging this as a legitimate market response to healthcare needs[35]. However, our discussion is in the context of measuring progress to UHC. In the road-maps to UHC in India, access to informal and unqualified providers has no place hitherto and, therefore, would not qualify as access to effective access to healthcare. Further, given our observations on inappropriate healthcare in informal providers as seen in this study, we are not even encouraging such a possibility.

Inappropriate care concealing unmet healthcare need. Merely being treated by a healthcare provider is not effective access. The care provided should conform to a normative understanding of what is the required healthcare. There could be excessive, inessential care, which is a problem in its own right, but if appropriate care is not part of the care provided, it remains an unmet healthcare need. To measure inappropriate care as a researcher is a challenging dimension, especially in a context where health records are very incomplete or absent, and communication to patients have been scanty. Only grossly inappropriate care as communicated by a physician or as uncovered by a researcher in consultation with a physician could be categorized as such. The Qualitative case studies in this study show that access to care providers did not often mean access to appropriate care. Further, sometimes individuals perceived inappropriateness of care and exited from the healthcare provider or despite this they continued as there was no other alternative available. And in other situations, the patient had not realized the inappropriateness of care provided to them. Previous studies done in India have also cautioned about inappropriate care[36, 37].

To this baseline, a new layer of problems is added with the introduction of government-funded health insurance programs. Our study documented several cases where private empanelled providers admitted and provided care to patients despite not having the capacity to provide appropriate care, billed the services for reimbursement, and then later referred the patient in a worse health condition with a depleted or exhausted insurance cover. A study conducted in an urban slum of Chhattisgarh also indicates high OOPE despite usage of health insurance cards [38].

The increase in inappropriate care has been reported from other studies in India [39]and elsewhere [40]. However, this study reveals that it has perhaps now become a distinct and increasing form of unmet healthcare needs. Further studies would be required to document the prevalence of such practices in different contexts.

The causes of the high proportion of unmet needs are evidenced both from survey data and the case studies. One major problem is that patients often access public healthcare facilities only to be told that the required services for that specific disease condition are not available. This is a consequence of health policies that have promoted very selective healthcare packages for government-provided primary healthcare- not only in the primary health centre but also in the district hospitals[41, 42]. Thus, there is either no population-based effort at uncovering latent needs or promoting appropriate healthcare-seeking behaviour for many healthcare needs. For many others where the need is perceived, though there is an accessible healthcare facility, the appropriate care is not available, and the patient is referred away. Recognizing this and as an effort to address this problem, universal coverage with government-funded public health insurance is introduced, but in the absence of private providers who can provide appropriate care, it does not lead to increased access. Rather because there are a reimbursement provisions available, inappropriate care increases in the local private sector. So often, the desired level of care is available in the government medical college hospital of the distant city and appropriate private care is also available at the same distance.

The simple learning is that improving mechanisms of financing health care and risk pooling is quite inadequate where the problems are large of access. In the Korba context, the immediate need is to strengthen the public health system and provide comprehensive primary health care. It is worth noting that the government has initiated a program called the Ayushman Bharat, Health and Wellness Centres for this. However, while early reports are encouraging, it is still too early to say whether it will address the considerable problem of unmet needs and equity in access to effective essential health care.

This study has its limitations. The morbidity prevalence rates measured in this study should be interpreted with caution because of several limitations. Firstly, the effect of seasonality is not captured since it’s a one-time cross-section study. Further latent need measurements are limited to five tracer conditions- if more had been included, more latent needs were likely to be uncovered.

Measurement of healthcare needs and measurement of unmet (or unfulfilled) healthcare needs are multidimensional and complex. Yet they are essential to understand the gaps in effective access to healthcare and plan to close these gaps to monitor progress towards universal healthcare. Further, because unmet needs, in contrast to perceived needs, are much higher in population sub-groups of lower socioeconomic status and greater marginalization, ensuring equity in access requires such measurement. Indicators measuring progress towards UHC should go beyond indicators reflecting utilization of services, including a few select indicators reflecting unmet needs and further seek to measure them across different relevant population sub-groups.

BADL: Basic activities of daily living

IADL: Instrumental activities of daily living

LASI: Longitudinal ageing study in India

MSBY: Mukhyamantri Swasthya Bima Yojana

NFHS: National Family Health Survey

PHQ-9: Patient health questionnaire-9

PPS: Probability proportional to size

RSBY: Rashtriya Swasthya Bima Yojana

SAGE: Study on global AGEing

UHC: Universal Health Coverage

UMPCE: Usual monthly per capita expenditure

Ethics approval and consent to participate: Written informed consent was taken from all respondents. Approval was obtained from the Institutional Ethics Committee of State Health Resource Centre, Chhattisgarh with Reference number: SHRC-19-07-2018. The data was anonymized.

