Covered, not paid
Six out of ten Indian households now have health insurance. Last year those insurers were asked for ₹1,32,488 crore and paid out ₹94,248 crore. Of the ₹38,240 crore that did not follow, ₹8,307 crore was still being processed at year end and ₹29,933 crore was refused. Most of that refusal was never recorded as a rejected claim.
- Insurers settled 87.5% of health claims by count and 71.1% by value. Of ₹1,32,488 crore claimed in FY25, ₹94,248 crore was paid and ₹29,933 crore was refused. A further ₹8,307 crore was still in process at year end.
- ₹18,521 crore of that was disallowance, not rejection. Money trimmed off bills insurers had already agreed to settle. It carries a claim count of zero, which is why it has never shown up in the headline settlement rate.
- Roughly 29% of claim value is paid by the patient first and reimbursed later. The hospital books it as cash, the insurer books it as insurance, and both are right. Hospital self-pay share therefore overstates household burden.
- Incurred claims ratios run from 61% to 103% across firms selling the same product. Standalone health insurers sit between 61 and 75; public sector insurers between 96 and 103. No overlap.
- Government-sponsored schemes cover 42% of lives and generate 8% of premium, at ₹385 a life against ₹7,750 for individual policies. It is the only segment shrinking on both lives and premium.
- Coverage rose 19 points while the public share of institutional births fell and the caesarean rate rose to 27.2%. In private facilities it is 54.1%. Across all 21 states with data, the private rate exceeds the public rate everywhere.
Your insurer settled 87% of claims. It paid 71% of the money.
IRDAI reports that insurers settled about 87% of claims and repudiated about 8%. That is the number by claim count, and it is the number everyone quotes. Run the same table by rupee value and a different picture appears.
Insurers were asked to pay ₹1,32,488 crore. They paid ₹94,248 crore.
What to look for: the second block. It is bigger than the third. More money was trimmed off approved bills than was lost to outright rejection, and only the third block has a claim count attached to it.
View this chart as a table · 4 rows
| What happened to it | ₹ crore | Share of all claims |
|---|---|---|
| Claims paid | 94,248 | 71.14% |
| Disallowed under policy terms | 18,521 | 13.98% |
| Repudiated | 11,412 | 8.61% |
| Outstanding at year end | 8,307 | 6.27% |
Source: IRDAI Annual Report 2024-25, Table I.29. Total claimed value Rs 1,32,488 crore.
Two different numbers, and both are correct. Subtract paid from claimed and you get ₹38,240 crore. But ₹8,307 crore of that was simply not settled yet, which is normal at any year end. The amount actually refused is ₹29,933 crore, and ₹18,521 crore of that refusal happened inside claims the insurer had already approved.
That refused ₹29,933 crore splits two ways. Some claims were rejected outright. But the larger share, ₹18,521 crore, was quietly trimmed off bills the insurer had already agreed to settle.
Repudiation carries a claim count. Disallowance carries a count of zero against ₹18,521 crore, because it is not rejection of whole claims. It is deduction from claims recorded as paid.
That distinction matters more than anything else in the report. ₹18,521 crore was cut out of bills that insurers agreed to settle. It is the gap between what hospitals billed and what insurers accepted, and until this table it was not a published number at all.
In 4 out of 10 claims, you pay first
Hospital payer-mix tables show self-pay running far higher than insurance penetration would suggest. This table explains why, and it resolves a puzzle that has sat unanswered in Indian provider analysis for years.
| Mode of settlement | Claims (lakh) | By number | ₹ crore | By value |
|---|---|---|---|---|
| Cashless only | 188.95 | 57.96% | 62,537 | 66.35% |
| Reimbursement only | 119.58 | 36.68% | 27,649 | 29.34% |
| Both cashless and reimbursement | 4.33 | 1.33% | 2,912 | 3.09% |
| Benefit based | 13.14 | 4.03% | 1,150 | 1.22% |
In about 4 out of every 10 claims, the patient pays the hospital first and gets the money back from the insurer later.
The hospital records that as a cash payment. The insurer records it as an insurance payment. Both are right, and the same ₹27,649 crore gets counted two different ways.
In reimbursement mode the patient pays the hospital and claims afterwards. The hospital books it as self-pay. The insurer books it as insurance. Both are telling the truth, and the same ₹27,649 crore appears in one place as household spending and in another as insured expenditure.
Roughly 29% of all insurance claim value is therefore invisible to hospital payer-mix reporting. Any analysis that reads hospital self-pay share as a measure of household burden is overstating it by something close to that margin.
