In Uttar Pradesh, a dialysis claim under PM-JAY took about 69 days to clear last year, according to the CEO of the state’s health agency. After the state switched on auto-adjudication for haemodialysis, she said, the same claim now clears in about five days [1].
That change, reported at Aarogya Manthan 2026 in New Delhi on 25 September, captures the direction of the day. Across sessions marking eight years of PM-JAY and five years of ABDM, regulators, state agencies, insurers and technology firms described claims moving from paperwork that is checked to data that is trusted.
This analysis draws on three official NHA livestreams: the inaugural session, the PM-JAY track and the ABDM track [1][2][3]. Figures attributed to speakers are as stated on stage and have not been independently audited.
At a glance
- Auto-adjudication is now policy. NHA released a framework for PM-JAY, and speakers put the share of claims already clearing without manual review at about 30% to 35% [1][2][3].
- NHCX is in its second phase. NHA says the exchange carries close to 1 lakh claims a day. The next step is private-market adoption and structured claims in place of PDFs [2].
- IRDAI previewed three building blocks for retail health insurance: ABHA on every policy, a treatment code on every bill, and a standard billing format [2][3].
- Insurers are pairing AI with human review. One insurer said every rejection still goes to a person, and the industry’s claims repository put real-time data quality at about 60% [1].
- Purchasing is shifting from volume to value, through tumour-board certification for oncology packages, a national cost-effectiveness threshold, and outcome-linked incentives under discussion [1].
- Claims could become financeable. A proposal to discount approved NHCX claims, modelled on the TReDS invoice platform, aims to ease hospital working capital [2].
Policy updates: what changed on 25 September
The Union Health Minister launched five initiatives during the inaugural session [3][4][5].
| Release | What it does |
|---|---|
| Framework for Auto-Adjudication of Claims under AB PM-JAY | Sets out the vision, governance and roadmap for AI-enabled auto-adjudication [3] |
| NHCX Implementation Guide for Payers | Covers claim-processing workflows and system requirements for payers and their implementation agencies [3] |
| Secure Data Environment (SDE), built with IISc Bengaluru | A governed workspace where approved entities work with de-identified data or run code in an isolated NHA environment, with outputs reviewed before release [5] |
| National Healthcare Providers Registry (NHPR) mobile app | Lets professionals and facilities register from a phone, with GPS-tagged facility photos and verification by authorised verifiers in the same app [3] |
| Guidebook for a Model Digital Health Facility | A step-by-step guide for digitising each stage of a hospital visit, from registration to discharge [6] |
The NHA CEO said more than 13 crore PM-JAY records will be opened for research and analytics through the SDE, and that NHA plans to extend auto-adjudication to high-value, complex claims [3]. Several announcements from the stage matter as much as the documents.
IRDAI’s three building blocks. The IRDAI Chairman said the regulator would issue a public consultation paper in about a month, proposing that insurers seed ABHA on every new and renewed policy, carry a treatment or procedure code on every bill, and use a standard billing format [2]. He set a 12-month horizon for visible progress on NHCX adoption.
In mid-October, the Chairman announced plans for a consultation paper on healthcare billing and claims, with a scope that also covers treatment costs, provider behaviour and digital public infrastructure. Reports described it as still planned, and it is not yet clear whether this is the paper previewed at Aarogya Manthan [7].
Clinical standards tied to payment. The Secretary of the Department of Health Research said ICMR has given NHA more than 150 Standard Treatment Workflows, and that linking reimbursement to standard treatment gives guidelines practical weight [3]. NHA’s Deputy Secretary for health policy announced that the NHA CEO has approved a mandatory multidisciplinary tumour board certificate for all oncology packages, to be enforced through IT, with doctors verified on HPR and facilities on HFR [1].
A national cost-effectiveness threshold. Prof. Shankar Prinja of PGIMER Chandigarh said India’s threshold, with possible modifiers for disease severity, transformative technologies and made-in-India products, was expected within weeks [1]. The willingness-to-pay study behind it, drawing on 21,840 observations across six states, is published in Value in Health [8].
