Introduction

The right to affordable quality healthcare is a basic right that every person should have and it is one of the objectives of healthcare systems globally. In India many people live in households that are very poor and subject to healthcare induced out of pocket expenses causing further impoverishment. As per the World Bank[1], OOP expenditure on health is 39.4%. and contributes to medical impoverishment. Hence, financial risk protection during ill health is an important goal in India’s journey towards Universal Health Coverage (UHC) and thus to address this issue, the Ayushman Bharat-Pradhan Mantri Jan Arogya Yogna (AB-PMJAY) scheme was launched on 23 September 2018. AB-PMJAY is a health assurance programme funded by the government to provide cashless and paperless hospital care for beneficiaries (with annual family income below INR 5 lakh (approximately USD 5000), informal workers and citizens above 70 years). Currently, as per these criteria there are around 120 million families (roughly 550 million individuals) eligible for this scheme. The scheme covers all direct expenses during hospitalisation (including preexisting diseases), day care procedures, pre-hospitalisation (3 days) and post hospitalisation (15 days) expenses (diagnostics and medicines)[2]. However, it excludes outpatient consultation expenses and medicine expenses not associated with hospitalisation. As on early 2026, the scheme has empaneled 37,385 hospitals nationally and distributed 435.2 million individual cards of which 41 million were distributed in the state of Bihar (which has 1188 empaneled hospitals).

Despite macro-level achievements, empirical data consistently indicate a substantial divergence between design intent and implementation outcomes, particularly evident within low-income states in India such as Bihar. Despite including over 10 million families eligible for the scheme, Bihar consistently demonstrates some of the lowest performance metrics regarding enrolment, utilisation, and financial protection[3]. Specifically, Bihar's card-to-hospitalisation ratio[4] is approximately 2.2%, a figure markedly lower than those observed in states such as Tamil Nadu (with 9.05 million admissions) and Karnataka (with 6.60 million admissions). Furthermore, Bihar accounts for 16% of all inter-state AB-PMJAY hospitalisations, which suggests a potential inadequacy in local healthcare facility[5]. Given the above context, this study aims at understanding the implementation of the AB-PMJAY scheme in Nawada district of Bihar.

Study objectives:

1. To assess the level of awareness of AB-PMJAY and its various features among eligible beneficiaries in Nawada district, Bihar.

2. To study the utilisation pattern of the AB-PMJAY scheme among eligible beneficiaries and understand the barriers to utilisation.

3. To measure the extent of financial risk protection for healthcare costs among the study population.

Methods

The study used a mixed-methods cross sectional design with data collected during May- July 2025. The quantitative component used a structured interview schedule (n=130) to capture data on awareness, utilisation, and financial risk protection from eligible households. The qualitative component comprised 20 in-depth interviews (IDIs) with eligible beneficiaries to contextualise, deepen, and interpret the quantitative findings. Healthcare provision within Nawada district is underpinned by a hierarchical network of governmental facilities, including the District Hospital, Sub-district Hospitals, Community Health Centres (CHCs), and Primary Health Centres (PHCs). This is complemented by a limited presence of private nursing homes and clinics, primarily situated in Nawada town. A substantial portion of secondary and tertiary care services is accessed either at the District Hospital or via referrals to larger urban centers such as Gaya and Patna.

Inclusion and exclusion criteria

For inclusion in the study, households had to satisfy several conditions: Firstly, they were required to be officially registered as AB-PMJAY-eligible based on the Socio-Economic Caste Census (SECC) data or possess valid AB-PMJAY cards. Secondly, at least one household member had to experience an inpatient hospital admission lasting a minimum of one night, within the 12 months preceding the data collection phase. Thirdly, an adult respondent from the household needed to provide written informed consent to participate. Accordingly, households were excluded if their reported hospitalisation fell outside this 12-month recall window, if they were unable to provide adequate confirmation of AB-PMJAY eligibility, or if they declined to participate in the study.

Data collection instruments

The research employed two distinct instruments for data collection. These included: (i) a pre-tested structured interview schedule, designed to gather information on household profiles, AB-PMJAY awareness, scheme utilisation patterns, and financial risk protection; and (ii) a semi-structured IDI guide, which facilitated exploration of scheme experiences, hospital encounters, financial impact, and suggestions from beneficiaries. The interview schedule underwent pilot testing with 10 households in a village not included in the main study. During this phase, ambiguities noted within the financial risk protection section and caste coding protocols were clarified. Structured interviews typically spanned 20–25 minutes, while IDIs extended to 40–60 minutes. All interviews were conducted in Hindi, audio-recorded with participant consent, subsequently transcribed, and then translated into English.