Two people, same premium, very different insurer
Incurred claims ratio is the share of premium paid back out as claims. Across Indian health insurers it spans 41 points, and the split by insurer type is almost perfectly clean.
| Insurer category | Health premium ₹ cr | Claims ratio |
|---|---|---|
| Standalone health specialists | 38,414 | 68.06% |
| Private general insurers | 43,144 | 83.46% |
| Public sector insurers | 45,860 | 99.84% |
For every ₹100 of premium collected, Niva Bupa pays out ₹61 in claims. Oriental Insurance pays out ₹103, more than it collected.
What to look for: the colours do not mix. Every standalone health insurer sits below 75, every public sector insurer above 96. That is not a spectrum of competitive performance; it is two different businesses wearing the same label.
View this chart as a table · 18 insurers
| Insurer | Type | Health premium ₹ cr | Paid out per ₹100 of premium |
|---|---|---|---|
| Niva Bupa | Standalone health | 6,762 | ₹61.22 |
| Care Health | Standalone health | 8,297 | ₹64.53 |
| Star Health | Standalone health | 16,716 | ₹70.30 |
| Aditya Birla Health | Standalone health | 4,822 | ₹71.50 |
| ManipalCigna | Standalone health | 1,798 | ₹74.81 |
| Tata AIG | Private general | 3,592 | ₹76.24 |
| SBI General | Private general | 4,731 | ₹82.19 |
| ICICI Lombard | Private general | 7,636 | ₹82.24 |
| Go Digit | Private general | 1,626 | ₹83.78 |
| HDFC ERGO | Private general | 6,100 | ₹84.85 |
| Bajaj Allianz | Private general | 8,043 | ₹87.31 |
| Reliance General | Private general | 2,217 | ₹87.34 |
| Future Generali | Private general | 1,859 | ₹95.29 |
| National Insurance | Public sector | 8,590 | ₹96.05 |
| Universal Sompo | Private general | 1,089 | ₹97.12 |
| United India | Public sector | 7,519 | ₹97.51 |
| New India Assurance | Public sector | 19,756 | ₹100.98 |
| Oriental Insurance | Public sector | 9,994 | ₹102.58 |
Source: IRDAI Annual Report 2024-25, Statement 10. Statement 10 includes personal accident within health, so these differ slightly from Table I.27.
The pattern is not random. Companies that only sell health insurance pay out the least. Government-owned insurers pay out the most, and lose money doing it.
Every standalone health specialist sits between 61% and 75%. Every public sector insurer sits between 96% and 103%. There is no overlap.
The public sector carries the government and group books at a structural loss while the specialists run retail individual business at roughly a thirty point margin. Functionally, public insurers are the risk pool that makes the private market profitable.
One thing moving in the other direction: every standalone health insurer's ratio rose in FY25. ManipalCigna by 11.03 points, Care Health by 6.84, Star Health by 3.83. Medical inflation is reaching the specialists while the public insurers reprice. Worth tracking as a two-year trend rather than calling now.
What you are worth to an insurer: ₹385 or ₹7,750
India covers 58 crore lives under health insurance. What each life is worth to an insurer varies by a factor of twenty.
| Class of business | Lives (crore) | Share of lives | Premium ₹ cr | Share of premium | Premium per life |
|---|---|---|---|---|---|
| Government sponsored | 24.54 | 42.3% | 9,459 | 8.0% | ₹385 |
| Group | 27.51 | 47.4% | 61,435 | 52.3% | ₹2,233 |
| Individual | 6.01 | 10.3% | 46,611 | 39.7% | ₹7,750 |
A person covered by a government scheme is worth ₹385 a year to an insurer. A person who buys their own policy is worth ₹7,750. Twenty times more.
Government schemes cover the most people and bring in the least money. Insurers are now writing less of that business every year.
The direction of travel is the finding. Government-sponsored lives covered fell 6.01% and government premium fell 10.03% year on year, the only segment shrinking on both measures while total premium grew 9.12%.
The reason is visible two tables up. Government business ran an incurred claims ratio of 115.28% in FY24. It is the least profitable line in Indian health insurance, and insurers are writing less of it.
