Healthcare Utilization of Below Poverty Line Population under Government-Funded Health Protection Scheme in one sub-district in Bangladesh: A Cross-Sectional Study

Abstract


Background
Universal health coverage (UHC) is now a familiar acronym for the global health agenda that combines a triple policy.It is de ned as ensuring universal access to quality healthcare services as per need without the risk of incurring nancial hardship [1].More recently, the adaption of the sustainable development goals (SDGs) by the UN assembly explicitly included the UHC goal under SDG 3.8 stating "Achieve Universal Health Coverage, including nancial risk protection, access to quality essential healthcare services, and access to safe, effective, quality, and affordable essential medicines and vaccines for all" [2].Access to healthcare is considered as an important determinant in assessing equity in healthcare delivery [3].Ideally, utilization of healthcare services re ects a need for care, however, that is not possible always for several reasons [4].In low-and middle-income countries (LMICs), healthcare utilization is in uenced not only by the demand side constraints but also by the supply side constraints.In some instances, such utilization is determined solely by the ability to pay or out-of-pocket (OOP) rather than the need for care [5].This situation can gradually impose heavy nancial burdens on individuals as well as households and in certain instances can lead to catastrophic healthcare expenditure (CHE) and economic impoverishment [6,7].OOP spending is the major sources of healthcare payment in most of the LMICs including Bangladesh.The healthcare system of this country predominantly relies on OOP payment nancing.It was estimated that, in 2015, around 67% of the total healthcare expenditure was made as OOP [8].Such high percentage of OOP expenditure for healthcare is associated with low nancial protection of the households.Among 14 Asian countries, 15.6% of households faced CHE because of the high burden of OOP payments [9].A recent study revealed that 14.7% of the household faces CHE from healthcare related OOP payments and 3.5% of the total population fall into poverty annually [10].Another study conducted among 11 Asian countries showed that about 5 million people get impoverished annually in Bangladesh from OOP healthcare expenditure [11].
Public spending on health is determined by the capacity of the Government to raise revenues and allocate it to health sector [12].Bangladesh still spends only 3% of its GDP in health sector while government expenditure in relation to GDP is only 0.69% placing Bangladesh at the least spending countries on health in South-East Asian Region [13].The signi cantly increasing healthcare cost from the non-communicable disease burden and epidemiological transition (responsible for 60% of the mortality) have led to increase the OOP spending for healthcare.The National Health Policy of 2011 of the Government of Bangladesh (GOB) acknowledges that health is a human right and to achieve UHC, it is necessary to ensure healthcare services for the poor in an affordable cost [14].For achieving UHC by 2030, effective measures have to be taken for reducing households' OOP healthcare related expenditure and increasing utilization of quality healthcare.Responding to the global call for UHC, the Government of Bangladesh adopted the rst ever Health Care Financing Strategy (HCFS) in 2012 aiming to achieve UHC by 2032 [15].Under the HCFS, the Shasthyo Surokhsha Karmasuchi (SSK), a health protection scheme of the GOB, is being piloted by the Health Economics Unit (HEU) of the Ministry of Health and Family Welfare (MoHFW) [16].Although the SSK has a comprehensive plan to bring all the population of the country under the health insurance, initially aimed to pilot targeting the below poverty line (BPL) population only.