Measurement of catastrophic health expenditure

Out-of-pocket expenditure (OOPE) per hospitalisation episode was calculated by integrating all direct financial expenses specifically consultation fees, surgical procedures, prescribed medications, and diagnostic services, while indirect expenditures included transportation, food, and accommodation for both the patient and any accompanying caregivers. Subsequent to this aggregation, any monetary reimbursements received were subtracted from the total. Consistent with the methodological framework articulated by Wagstaff and van Doorslaer[6], households were deemed to experience catastrophic health expenditure (CHE) if their computed OOPE surpassed 10% of their total monthly consumption expenditure. This evaluation of CHE was performed twice, first considering only direct costs, and then incorporating the combined sum of both direct and indirect costs.

Quantitative analysis

Initially, quantitative data underwent double-entry validation to ensure accuracy. Subsequent analysis involved descriptive statistics, specifically frequencies, percentages, means, and standard deviations. To ascertain the independent predictors of scheme awareness, utilisation, and the incidence of catastrophic health expenditure (CHE), binary logistic regression was performed. This model adjusted for several covariates, including education level, caste, gender, housing type, and possession of the AB-PMJAY card. The findings are presented as adjusted odds ratios (AOR) with corresponding 95% confidence intervals.

Qualitative analysis

The IDI transcripts were analysed using thematic analysis as described by Braun and Clarke[7]. First a list of themes was made based on the three goals of the study. Additional themes emerging while analysing the IDI transcripts were also included in the analysis. A narrative synthesis of the qualitative findings was conducted and interpreted.

Ethical considerations

The approved IRB guidelines of the Tata Institute of Social Sciences, Mumbai were followed in this study. Informed written consent was obtained from all participants who were able to read and write. For individuals who were unable to read, the consent form was read aloud completely, and their thumb impression was accepted as documented assent. Participation in the study was entirely voluntary, and all participants were clearly informed of their right to withdraw at any point without facing any adverse consequences. Confidentiality was maintained by anonymising all collected data and storing audio recordings on a password-protected device.

Results:

Socio-demographic profile

Table 1 provides the socio-demographic attributes of the 130 eligible beneficiaries who consented for this study. 75.4% of respondents were from the Scheduled Caste (SC) or Other Backward Classes (OBC), who are among the targeted beneficiaries of this scheme. Around 32% of respondents were illiterate and over a quarter of the respondents lived in poor housing conditions.

Objective 1: Awareness of AB-PMJAY and its features

As shown in Table 2, 55.4% (n=72) were aware that AB-PMJAY was a national scheme. Most people got information about the scheme from family or friends (19.2%), followed by TV/radio (15.4%) and community health workers (ASHA/ANM -13.8%). Very few people got information from Panchayat (village administration) based outreach (4.6%) or the internet (2.3%).

Of the 72 people who had heard about the scheme not many knew about its features (Figure 1). Only 25% knew that pre-existing conditions were covered, while 20.8% knew about cashless and paperless processing. Fewer people (9.7%) knew about the transport allowance, while only 8.3% knew about the Arogya Mitra (scheme facilitator within empaneled hospital). Nobody knew about the October 2024 Ayushman Vay Vandana initiative (inclusion of elderly) or the March 2024 expansion of eligibility to include the grassroots healthcare workers (ASHA, Anganwadi workers) and their families. This lack of awareness about eligibility was also encountered among the community health workers themselves.

Figure 1: Knowledge of scheme features among those aware of the national scheme (N=72) (BPL= Below Poverty Line; Arogya Mitra = Healthcare friend/facilitator)

Logistic regression analysis revealed that education of the HoF (AOR = 9.20; 95% CI: 3.41–24.82; p < 0.001) and SC caste status (AOR = 0.41; 95% CI: 0.19–0.89; p = 0.024) were independently significant predictors of PMJAY awareness after controlling for confounders (R² = 0.384).

Objective 2: Utilisation patterns and barriers

The scheme has been implemented for over seven years now but only 47.7% of the people who are eligible received their Ayushman golden card (Figure 2). Among the respondents, 62 (47%) went to empanelled hospitals and 68 (53%) went to non-empanelled private hospitals. Only 15.4% of the respondents were able to utilise the scheme, while 45% were refused treatment even among the empaneled hospitals. One person was turned away from a hospital on the pretext that the hospital had already helped too many people that month. Another person had to pay all the money OOP because the person who was supposed to help them in the hospital (Arogya Mitra) was not there. A third person was asked to pay some money and then get it back later but that never happened. None of these respondents knew about the grievance redressal mechanism.

