It’s the 28th at a 150-bed hospital in a Tier-2 city near Pune. The billing desk has a queue at the counter, three cashless files open on one screen, and a WhatsApp thread with a TPA co-ordinator who last replied at 11:40am. Two discharges are stuck because the pre-auth came back with a query. The revenue for the month is real, treated, documented somewhere. It just hasn’t turned into money yet.
That gap, between care delivered and cash collected, is the revenue cycle. In an Indian hospital it runs through more hands than most vendors admit, and the weakest link is almost never the billing engine.
Key takeaways
- The Indian revenue cycle has a cashless and a scheme layer on top of ordinary billing, so a claim passes through the TPA, sometimes the insurer, and sometimes a state health agency before it settles.
- IRDAI’s 29 May 2024 Master Circular puts a one-hour clock on cashless authorisation and a three-hour clock on discharge authorisation, and both depend on a pre-auth that’s complete on first read.
- NHCX is standardising the claim message across insurers and TPAs, but as of the July 2024 NHA update it was still in rollout, not a universal mandate.
- AB-PMJAY covers ₹5 lakh per family a year across 36,229 empanelled hospitals; scheme billing runs on its own tariff master and portal, separate from private cashless.
- Most denials trace back to documentation, not the billing software. That’s the part a hospital can fix first.
What the revenue cycle actually covers
Think of it as one chain with several handoffs, each of which can leak money.
A patient registers and gets a UHID. The front desk gives an estimate. If it’s a cashless case, someone raises a pre-authorisation with the TPA and waits for approval before the surgery is scheduled. Services get delivered and captured against a tariff. The bill is assembled with GST where it applies. The claim goes out. It either clears, gets queried, or gets partly disallowed. Collections chase whatever the insurer, the scheme, or the patient still owes. Then the finance head reconciles all of it at month-end and finds the shortfall.
Nine steps, and only two or three of them are things the billing module does well on its own. The rest is coordination between the clinical side, the insurance desk, and the counter.
Where an Indian hospital’s cycle gets complicated
A cash-paying patient is simple. You treat, you bill, they pay at discharge. That’s a small and shrinking share of most hospitals’ revenue.
The volume sits in cashless and scheme cases, and each one splits the cycle into a parallel track that runs on someone else’s rules.
The TPA and cashless track
A Third Party Administrator runs the cashless desk on the insurer’s behalf. It processes the pre-authorisation, assesses the claim, and settles on the insurer’s account. Worth being precise about one thing: a TPA can flag a claim as not admissible, but it can’t deny it. Only the insurer can repudiate a claim, and it has to state reasons against the policy terms (IRDAI Policyholder portal). So a cashless rejection that lands back through the TPA traces to an insurer decision. The TPA and cashless glossary walks through that chain if your desk staff are new to it.
The pre-auth is where the money either flows or stalls. It’s built from the treating doctor’s diagnosis, indication, and plan. If that documentation is complete and consistent, the request is approvable on first read. If it’s a phrase in a text box, the TPA sends a query, and the whole thing waits.
The IRDAI clock
The 29 May 2024 IRDAI Master Circular put hard timelines on this. The insurer must decide on a cashless authorisation request within one hour, and grant final discharge authorisation within three hours of the hospital’s discharge request (IRDAI Master Circular, 29 May 2024). If final authorisation is delayed past three hours, the extra hospital charges from that delay are borne by the insurer.
Here’s the operational catch. The clock is only useful to the hospital if the pre-auth is clean. A query resets the practical timeline and holds the bed. Discharges pile up in the afternoon because the documentation to close them arrived late. The IRDAI cashless timelines breakdown covers exactly where a note sits on that clock.
NHCX and the claims rail
For years the billing desk kept a mental map of which insurer wants which portal and which TPA still takes a PDF over email. The National Health Claims Exchange is aimed at that mess. NHCX is a single digital channel, built by the National Health Authority under ABDM and shaped with IRDAI, for hospitals, insurers, and TPAs to pass claim messages in one FHIR-based format.
Where does it actually stand? As of the July 2024 NHA update, 34 insurers and TPAs were live on the exchange and roughly 300 hospitals were ramping up (Press Information Bureau, 21 July 2024). That’s active rollout, not a mandate. Most hospitals aren’t transacting on it today. There’s a financial nudge attached: for an AB PM-JAY claim filed in FHIR format through NHCX and linked to an ABHA address, the Digital Health Incentive Scheme pays the health facility ₹200 per claim or 10% of the claim amount, whichever is lower (NHA / ABDM, Digital Health Incentive Scheme). The full mechanics live in the NHCX for clinics explainer.