Shasthyo Surokhsha Karmasuchi (SSK)
This scheme is a model of health insurance that has started for piloting at Kalihati Upazila of Tangail District in 2016 and other two Upazilas -Modhupur and Ghatail of the same district in 2017.The scheme provides healthcare to the identi ed and enrolled BPL population.Upazila Health Complex (UpHC) is the rst contact point of the bene ciaries.Enrolled households get inpatient care through UpHCs and on a structured referral basis from the Tangail District Hospital (TDH).Each household is provided with an electronic health card ensuring 50,000 BDT coverage per year for inpatient care of 78 different diseases against a government nanced premium of 1,000 BDT per household per year.The key actors in the implementation of the SSK scheme were SSK cell, Scheme Operator (SO) and contracted hospitals.A detail description of the scheme is published elsewhere [17].
Health insurance scheme is viewed as an alternative to improve access to healthcare and reduce the direct nancial burden in using care [18,19].Though it is often predicted by the theory of health insurance, it does not always provide the expected nancial protection for the "insured" people [20][21][22], since health is a multidimensional issue.A study found that insurance scheme increases healthcare utilization and reduces OOP payments among the bene ciaries [7].The outpatient visits were favorable to the poor but the poorest had less access to inpatient care [23].In India and Africa, Community Based Health Insurance (CBHI) had the positive effect on reducing the event of CHE [24][25][26][27].On the other hand, subsidized health insurance scheme also worked for Columbia as it had positive effect on OOP payment reduction and health care utilization [28,29].Askeskin program, a subsidized social health insurance in Indonesia had improved access to healthcare facilities and led an increase in outpatient utilization [30,31].The evidence from Sokapheap Krousat Yeugn (SKY) micro-health insurance programme showed positive effect on utilization, reduced OOP health expenditure, but no effect on health status [32].Another study conducted in Cambodia has shown that the integrated Social Health Protection Scheme (iSHPH) has enhanced HEF bene ciaries' utilization of public health services [33].The evidence from Vajpayee Arogyashree scheme (VAS) suggests that it led to a substantial reduction in mortality driven by increased tertiary healthcare utilization.[34].Focusing on Indian publicly nanced health insurance schemes, an evidence prevails that the hospital services utilization increased after the introduction of these insurance schemes.Moreover, this increase in utilization has sustained over time and across regions [35].
Empirical evidence on the utilization of healthcare through SSK scheme is rare.Moreover, the Government is going to expand the scheme in some Upazilas initially and plan to implement across the country gradually.In this context, evidence on utilization of SSK scheme will help the key policymakers to develop more effective strategy for future implementation of the scheme.Thus, we aimed to assess the utilization of SSK scheme among the enrolled BPL population and identi ed the factors associated with the utilization.

Study design
An exploratory study design with a cross sectional survey was used to examine the utilization of healthcare services through SSK scheme among insured BPL households.A community survey was conducted among the enrolled households of the scheme.

Study setting
The study was conducted in the Kalihati Upazila (sub-district) of Tangail District where the SSK is currently being piloted.The total households of this Upazila is 89,351 of which 35,740 (40%) BPL households were enrolled in the SSK scheme.We included only Kalihati subdistricts for this study as the scheme here was comparatively matured than other two intended sub-districts i.e.

Data collection
A cross-sectional community survey was conducted using a standard pretested questionnaire.Information on socio-demographic characteristics of the households, illness and treatment related information of the members, and knowledge about the SSK scheme were collected.A six-member team of trained and experienced eld research assistant (5 data collectors and a supervisor) interviewed the respondents from the selected households.Informed written consent was obtained before their participation in the survey.The survey was conducted between July to September 2018.

Sampling
The sample of the community survey was determined based on the utilization of such health protection scheme among the BPL population.A study showed that overall 19% of the insured population utilized healthcare from a health protection scheme called Rashtriya Swasthya Bima Yojana (RSBY) in India [36].Based on this utilization rate, considering 95% con dence interval and 3% error level, a minimum of 657 households were required to interview for this study.Assuming 5% non-response rate and 1.2 design effect for clustering, total 828 households were selected for the interview.
A two-stage cluster sampling method was adopted two select the sampled households in the selected Upazila..The SSK household list maintained by the Scheme Operator was used as sampling frame.There are 15 unions (collection of villages), in Kalihati Upazila.In the rst phase, the unions were classi ed into three subgroups i.e. nearer, medium, and far using the distance of the unions to the Kalihati UpHC.Information on the distance was collected from the SSK Scheme Operator before the sampling.As the number of enrolled SSK households were not equal in each union, the estimated total sample was proportionately allocated for a union based on the total SSK enrolled households of the union.In the second phase, simple random sampling technique was used to select the SSK households from the lists of SSK households in each union.However, while visiting the sampled SSK households, we found that some of the SSK households didn't receive SSK cards yet and were not eligible to receive health care under the scheme.In such cases, instead of interviewing the listed households, we interviewed the adjacent households who had SSK card.