Figure 2: Utilisation patterns and barriers to utilisation

The most common reason for non-utilisation of the scheme was ignorance about empanelled hospitals (48.5% of non-users), followed by the golden card (beneficiary identity card) not yet having been issued at the time of hospitalisation (22.1%), prior negative experiences with empaneled hospitals (10.3%), and distance to facility (7.4%). Logistic regression analysis found card possession (AOR = 14.7; 95% CI: 4.1–52.6; p < 0.001) and awareness of empanelled hospitals (AOR = 6.3; 95% CI: 2.2–18.1; p = 0.001) were the strongest independent predictors of utilisation (R² = 0.412).

As seen from Table 3, although all respondents (130) were eligible for the scheme more than half (52%) did not seek care in empanelled hospital. Only 20 (32%) among the 62 respondents who had the beneficiary identity card (golden card) to access the scheme benefits, used it.

Objective 3: Financial risk protection and catastrophic health expenditure

The financial risk protection among those who used the scheme is shown in Table 4. Out of 20 people who successfully used the scheme, the average direct out-of-pocket expenses (OOPE) was Rs. 581.9 per episode which was much lower than those who did not avail the scheme (mean direct OOPE Rs. 13,355.8). However, among those who successfully used the scheme, indirect costs averaged Rs. 2,650 per episode (more than four times the direct costs). These indirect costs were mainly for transport, food and lodging when they went to district or urban empaneled hospitals. No one got the transport allowance or reimbursement for follow-up after hospitalisation that the scheme promised. As a result, 75.0% of PMJAY users had CHE when indirect costs were included (Table 4).

Among 68 non-empaneled hospital users, 95.6% had CHE on combined costs. 30% of them had OOPE that was more, than 100% of their monthly household consumption expenditure. In this group, 20 households (29.4%) had to borrow money or sell assets to pay for care. Among households that borrowed, 55.0% sold jewellery or physical assets. 25.0% borrowed from family or friends and 20.0% took loans from moneylenders. 50.0% of borrowing households remained in debt at the time of data collection. Logistic regression identified non-utilisation of the scheme (AOR = 38.4; 95% CI: 9.2–160.7; p < 0.001) and illiteracy (AOR = 4.1; 95% CI: 1.4–12.2; p = 0.010) as independent predictors of total CHE (R² = 0.468).

Qualitative findings from in-depth interviews

The following thematic issues emerged from the analysis of IDI’s.

Treatment without dignity and informal payments: Some people said they were treated poorly in empaneled private hospitals with staff asking for extra money for better care.

Non-Transparency in expenditure information during hospitalisation: None of the respondents received a written summary or a detailed bill when they were discharged from the hospital. To know how much money is left on their scheme card, people need to use the internet which is hard for those who do not have internet access or are digitally illiterate.

Minimal Role of Arogya Mitras (Scheme facilitators in hospitals): Most people said the facilitators were not helpful or did not exist. When someone had an emergency, the hospital asked for a monetary deposit, which was promised to be paid back but never returned.

Exclusion of terminal conditions: Two people said their family members with serious diseases like kidney disease and blood cancer were not allowed to get treatment under the scheme, although the scheme includes these conditions.

Non-awareness of scheme expansion: None of the respondents knew about the scheme’s expansion to include elderly and community healthcare workers.

Discussion

This study analyses the implementation of the AB-PMJAY scheme in one of the poorest districts of India. The study shows a gap between policy on paper and field level implementation. Poor awareness of the scheme, non-utilisation of the scheme (even among those aware) and high indirect costs negatively affect the intended financial risk protection of this scheme. 55.4% of respondents in Nawada know about PMJAY, which is slightly higher than what a previous study[8] found (49.3%) in rural areas of Bihar. This shows that more people know about PMJAY now than they did in the past. Family or friends were the most common source of information about the scheme and not the community health worker. This reveals a gap in the health system implementation of this scheme because community health workers are supposed to be spreading awareness about this scheme. Moreover, many of the community health workers were themselves unaware of their eligibility for this scheme. Similar to this study, other studies[9-11] have found low education, lower caste and rural dwelling, as significant factors negatively affecting awareness and knowledge about the scheme benefits. Although the scheme was expanded in March and later in October 2024, with new awareness initiatives, it failed to reach the rural poor in Nawada district.