Onboarding and industry data. NHA said a revamped NHCX sandbox would follow within about a week, expected to cut integration time from months to days [2]. The Insurance Information Bureau of India (IIB) said 51 new abuse triggers would join its existing 32 fraud and waste triggers, and that it is in talks with NHA to link the HFR ID with IIB’s ROHINI hospital ID [1].
IRDAI’s consultation on a Public Insurance Registry, which would let people compare policies and claims performance, closed on 30 September [9]. The Chairman has since said a beta version is being considered for January 2028 [10].
PM-JAY: auto-adjudication moves from pilot to policy
NHA started with a handful of low-value, high-volume packages, and a hackathon earlier in 2026 helped show the approach worked. The NHA CEO put the share of auto-adjudicated PM-JAY claims at about 30%, and the ABDM Mission Director cited about 35% on the same day [1][2][3].
The CEO was specific about the conditions for scaling. Results must be explainable, auditable and accountable. A share of machine decisions must be audited to confirm consistent performance, and trust has to be earned continuously, not once [1].
What the states reported
Uttar Pradesh handles 10,000 to 15,000 claims a day. Its agency CEO said daily volumes had grown beyond what manual review could absorb consistently, so the state trained hospitals first and then switched the engine on [1]. She said queries have fallen month on month since.
She added that some well-regarded hospitals saw claims rejected where their documentation did not yet match standard treatment guidelines, and that those hospitals are now asking for training [1].
The CEO of a state that recently joined PM-JAY reported about 6,000 claims a day, of which about 1,500 are dialysis and 300 chemotherapy [1]. He described a rule that checks earlier creatinine reports before approving dialysis, and called for a gradual rollout with continuous training for hospitals, TPAs and insurers.
The payment side is next
The panel moderator noted that clinical adjudication in PM-JAY is now automated, while the payment workflow that follows it has not yet changed [1]. The full benefit depends on digitising both.
The Chief Controller of Accounts at the Health Ministry, who oversees CGHS payments through NHA’s Transaction Management System, listed what scaling would need on that side [1]:
- Financial rules (General Financial Rules, CGHS rates and beneficiary categories) encoded in the engine
- Parallel runs that compare machine and manual decisions on samples
- A maker-checker principle, including a preparer and checker inside hospitals before submission
- An audit trail extended to cover every approved transaction
The NHA CEO added that engines today mostly extract data from scanned documents, and that native digital records from providers would make them far more efficient [1]. For hospitals, the practical point is that documentation is now read by a machine before a person.
The insurance market: NHCX enters its second phase
The IRDAI Chairman described the market in three bands. Market-based insurance reaches perhaps the top 10% of the population, and government schemes cover the bottom half. A “missing middle” of about 40% sits between them, and he asked NITI Aayog to help design an affordable product for that group [3].
NHCX is the shared rail meant to serve all three. According to the ABDM Mission Director, PM-JAY runs every claim through it, alongside 10 central and 56 state government schemes, and the exchange handles close to 1 lakh claims a day [2]. The IRDAI Chairman said NHCX already handles PM-JAY at scale, with more than 8 crore claims paid and 15 crore pre-authorisations processed [2][3]. The Health Minister separately said PM-JAY has facilitated more than 13 crore hospital admissions to date [3][11].
NHA said most insurers have integrated on the payer side, and that private-market claims, especially from small and mid-sized hospitals, are the next stage [2].
What each participant needs from the exchange
The Chairman argued that NHCX works only if every participant gets value from it [2].
| Participant | What NHCX has to deliver |
|---|---|
| Policyholders | Time-bound pre-authorisation (30, 60 or 90 minutes) and claims settled promptly and consistently |
| Hospitals | Less uncertainty at admission and discharge, faster settlement, and a workable path for small district hospitals |
| Insurers | Consistent, comparable claims that follow the contract |
| Technology providers | A viable return on the cost of connecting |
The provider view
The managing director of Artemis Hospital, who co-chairs the IRDAI–CII working group on NHCX, described the friction hospitals see today: a different form for each insurer, policy terms not visible at admission, back-and-forth over pre-authorisation amounts, and discharge waits of four to six hours while additional documents are requested [2]. She said hospitals and insurers in the working group have agreed two targets: pre-authorisation in under 30 minutes and discharge within one hour.