The birth you get depends on which door you walk through
Insurance expanded nineteen points between survey rounds. Over the same period the public share of institutional births fell, and the caesarean rate rose by more than a quarter.
| Indicator | NFHS-5 | NFHS-6 | Change |
|---|---|---|---|
| Households with health cover | 41.0% | 60.2% | +19.2 |
| Institutional births in a public facility | 61.9% | 58.6% | −3.3 |
| Caesarean section rate, all facilities | 21.5% | 27.2% | +5.7 |
| Caesarean rate in private facilities | 47.4% | 54.1% | +6.7 |
| Caesarean rate in public facilities | 14.3% | 16.9% | +2.6 |
More people got insured. Fewer of them gave birth in a government hospital.
And once inside a private hospital, more than half of all births are now by caesarean, against 1 in 6 in government hospitals.
In urban India the public share of institutional births is now 48.2%. A majority of urban institutional births happen in private facilities, a threshold crossed without announcement. The urban caesarean rate is 40.5%, and in urban private facilities it is 57.4%.
Twenty-one states, no exceptions
Read one row at a time. The blue dot is the caesarean rate in that state's government hospitals. The red dot is the rate in its private hospitals. The line between them is the difference.
What to look for: a single row where the blue dot sits to the right of the red one. There is not one. In all 21 states the private rate is higher, and the gap ranges from 3 points to 64.
View this chart as a table · 21 states, largest gap first
| State or UT | Government hospital | Private hospital | Gap, percentage points |
|---|---|---|---|
| Assam | 17.4% | 81.4% | +64.0 |
| Odisha | 19.9% | 76.8% | +56.9 |
| Chhattisgarh | 12.2% | 64.9% | +52.7 |
| Madhya Pradesh | 10.4% | 61.7% | +51.3 |
| Jharkhand | 6.1% | 54.1% | +48.0 |
| Bihar | 2.7% | 49.3% | +46.6 |
| Arunachal Pradesh | 16.7% | 62.1% | +45.4 |
| Jammu and Kashmir | 48.6% | 90.0% | +41.4 |
| Himachal Pradesh | 23.3% | 63.7% | +40.4 |
| Goa | 31.4% | 69.7% | +38.3 |
| NCT of Delhi | 19.6% | 50.9% | +31.3 |
| Meghalaya | 5.0% | 36.2% | +31.2 |
| Karnataka | 34.0% | 63.8% | +29.8 |
| Punjab | 34.2% | 63.3% | +29.1 |
| Haryana | 15.2% | 40.2% | +25.0 |
| Puducherry | 37.4% | 62.4% | +25.0 |
| Mizoram | 10.3% | 35.2% | +24.9 |
| Maharashtra | 23.8% | 48.5% | +24.7 |
| Gujarat | 15.6% | 39.8% | +24.2 |
| Nagaland | 9.4% | 32.5% | +23.1 |
| Kerala | 39.3% | 42.5% | +3.2 |
Source: NFHS-6 State and UT Fact Sheets, IIPS, May 2026. Thirteen jurisdictions have no extractable numeric layer and are excluded rather than estimated.
Same state. Same women. Same year. The only thing that changes is which door she walks through.
In Bihar, about 3 out of every 100 births in a government hospital are caesarean. In a private hospital in the same state, it is 49 out of 100.
Across all 21 states with data, the private rate is higher than the public rate. Not most of them. All of them.
The usual explanation does not survive testing. People assume private hospitals see older, wealthier, higher-risk mothers, so naturally they operate more often. If that were true, richer and better-educated states should show higher private caesarean rates.
They do not. Across the 21 states, the link between women's education levels and the private-hospital caesarean rate is almost exactly zero. Education explains how many caesareans a state performs overall. It explains nothing about what happens inside a private hospital.
What does predict it is the size of the private sector, and it runs backwards. Where private hospitals handle the fewest births, they perform the most caesareans. In Jammu and Kashmir private hospitals handle one in five births, and 90 out of every 100 of those are surgical. In Kerala they handle two thirds of births, and the figure drops to 42, almost the same as the government rate.
One honest caveat. Bihar's public rate of 2.7% is far below what genuine clinical need would produce. So this is not only private hospitals doing too many. It is also government hospitals doing too few. The gap measures a system with no agreed medical threshold on either side.
What still cannot be checked
Everything above comes from two government sources. Reconciling them against what hospitals report is impossible, because hospitals do not report it.
What to do with this
Four readers, four different actions. The numbers above are the same for all of them.
Sources and evidence status
What is officially reported, what I calculated from it, and what is directional.
Before citing externally: Statement 10 includes personal accident within health, so firm-wise ratios differ slightly from Table I.27, which excludes personal accident and travel. The caesarean correlations rest on 21 states and cannot establish cause. The reimbursement double-counting estimate follows from the mode-of-settlement split and is arithmetic rather than a measured figure.