Analysis
Descriptive statistics of the household members and statistical inference test has been performed.In the descriptive statistics, the characteristics of the study participants were presented regarding frequency (n) and percentages (%) with 95% con dence interval (CI).Utilization of healthcare has been presented by several dimensions e.g.sources of care, types of illness suffered (inpatient was de ned as who required admission and stayed in the hospital overnight, and outpatient was de ned as who didn't require admission to the hospital or require admission but didn't stay overnight), and types of care utilized.Beyond the utilization of SSK scheme we classi ed other two sources of healthcare utilization, rstly, medically trained providers (MTP) which includes care from medical college hospital, UpHCs, general practitioner, non-government organization (NGO) hospital, quali ed private practitioner, Union Sub-Centre (USC), Maternal and Child Welfare Centre (MCWC) and secondly medically non-trained provider that includes unquali ed private practitioner, traditional healer, drug sellers, and other unquali ed sources.
A multiple logistic regression model was used to predict the likelihood of healthcare utilization from the SSK facility while controlling for demographic and household socioeconomic characteristics.The model is speci ed as: Where, Y i = utilization status from SSK (1=yes, 0=No), X 1 X 2 … are determining characteristics of healthcare utilization, β 1 , β 2 , … represents the coe cients of association and u i of the model.

Results
Socio-demographic Characteristics Among the selected 828 households, nally 806 were interviewed in this study (97% response rate).These households consisted of 3,178 members, mostly adults (58.4%) followed by children (35.6%), and elderly (6.1%) (Table -1).The proportion of male to female was almost equal.More than half of these members were married.The majority of the members (47%) had no education, while 27.5% of had primary, and 25.5% had secondary or above education level.About 27% of the household members were housewives and around 15% were agricultural workers or day labors.Most of the households (58.6%) had 4-5 members, followed by less than 4 members (24.9%), and more than six members (16.5%).Around 36% individuals lived 15+ km away from the SSK facility while around 20% individuals stayed closer to the SSK facilities (less than 5km).More than one third of members belonged to the households having less than 10,250BDT (122 USD) income per month.
As per BPL household selection criterion, we found 66% of the SSK households were remained BPL at the time of interview.
Regarding the knowledge of insurance scheme, about 68% of the visited SSK households knew about the SSK scheme.A higher proportion of bene ciaries (63%) knew about the service provision of SSK scheme, however, their knowledge about the bene t package of SSK scheme was found to be low.The majority (56%) of the bene ciaries reported of knowing about SSK scheme from the SSK representatives, followed by 25% from familiar people in locality, and 17% from relatives.

Utilization of healthcare
In the last 90 days recall period, almost 25% (n=781) of the surveyed individuals had self-reported illness or symptoms.Among the ill individuals, 81.8% (n=639) sought healthcare for their illness or symptoms from any types of healthcare providers.Only 8% of these individuals went to SSK facility for seeking healthcare.A large percentage (63.8%) of them went to 'non-medically trained provider'.The majority of the individuals utilized healthcare from pharmacy (45.9%), followed by public providers (20.7%), and private providers (18.9%).Of the total ill individuals, 3.6% (n=23) utilized inpatient care (IPC) from any sources and only 10 of 23 (43.5% who required IPC)utilized IPC from SSK healthcare facility within 90 days prior to the survey.
While examining the utilization of SSK facilities by the background characteristics of the individuals, we found that the elderly (10.4%), male (7.2%), unemployed (12.1%), having primary level education (9.1%), having more than 6 members in the households (9.1%), had accident or injury (17.2%), and income level lower than 10,250 BDT (15.3%) and lived in 5 km or less distance from SSK facility (17.9%) had comparatively higher utilization from SSK facility compared to other corresponding groups. .