She described hospitals moving “from coercion to curiosity”, with about 15 payers and more than 11 hospitals ready [2]. Artemis built 22 APIs and found testing took longer than expected, so she asked for a readiness checklist that would make onboarding possible in about 48 hours. The ABDM Mission Director said the payer implementation guide answers that request.
How insurers are using AI
ICICI Lombard’s CEO said the industry paid about ₹1 lakh crore in health claims last year. By his estimate, 10% to 15% of that was fraud, waste and abuse [2]. He said his company auto-adjudicates about 70% of claims with AI, and called NHCX a potential “UPI moment” for claims, since bilateral integrations with tens of thousands of hospitals cannot match a shared exchange.
Niva Bupa described a layered approach: OCR to digitise documents, then separate tools for financial adjudication, medical admissibility and fraud detection, with rules drawn from NHA, ICMR and WHO guidelines [1]. The company’s representative said about 40% of claims, those without flags or queries, receive pre-authorisation in about 30 minutes, and that reimbursement follows within five days.
By the same account, every rejection goes to a person, about 20% of AI decisions pass a second human check, and a further 10% are audited after payment [1].
IIB, which holds the industry’s policy and claims data, put real-time data quality from the 32 health insurers at about 60% [1]. IIB’s representative said some gaps sit with insurers and others with coding and documentation standards, and that five IRDAI working groups now bring hospitals and insurers together to close them.
Turning approved claims into working capital
M1xchange, an RBI-licensed TReDS platform that lets suppliers raise finance against approved invoices, proposed a new use for NHCX. Its CEO said the platform handles about ₹5 lakh crore a year, with bank competition bringing discounting rates to about 6% [2]. He suggested treating an NHCX-approved claim as a receivable a bank can finance, and estimated that hospitals’ working-capital costs could fall by about half.
Gujarat’s Additional Chief Secretary for Health said the idea addresses provider liquidity, provided adjudication is predictable, claims are visible and fraud control is strong [2]. The ABDM Mission Director explained that the payer would pay the financier, so the hospital receives payment at around T+1 at a small discount instead of borrowing at 12% to 13%.
The IRDAI Chairman called claim discounting a “next-order use case” and stressed that its cost must never pass to the policyholder [2]. NHA plans workshops for states. His advice on adoption was that certainty, not extra payment, should be the main incentive, with insurers publishing turnaround standards and larger hospitals with existing HMIS and EMR systems going first.
Market trend: from purchasing volume to purchasing value
The second PM-JAY panel asked whether a scheme covering about 60 crore people is getting value for money. NHA’s moderator used the WHO definition of strategic purchasing, which links payment to quality and population need [1]. Four models from the panel show what that can look like.
| Model | Mechanism | Result reported on stage |
|---|---|---|
| Tamil Nadu (CMCHIS converged with PM-JAY) | Cover for all 1.45 crore families, tenders among public-sector insurers, and an incurred claims ratio band of 90% to 110%, above which the state pays half the excess | Premium of ₹849 per family per year; 54% of claim value goes to government facilities [1] |
| National Cancer Grid pooled procurement | “Quality first, cost next” scoring on about 25 parameters; central price discovery, with each member contracting directly | Average price reduction of 82% from MRP, with no drug substitution [1] |
| Indonesia’s BPJS Kesehatan | Performance-based capitation in primary care (contact rate, referral rate, chronic cases referred back) and DRG-style hospital payment | A primary-care referral rate above about 15% is considered poor performance [1] |
| PM-JAY next steps (proposed) | Keep case-based bundles, add outcome-linked incentives using ABDM and transaction data, and adjust for case mix | Under discussion [1] |
Prof. Prinja cautioned that paying purely for outcomes would need heavy measurement, since outcomes depend on factors beyond care [1]. He favoured simple, low-burden metrics that are hard to game, layered on top of existing bundled payments.