Consent for publication: Not Applicable.

Availability of data and materials: The anonymized dataset, is available in the public domain at[https://drive.google.com/file/d/1iXXHznG9hBQCU8ODHZxE51TZllgu1qH7/view?usp=sharing].

Competing interests: The author declares there was competing interest.

Funding: None.

Author’s contributions: RA, TS, GS contributed to the study design, to the interpretation of results, and writing the manuscript. RA analyzed the data and TS and GS verified the results. All authors read and approved the final version of the manuscript.

Acknowledgements: Authors would like to thank Dr. Prabir Chatterjee, former Executive Director, SHRC, Dr. Indranil Mukhopadhyay, and Dr. Bino Paul for the valuable inputs in designing the study. Special thanks to Mr. Rajamani Pawar, Ms. Sarita Sarathi, Ms. Seema Sarathi, Ms Ambika Devi, Ms. Rudrani Prajapati, Ms. Kewara bai, and Ms. Pratibha Devangan for helping in data collection, and Mr. Uttam Thakur and Ms. Nisha Dewangan in data entry.

Boerma T, Eozenou P, Evans D, Evans T, Kieny M-P, Wagstaff A. Monitoring Progress towards Universal Health Coverage at Country and Global Levels. PLOS Medicine. 2014;11:e1001731.
WHO. Making fair choices on the path to universal health coverage: Final report of the WHO Consultative Group on Equity and Universal Health Coverage. 2014.
Goddard M, Smith P. Equity of access to health care services:: Theory and evidence from the UK. Social Science & Medicine. 2001;53:1149–62.
Gulliford M, Figueroa-Munoz J. Introduction: Meaning of’Access’ in Health Care. Access to Health Care. M. Gulliford and M. Morgan. London and New York. Routledge; 2003.
Aday LA, Andersen R. A framework for the study of access to medical care. Health services research. 1974;9:208.
Guagliardo MF. Spatial accessibility of primary care: concepts, methods and challenges. International journal of health geographics. 2004;3:3.
Wang F, Luo W. Assessing spatial and nonspatial factors for healthcare access: towards an integrated approach to defining health professional shortage areas. Health & Place. 2005;11:131–46.
Falkingham J. Poverty, out-of-pocket payments and access to health care: evidence from Tajikistan. Social Science & Medicine. 2004;58:247–58.
McIntyre DI, Thiede M, Birch S. Access as a policy-relevant concept in low-and middle-income countries. HealtH economIcs, polIcy and laW. 2009;4:179–93.
Abiiro GA, De Allegri M. Universal health coverage from multiple perspectives: a synthesis of conceptual literature and global debates. BMC International Health and Human Rights. 2015;15:17.
Tanahashi T. Health service coverage and its evaluation. Bull World Health Organ. 1978;56:295–303.
Ranjan A, Dixit P, Mukhopadhyay I, Thiagarajan S. Effectiveness of government strategies for financial protection against costs of hospitalization Care in India. BMC Public Health. 2018;18:501.
Kettner PM, Moroney RM, Martin LL. Designing and managing programs: An effectiveness-based approach. Sage Publications; 2012.
Govt. of Chhattisgarh. Government of Chhattisgarh: Rashtriya Swathya Bima Yojana ( RSBY) and Mukhyamantri Swasthya Bima Yojana (MSBY). 2019. http://cg.nic.in/healthrsby/. Accessed 6 Mar 2019.
Govt. of Chhattisgarh. List of Hospitals under Ayushman Bharat PMJAY & Mukhyamantri Swasthya Bima Yojana, Chhattisgarh as on 07/12/2018. 2018.
Govt. of India. 11th Common Review Mission Report 2017. New Delhi; 2017.
Sonwani F. District Wise Comparative Analysis of Health Facilities in Chhattisgarh State. International Journal of Healthcare Sciences. 2014;2:57–65.
Government of India. National Family Health Survey. 2018. http://rchiips.org/nfhs/nfhs4.shtml. Accessed 11 Feb 2022.
Bloom DE, Sekher TV, Lee J. Longitudinal Aging Study in India (LASI): new data resources for addressing aging in India. Nat Aging. 2021;1:1070–2.
Government of India. NSS 75th Round-Key Indicators of Social Consumption in India: Health. New Delhi; 2019.
Govt. of India. Standard Treatment Guidelines. 2016. http://www.nhm.gov.in/nrhm-instate/520-standard-treatment-guidelines.html. Accessed 6 Mar 2019.
Patel V, Araya R, Chowdhary N, King M, Kirkwood B, Nayak S, et al. Detecting common mental disorders in primary care in India: a comparison of five screening questionnaires. Psychological medicine. 2008;38:221–8.