One thing NHCX doesn’t change: the IRDAI rules and the documentation. It moves the message faster. It doesn’t write the note.
The PMJAY and scheme track
Then there’s the scheme layer, which runs on its own tariff and its own portal. AB-PMJAY provides ₹5 lakh per family a year for secondary and tertiary care, across 36,229 empanelled hospitals nationwide, covering 1,961 procedures across 27 specialties (Press Information Bureau, 17 March 2026). CGHS and state schemes add their own package rates and pre-authorisation flows.
Scheme billing isn’t cashless-with-a-different-logo. The package rate is fixed, the documentation requirements are specific to the procedure, and a mismatch between the coded procedure and the operative note is a straight rejection. Hospitals that run a heavy PMJAY load usually have a separate desk for it, because the workflow shares almost nothing with private cashless beyond the word “cashless.”
What software should own, and what it can’t
A good hospital RCM setup handles the deterministic work without drama:
- Registration, UHID, and the patient estimate at the front desk.
- Tariff and package pricing, including scheme masters and TPA rate lists.
- GST-correct invoicing and receipts.
- Pre-auth request generation and a claim-tracking status board.
- A denial worklist so nothing ages past the appeal window unseen.
That list is mature in most Indian hospital systems. If you want the module-by-module view of where each of these lives, the hospital management system breakdown maps it.
What software can’t manufacture is the clinical content the claim is built from. The billing team codes what the note supports. When the operative note says “as above” and the discharge summary is three lines a resident typed at 11pm, the coder guesses, downcodes, or leaves the field blank. That’s the root of a large share of denials, and no billing engine fixes it, because the problem starts in the consulting room and the OT, not the finance office.
I’ll say it plainly: most hospitals buy more RCM software when the actual bottleneck is documentation. The new module changes the plumbing, not the note that feeds it.
Where the denials really start
Trace a denied cashless claim backwards and it usually lands in one of four places.
The diagnosis on the pre-auth doesn’t match the final coded diagnosis. The indication for surgery isn’t documented in a way the TPA’s medical officer accepts. The discharge summary is missing the continuity detail a scheme auditor needs. Or the procedure code and the operative note simply disagree. Each of those is a documentation defect wearing a billing costume.
The finance head sees it as a rejection rate. The billing manager sees it as rework. The reason sits upstream, in how the encounter got written down under time pressure, in the two minutes a doctor had between one patient and the next.
Where Hospital Copilot fits
Documentation and claim submission are different problems, and Patient Square works on the first one only.
Hospital Copilot and Practice Copilot are the products here, with an ambient documentation module inside them. During the consult or the ward round, the module listens and drafts a structured note, ICD-10 code suggestions, and a prescription draft, ready to review and sign a couple of minutes after the encounter. That’s the raw material the pre-auth desk builds on.
For the revenue cycle, that means the coder is reading a documented encounter instead of interpreting a terse string. The indication, the diagnosis, and the plan are present and consistent, so the pre-auth goes out approvable on first read more often, and fewer of them bounce back as queries against the IRDAI clock.
A few design choices matter for an Indian hospital. Ward rounds and OPD consults come in braided Hindi and English, so the module takes code-mixed speech on input and returns the note in clean clinical English, which is what the medico-legal record and the claim need. The ICD-10 output is a suggestion the clinician confirms, not an auto-code. The prescription is a draft the doctor reviews and signs. And visit audio is processed in memory and discarded once the note drafts, so there’s no recording sitting on a server.
To be straight about the boundary: this is not a claims tool. Hospital Copilot does not connect to NHCX, any insurer, or any TPA, and it does not file pre-authorisations. ABDM and NHCX integration are on our roadmap, not shipped, and we won’t wear an ABDM badge we don’t hold. A hospital that needs live NHCX transacting today should pick a claims platform that has it and use a documentation layer alongside.
Fix the cycle in the order that pays
If your rejection rate is climbing, the instinct is to look at the billing software or the insurance desk. Look at the note first. It’s cheaper to fix and it’s usually the real leak.
Run the trace on your last 50 denied cashless claims. Sort them by root cause. If more than a third come back to thin or late documentation, a new RCM module won’t move your numbers, and a documentation layer might. That 150-bed hospital on the 28th doesn’t have a billing-engine problem. It has two discharges stuck behind a query on a pre-auth that went out incomplete.
Book a short demo to see what an ambient consult produces as a note, and how much cleaner a pre-auth reads when the encounter behind it is fully written down. Or start with the hospital management system map if you’re still working out which module actually owns which part of your cycle.