The utilisation rate of the scheme among those with the card at 32% is very low and reflects the poor state of implementation in the entire state of Bihar, which has a very low card-to-admission ratio (2.2 percent). This is very different from states like Tamil Nadu and Karnataka, where the utilisation rate is much higher. Three reasons for this low utilisation rate emerged from this study: lack of awareness regarding empaneled hospitals; long waiting time for getting the AB-PMJAY card; and empanelled hospitals refusal to help. The high refusal rate of hospitals (45%) reflects a lack of regulatory oversight and compliance monitoring over the private healthcare sector, resulting in ‘risky insurance’[12]. People have health insurance but the hospitals do not follow the rules. Empanelled hospitals refusal to hospitalise eligible beneficiaries of the scheme is not limited to Bihar state and has also been reported in a pan-India study[13]. The National Health Authority has a system to help people with their grievances but low awareness affects its implementation.

The study showed that the scheme was effective in reducing direct OOPE, among those who utilised it correctly (15% of respondents). Studies done in other states[14] and independent reports[5] also provide evidence on the reduction in OOPE due to this scheme. However, the high indirect costs experienced among the scheme users is a major area of concern and has been reported[15,16] in studies from other states of India. Travel costs were significant contributors to the indirect costs and although the scheme provided for transport allowance, none of the beneficiaries utilising the scheme in this study received the transport allowance, which reflects another implementation gap.

The findings of this study can be interpreted using the theory of planned behavior[17], which says that that people are more likely to do something if they know about it. If people do not know about the scheme (as in this study) or if they have had bad experiences with it before (also a finding from this study) they are less likely to use it. Rogers[18], theory of diffusion of innovations suggests that new ideas and programs like AB-PMJAY are often adopted slowly by people who are marginalised as is the situation of the respondents of this study.

In terms of numbers, national data shows that AB-PMJAY scheme has distributed 435.2 million cards with 37,385 hospitals empaneled as part of the scheme. The OOPE for health has come down from 55.63% when the scheme started to the current value of 39.4%, which is a significant achievement. The AB-PMJAY 2.0 reforms, which include AI based fraud detection mechanisms, rationalized health benefit package, upgraded Hospital Engagement Module and creating a unified 14-digit health id for national portability are all steps in the right direction. However, as shown in this study the effectiveness of this scheme lies in overcoming last mile implementation hurdles.

Conclusions and Recommendations

The implementation issues identified in this study can be framed into four dimensions of concern (4 A’s). These 4A’s also serve as the four actionable areas for recommendations.

Dimension 1: Awareness

Only 55% of eligible beneficiaries in this study were aware of the scheme and none had information about all the benefits of the scheme. Also, among those enrolled, due to ignorance, less than half [62 (48%)] went to empanelled hospitals and among these only 20 (32%) actually utilised the card to access the scheme. Hence, enrolment in the scheme or having the beneficiary identity card does not in itself, equip beneficiaries with the operational knowledge required for accessing the scheme (knowledge about empanelled hospitals, eligible services, benefits and rights).

Dimension 2: Accessibility

In a rural district like Nawada, the distribution of empanelled hospitals is heavily skewed toward the district headquarters and larger towns, leaving residents of more distant rural villages without a reasonably geographically accessible empanelled facility. This study found that among those who successfully used the scheme, indirect costs averaged four times the direct costs. These indirect costs were mainly for transport, food and lodging when they went to district or urban empaneled hospitals. Hence, the effectiveness of the financial risk protection was negatively affected.

Dimension 3: Administrative

Non-issuance of e-card among eligible beneficiaries, non-reimbursement of transport expenses, absent or non-helpful scheme facilitator in the empanelled hospital, refusal of hospital to admit eligible beneficiary, and informal payments reflect poor administrative, regulatory and governance mechanism.

Dimension 4: Affordability

The study showed that direct OOP payments were significantly lower for those who utilised the scheme as compared to those who did not utilise it. However, 75% of the scheme beneficiaries experienced CHE due to indirect costs suggesting that the schemes core objective of financial risk protection was negatively affected. Hence, the beneficiaries faced a paradoxical situation of experiencing CHE to access a scheme aimed at preventing OOPE.