Healthcare seeking behavior
We found that, majority of the people utilized healthcare from the drug sellers (pharmacies) (45.9%).Among the respondents utilized healthcare from medically trained provider, 12.5% of them went to private/NGO hospitals followed by scheme facility (SSK UpHC), and general practitioner (7.5%), Medical college/specialized hospitals (5.5%), and other quali ed providers (USC/MCWC) (2.7%).
While comparing the utilization by types of provider and types of reported illness, we found that 19.2% individual utilized SSK facility for accident/injuries while 34.6% from other MTP and 46.2% from non-MTPs.. Utilization from other MTP was the highest individuals who required female reproductive health and delivery services (76.5%) and had non-communicable diseases (47.2%).Utilization of non-MTP was the highest for communicable disease (66%), Accident and injury (46%) and for other illness (57%).

Associated factors of healthcare utilization from SSK facility
Table 4The utilization of health care from SSK facility were signi cantly associated with occupation, education, types of selfreported illnesses, knowledge of SSK scheme, living distance of health facility, and current BPL status of the household.We found that unemployed are about four times (OR: 4.175; 95% CI 1.01 -17.24) more likely to utilize services from SSK facility compared to the agricultural/labor.Individuals having secondary or higher level of education were less likely to utilize healthcare from the SSK facility.Patients with accident and injury were almost ve times more likely (OR: 5.11; 95% CI: 1.26 -20.78) to utilize healthcare from SSK facilities than patients suffered from non-communicable disease.Member of households with 4-5 members were less likely to utilize healthcare from SSK facility compared to the households with less than 4 members.Individual who heard about SSK scheme were more likely to utilize healthcare from the SSK facility (OR: 10.61; 95% CI: 2.39 -47.06).We observed that non-BPL household were above two times more likely to utilize healthcare from the SSK facility.Household distance of 15 km from the SSK facility were less likely to utilize healthcare from the scheme.We found that individuals from second income quintile were signi cantly utilized healthcare from the SSK facility compared to the poorest income quintile.Associated factor of healthcare utilization from any medically trained provider (MTP) Considering healthcare utilization status from any MTP (including SSK facility) as dependent variable, we found that having secondary and above level education were signi cantly associated with the MTP utilization compared to the no-institutional group.Similarly, people with self-reported injury and female reproductive problem utilized healthcare more compared to the people who suffered from non-communicable disease.Household size, knowledge of SSK scheme, BPL status and distance of nearest healthcare facility, and income quintiles didn't have any association with the utilization of MTPs.

Discussion
Overall, 8% of the cardholders who sought care for illness in the last 3 months visited SSK facilities and of the 23 cardholders who sought IPC, less than half (10 of 23 patients) of them sought IPC under SSK.The results of this study revealed that although the SSK health protection scheme provides free inpatients care and outpatient consultation services without fee and regular medicines, it is yet to be popular among the enrolled BPL households.A study conducted in India showed that such scheme signi cantly increased the utilization of inpatient healthcare among the bene ciaries [37].. Most of the enrolled people usually had symptoms and majority of them took healthcare services from pharmacies (46%)..This segment of individuals who require outpatient cares can be attracted under the scheme through introducing outpatient bene t package.Evidence showed that involvement in insurance scheme resulted in increased utilization of outpatient care at public facilities [38].
One of the key reasons for such low utilization of the SSK scheme was the lack of knowledge about the scheme among the enrolled households.We found households' knowledge was signi cantly associated with the utilization of healthcare by the members from the scheme facility.Studies showed that knowledge of insurance scheme signi cantly affect the utilization of healthcare and dropout rates [39,40].Although, the SSK management has developed an information, education, and communication (IEC) strategy, majority of the households didn't have knowledge on SSK scheme during the survey.This IEC strategy should be properly implemented among the target population.Moreover, community campaigns need to be strengthened to increase the demand for healthcare under this scheme.Households' distance from health facility also in uenced the utilization of healthcare from the scheme.It is evident in literatures that spatial distribution signi cantly affect the utilization of healthcare from health facilities [41][42][43].Even it is true for such health protection scheme that provides' free inpatient healthcare through the existing health facility.We observed that individuals from the currently non-BPL households were more likely to utilize healthcare from the SSK scheme.These individuals may reside near the scheme facility and may be better informed about the SSK services.It was evident that the individuals who belonged to the second income quintile were more likely to utilize healthcare from the SSK scheme compared to the poorest income group.The stylized fact is that utilization of healthcare is the lowest among the least well off group [44].
We found that unemployed individuals were more likely to utilize healthcare services from SSK facilities.The target population were poor; thus, the wage earner may have the higher opportunity costs of visiting to the healthcare facilities compared to the unemployed.Utilization of the scheme facility was signi cantly higher among the individuals suffered from accident/injury compared to the individuals suffered from communicable diseases.Similar nding was observed for a study conducted in a community-based health insurance scheme in Bangladesh [45].
The SSK scheme has a potential to increase the utilization among the enrolled BPL households by addressing the identi ed factors namely increasing knowledge through rigorous community campaign [46] and adding bene ts for the out-patients under the scheme rather than only providing the free consultation services [47].However, such mechanism needs to be examined in terms of sustainability and moral hazard.Further, list of BPL households should be updated at regular intervals for precise targeting and providing services to them who cannot afford it.Such precision may be possible by increasing local level political commitment.As the utilization of care was signi cantly negatively associated with the distance of healthcare SSK facility , the authority can probably think of linking the lower level facilities such as union sub-centre where the patient will get outpatient care and be referred to UpHC for inpatient services.However, further exploration would be required for such initiative.A comprehensive evaluation of this scheme including all three upazilas for examining the effect of the scheme in increasing healthcare utilization in comparison to the people from similar socio-economic group would be useful.