A health-financing specialist on the panel described the next role for claims data. Beyond confirming that a claim is genuine, adjudication can check whether treatment matched symptoms, standard guidelines and the provider’s capability, and eventually outcomes [1].
The Secretary of the Department of Pharmaceuticals named his single priority as using IT to analyse hospital performance and track outcomes such as repeat procedures, transfers and mortality, which he sees as the next layer to monitor [1]. He also set out three ways to finance high-cost cancer therapies beyond the ₹5 lakh cover: full public funding, co-payment, or the family paying at the scheme’s negotiated rate. For providers, the direction is that claims data will increasingly describe the quality of care, not only its cost.
The data layer underneath: standards and structured records
Each of these trends depends on claims a machine can read. The founder of Plus91, a hospital software firm, noted that insurers pay for treatments and outcomes, and those live in the medical record rather than the bill [2].
The standards to support that are published and usable. NRCeS at C-DAC Pune described a deliberately small set: FHIR for structure, and SNOMED CT, LOINC and ICD for meaning [2]. Recent additions include:
- Common Drug Codes for India (CDCI), built partly from a drug database donated by Tata 1mg
- About 1,500 LOINC-based lab codes, set at parameter level after participating labs found 18 different CBC panel variants
- Bharat Health Terminology Services, a single bundle with search, a SNOMED browser and FHIR API servers
The AIIMS Director illustrated the need with a familiar example: death summaries that record “cardiopulmonary arrest” rather than the underlying condition [2]. Shared vocabulary makes records comparable across hospitals, and therefore easier for payers to assess.
State experience shows the work in progress. Uttar Pradesh’s Health Secretary said most of the state’s systems are ABDM-compliant and the next step is connecting them, so that a person’s journey from an ASHA visit to a district hospital forms one continuous record [2]. Tripura’s representative said the state onboarded 85% of its public facilities onto an M3-compliant hospital system in four months by delivering software, hardware, instructions and training together [2].
What this asks of hospital revenue teams
For a hospital CFO or revenue cycle head, the forum pointed to one shift: the claim is becoming a structured, coded record that a machine reads first and a person reviews second. Six practical changes follow from what speakers described.
- Documentation now sets the pace of payment. Uttar Pradesh’s account suggests that engines apply standard treatment guidelines consistently, so documentation gaps surface as queries [1].
- Validation is moving before submission. The director of Akhil Systems, a hospital software company, described practice in Saudi Arabia and Dubai: eligibility checked at registration, coverage and coding alerts during the consultation, and a coding check before the claim leaves [1]. He said AI-based pre-submission checks cut rejections by about 70% at one private institution, and asked NHA to offer providers a validation engine with payer rules built in.
- Coding is moving to the point of care. IRDAI’s proposed treatment code on every bill and the national terminology services both point towards coding during care rather than afterwards [2].
- ABHA at admission is likely to become routine if IRDAI’s proposal goes ahead. Seeded on a policy, it lets a hospital see entitlements such as room category at admission [2].
- Structured claims are replacing PDFs on NHCX. The second phase is built around structured data, and the payer guide and revamped sandbox are meant to shorten the path to going live [2].
- Clean approvals could carry financial value. If claim discounting develops, a predictable, non-repudiable approval becomes a financeable asset, which favours claims that clear the first time [2].
Accuracy upstream now has a direct effect on settlement time downstream.