Based on the 4 A’s identified, the study recommends a strong offline community awareness campaign by using existing community health workers/local panchayat to address the awareness concern. Development of rural healthcare infrastructure should complement expansion of the scheme as geographical location of empanelled hospital is a significant contributor to indirect costs in accessing the scheme. Strong monitoring of empanelled hospitals is required, while ensuring the presence of a trained scheme facilitator in all empanelled hospitals to improve the administrative bottle necks in implementation. Locally based grievance redressal mechanisms need to be made functional to help patients in their time of need. All the above steps, would help address the affordability paradox in the scheme’s implementation.

Study limitations

The study is limited to the beneficiary perspective and does not include the provider perspective on this issue. Since the recall period regarding hospitalisation experience was one year, it is subject to recall bias/error.

References

1. World Bank. World development indicators: Health expenditure data. World Bank DataBank; 2025. Available from: https://databank.worldbank.org
2. Ministry of Health and Family Welfare, Government of India. Health benefit package under PM-JAY: Revised 2025 list of 1,961 procedures. National Health Authority; 2025. Available from: https://pmjay.gov.in
3. Bhatnagar A, Chhabra P, Singh S, Banerjee S. Coverage and utilization of Pradhan Mantri Jan Arogya Yojana (PM-JAY) across Indian states: Findings from the National Family Health Survey-5. Lancet Reg Health Southeast Asia. 2022;3:100017. doi: 10.1016/j.lansea.2022.100017
4. Medical Buyer. Bihar PM-JAY: State-wise hospitalization and card issuance data, October 2024. Medical Buyer India; 2024. Available from: https://www.medicalbuyerindia.com
5. Drishti IAS. Ayushman Bharat Pradhan Mantri Jan Arogya Yojana (AB-PM-JAY); 2025. Available from: https://www.drishtiias.com/to-the-points/paper2/ayushman-bharat-pradhan-mantri-jan-arogya-yojana
6. Wagstaff A, van Doorslaer E. Catastrophe and impoverishment in paying for health care: With applications to Vietnam 1993–1998. Health Econ. 2003;12(11):921–934. doi: 10.1002/hec.776
7. Braun V, Clarke V. Using thematic analysis in psychology. Qual Res Psychol. 2006;3(2):77–101. doi: 10.1191/1478088706qp063oa
8. Prasad V, Singh S, Kumar P. PMJAY awareness and utilization in rural Bihar: A cross-sectional study. J Family Med Primary Care. 2023;12(5):987–993.
9. Thomas A, Murthy S, Bhattacharya A. Determinants of PMJAY awareness, enrolment, and utilization: Evidence from six Indian states. Health Policy Plan. 2023;38(6):645–657. doi: 10.1093/heapol/czad034
10. Patel R, Bhimani D. Awareness of Ayushman Bharat PM-JAY among eligible beneficiaries in Gujarat. Indian J Public Health. 2025;69(1):78–84.
11. Dixit R, Chauhan A, Juneja K. Awareness and Utilization of Ayushman Bharat Pradhan Mantri Jan Arogya Yojana (AB-NHPM) among Beneficiaries of Gautam Buddha Nagar District: A Comparative Study. Indian J Community Med. 2025;50(1):213-219. doi: 10.4103/ijcm.ijcm_728_23
12. Agnes N. Risky insurance: The Pradhan Mantri Jan Arogya Yojana in Jharkhand. Econ Polit Wkly. 2020;55(45). Available from: https://www.epw.in/journal/2020/45/reports-states/risky-insurance.html
13. Trivedi M, Patel S, Sharma R. User experiences and access barriers under PMJAY: A national qualitative study. PLOS ONE. 2022;17(4):e0266732. doi: 10.1371/journal.pone.0266732
14. Kumar R, Sharma P, Gupta M. Reduction in catastrophic health expenditure under AB-PMJAY at a tertiary hospital in Himachal Pradesh. Indian J Community Med. 2024;49(2):234–239.
15. Garg S, Tripathy JP, Roy A, Bhattacharyya K. Four-year impact evaluation of Ayushman Bharat PMJAY in Chhattisgarh. BMC Health Serv Res. 2024;24(1):312. doi: 10.1186/s12913-024-10812-y
16. Parmar D, Williams G, Dkhimi F, Ndiaye A, Asante FA, Arhinful DK, Mladovsky P. Enrolment of older people in social health protection programs in West Africa: Does social exclusion play a part? Health Syst Reform. 2023;9(1):e2198040. doi: 10.1080/23288604.2023.2198040
17. Ajzen I. The theory of planned behavior. Organ Behav Hum Decis Process. 1991;50(2):179–211. doi: 10.1016/0749-5978(91)90020-T
18. Rogers EM. Diffusion of innovations. Free Press. 1962.