Limitations
One possible shortcoming of this study is that although the sample household was selected using simple random sampling technique, due to the real scenario at led we found the sampled households were not eligible to get SSK services due to not having SSK cards, we interviewed the adjacent households instead of the selected households.Interview of such adjacent household may create selection bias.However, we used 3% margin of error level to capture more households to reduce such bias.
This was the rst study that examined the healthcare utilization pattern and associated factors among the SSK BPL households in Bangladesh.Another limitation of this study was that the seasonal variation in utilization of healthcare has not been captured as the survey carried out from July to September 2018 which restricts us to establish the causality in vigorous way.Although we used the distance variable in the regression model, some other important variables (e.g.travel time and cost) were not considered as these may vary according to the road types and transportation facility.There was a possibility of recall bias as health service utilization data was collected using a self-reported questionnaire.Earlier studies used recall periods of 1 to 12 months for collecting similar data [48][49][50].We used a 90 days recall period to reduce the possibility of recall bias.

Conclusion
This study investigated the level of healthcare utilization of the nancially disadvantaged BPL population from the Government funded health protection scheme in Bangladesh.Utilization of healthcare from the scheme facility is low despite the provision of free IPC for 78 different types of diseases and outpatient consultation.Here, inadequate knowledge on scheme is one of the key barriers to utilization which might increase over time with appropriate measures e.g., developing and implementing interpersonal communication strategy, and organizing regular awareness building campaign.Further, as large number of people utilized care from non-quali ed and other MTPs, the policymakers can think of redesigning the bene t package based on needs of the BPL population and motivate them to seek care from SSK facility.The evidence generated from this study will be useful to address the reasons behind low rate of scheme utilization and challenges to increase the effectiveness of SSK in terms of healthcare utilization among the target vulnerable group.Addressing the drawback associated with the healthcare utilization is important before further expansion or scale of the scheme.A successful scale of this scheme will be a vehicle to achieve UHC through addressing the need of healthcare for the people who cannot afford quality healthcare in Bangladesh.All respondents of the study were interviewed after obtaining written informed consent and their participation was be voluntary.

Consent for publication
Not applicable

Funding
The study was funded by Swedish International Development Cooperation Agency -Sida (Grant #: GR-01455).The funding body was not involved in the design of the study, collection, analysis, and interpretation of data, and writing of this manuscript.

Figures Figure 1
Figures MEC, and SA contributed to conceptualize the research idea, study design, literature search, writing, revising, and nalizing the manuscript with the support from GGM, ZI and JAMK.All authors read, revised, and approved the nal version of the manuscript.Ethics approval and consent to participateThis study was approved by the Research Review Committee and Ethical Review Committee of the icddr,b (Protocol# PR-17047).

Table 1 .
Background characteristics of the study population

Table 2 :
Households' current BPL status, knowledge, and sources of information about the SSK scheme.