Avasthi A, Varma SC, Kulhara P, Nehra R, Grover S, Sharma S. Diagnosis of common mental disorders by using PRIME-MD Patient Health Questionnaire. Indian Journal of Medical Research. 2008;127:159.
Poongothai S, Pradeepa R, Ganesan A, Mohan V. Prevalence of depression in a large urban South Indian population—The Chennai Urban Rural Epidemiology study (CURES–70). PLoS One. 2009;4:e7185.
Kochhar PH, Rajadhyaksha SS, Suvarna VR. Translation and validation of brief patient health questionnaire against DSM IV as a tool to diagnose major depressive disorder in Indian patients. Journal of postgraduate medicine. 2007;53:102.
Jabine TB. Reporting chronic conditions in the National Health Interview Survey. A review of findings from evaluation studies and methodological test. Vital Health Stat 2. 1987;:1–45.
Lee J, Wilkens J, Meijer E, Sekher TV, Bloom DE, Hu P. Hypertension awareness, treatment, and control and their association with healthcare access in the middle-aged and older Indian population: A nationwide cohort study. PLOS Medicine. 2022;19:e1003855.
Marshall EG, Wong ST, Haggerty JL, Levesque J-F. Perceptions of unmet healthcare needs: what do Punjabi and Chinese-speaking immigrants think? A qualitative study. BMC Health Services Research. 2010;10:46.
Rhodes AE, Fung K. Self-reported use of mental health services versus administrative records: care to recall? International Journal of Methods in Psychiatric Research. 2004;13:165–75.
Prinja S, Jeet G, Kumar R. Validity of self-reported morbidity. Indian Journal of Medical Research. 2012;136:722.
Sen A. Positional Objectivity. Philosophy & Public Affairs. 1993;22:126–45.
ICMR, PHFI, IHME. India: Health of the Nation’s States : The India State-Level Disease Burden Initiative. New Delhi; 2017.
Dandona L, Dandona R, Kumar GA, Shukla DK, Paul VK, Balakrishnan K, et al. Nations within a nation: variations in epidemiological transition across the states of India, 1990–2016 in the Global Burden of Disease Study. The Lancet. 2017;390:2437–60.
Mohanty SK, Pedgaonkar SP, Upadhyay AK, Kämpfen F, Shekhar P, Mishra RS, et al. Awareness, treatment, and control of hypertension in adults aged 45 years and over and their spouses in India: A nationally representative cross-sectional study. PLOS Medicine. 2021;18:e1003740.
Das J, Chowdhury A, Hussam R, Banerjee AV. The impact of training informal health care providers in India: A randomized controlled trial. Science. 2016;354.
Barker AK, Brown K, Ahsan M, Sengupta S, Safdar N. What drives inappropriate antibiotic dispensing? A mixed-methods study of pharmacy employee perspectives in Haryana, India. BMJ Open. 2017;7:e013190.
Bhandari N, Bahl R, Taneja S, Martines J, Bhan MK. Pathways to Infant Mortality in Urban Slums of Delhi, India: Implications for Improving the Quality of Community- and Hospital-based Programmes. Journal of Health, Population and Nutrition. 2002;20:148–55.
Nandi S, Schneider H. When state-funded health insurance schemes fail to provide financial protection: An in-depth exploration of the experiences of patients from urban slums of Chhattisgarh, India. Global Public Health. 2020;15:220–35.
World Bank. Is that surgery really necessary? Ensuring the Medical Necessity of Care. World Bank. 2014. http://www.worldbank.org/en/news/feature/2014/07/30/is-that-surgery-really-necessary. Accessed 4 Mar 2019.
Mi J, Liu F. Rate of caesarean section is alarming in China. The Lancet. 2014;383:1463–4.
Sengupta A, Zaidi S, Sundararaman T, Onta S, Weerasinghe MC. Tackling the primary care access challenge in South Asia. bmj. 2018;363:k4878.
Sundararaman T, Muraleedharan VR, Ranjan A. Pandemic resilience and health systems preparedness: lessons from COVID-19 for the twenty-first century. Journal of Social and Economic Development. 2021;:1–11.

No competing interests reported.

AdditionalFile.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Measuring unmet healthcare needs - findings from a tribal district in the Chhattisgarh State of India

Status:

Version 1

Abstract

Figures

Background

Methods

Results

Discussion

Conclusions

List of abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1