Where CaladriusHealth.AI’s platform fits
CaladriusHealth.AI builds revenue cycle software for hospitals that works directly on ABDM and NHCX, covering eligibility, pre-authorisation, claim submission, denial management and reconciliation. Several of the provider-side needs raised at Aarogya Manthan correspond to capabilities CaladriusHealth.AI describes in its platform product brief.
| Need raised at the forum | What the platform does, as described in its product brief |
|---|---|
| Check claims before submission [1] | A Claims Scorecard gives a pre-submission score across denial risk, payer-rule compliance, coding integrity, documentation completeness and tariff variance, with a suggested next action for each flag |
| Structured claims on NHCX, not PDFs [2] | Claim bundles are assembled from the encounter, with the facility’s HFR identifier, an HPR-verified care team and live payer policy discovery, and sent as NHCX ClaimBundles |
| Treatment codification [2] | SNOMED CT lookup and ICD-10 translation through the NRCeS Bharat Health Terminology Service, with fields left for a person where no confident code match exists |
| Explainable, auditable automation [1] | Each step can be inspected before submission, and model behaviour is tracked against a set of application and runtime evaluations |
CaladriusHealth.AI states on its website that its ABDM, NHCX and WASA milestones are in progress, not complete.
What to watch next
| Item | Status | Source |
|---|---|---|
| IRDAI consultation paper on ABHA, treatment codes and standard billing | Previewed for about a month after the event; a billing and claims paper was reported as still planned in mid-October | [2][7] |
| India’s cost-effectiveness threshold and modifiers | Expected within weeks of 25 September | [1][8] |
| Revamped NHCX sandbox | Announced for release within about a week of 25 September | [2] |
| IIB’s 51 abuse triggers | Announced for October | [1] |
| Public Insurance Registry | Consultation closed on 30 September; beta being considered for January 2028 | [9][10] |
| NHA workshops for states on claim discounting | Planned | [2] |
| Auto-adjudication of high-value, complex PM-JAY claims | Planned by NHA | [3] |
| NHCX adoption in the private market | IRDAI Chairman’s 12-month horizon | [2] |
Conclusion
Aarogya Manthan 2026 showed that NHCX is built and that attention has moved to the quality of the data flowing through it. Regulators, states and insurers described the same direction: structured, coded claims, automated where the rules are clear and reviewed by people where they are not. Hospitals that prepare their documentation and coding for that standard are positioned to benefit from the shorter cycles the forum described.
Sources
- National Health Authority, Aarogya Manthan 2026: PM-JAY session (livestream), AyushmanNHA YouTube channel, 25 September 2026. Accessed 26 October 2026.
- National Health Authority, Aarogya Manthan 2026: ABDM session (livestream), AyushmanNHA YouTube channel, 25 September 2026. Accessed 26 October 2026.
- National Health Authority, Aarogya Manthan 2026: Inaugural session (livestream), AyushmanNHA YouTube channel, 25 September 2026. Accessed 26 October 2026.
- DD India, Arogya Manthan 2026: J.P. Nadda highlights expansion of AB PM-JAY and digital health ecosystem, 25 September 2026. Accessed 26 October 2026.
- Digital Health News, Union Health Minister unveils series of digital health initiatives, September 2026. Accessed 26 October 2026.
- Digital Health News, NHA’s new guidebook maps out what a model digital health facility looks like, step by step, 2026. Accessed 26 October 2026.
- Newslaundry, Health insurance reform should cut costs, not shift them on to patients, 21 October 2026 (citing a Financial Express report). Accessed 26 October 2026.
- Chugh Y, et al., Estimating the willingness-to-pay per quality-adjusted life-year to aid health technology assessment in India, Value in Health, 2026. Accessed 26 October 2026.
- MediaNama, Lowdown: Inside IRDAI’s consultation paper on the Public Insurance Registry, September 2026. Accessed 26 October 2026.
- Business Standard, Irdai’s proposed distribution reforms aim for orderly transition: Ajay Seth, 15 October 2026. Accessed 26 October 2026.
- The Pioneer, PM-JAY facilitated over 13 crore hospital admissions, says Nadda, 26 September 2026. Accessed 26 October 2026.
Some external sources spell the event “Arogya Manthan”; this article follows the Health Ministry’s spelling, “Aarogya Manthan”. Source titles are reproduced as published.
CaladriusHealth.AI covers India’s health technology landscape with a focus on ABDM, NHCX, medical billing, and AI in healthcare.
Editorial disclosure: CaladriusHealth.AI builds revenue cycle software for hospitals that works on ABDM and NHCX, and has a commercial interest in the developments discussed in this article.
