BestWorstInsurance

India · Doctors indemnity · Research

How Do Indemnity Insurers Actually Behave When Indian Doctors Claim?

By Best Worst Insurance Editorial TeamUpdated 25 July 2026

This is the dataset that does not exist anywhere else, and it exists because a better one is impossible. No Indian regulator publishes professional-indemnity claim-settlement data — the phrase does not occur in the IRDAI Annual Report, the statistics Handbook, the GI Council Yearbook or the Insurance Ombudsman's annual report, each of which we text-searched to confirm it [1,2,3,4]. So we catalogued the only public record of these policies being tested: 25 Indian court and consumer-forum judgments in which a doctors' indemnity insurer was a party. Insurers contested liability or coverage in roughly 22 of the 25 and paid voluntarily in none [5]. In 10 of the 25 they ran or adopted the doctor's own defence against the patient — the product working as intended [5]. Only three outright repudiations were upheld, on three distinct grounds [5,6,7,8]. One warning travels with every number here: claims insurers pay quietly never generate a judgment, so this sample over-represents fights and can never be read as a settlement ratio.

Every ground insurers ran, and whether it worked

A litigated-disputes sample. Claims paid without a fight never generate a judgment, so these are conduct patterns within litigation, not settlement rates [5].

Ground the insurer ranFrequency in the 25-case sampleDid it work?
Late or missing notification of the claim4 of 25 — the most litigated single ground [5]Upheld where the delay was gross: a nine-year gap ended the claim, the NCDRC holding notice conditions substantive rather than formal [6]. Rejected where the insurer was on record and negligence was proven
No negligence — the merits defence10 of 25 [5]Works on thin evidence; in one case the insurer appealed after its insured had stopped and got a ₹9 lakh award set aside entirely [9]. This is the product working
Not a necessary party, or no direct payment to the complainantRaised repeatedly [10]Increasingly succeeds — the NCDRC held in 2024 that insurers cannot be directed to pay the complainant directly [11]
No policy was ever issued to this doctor or hospital3 of 25 [5]Never conclusively tested, because those complaints failed on other grounds — but it means a doctor may have to prove their own policy years later
Deliberate, conscious or intentional disregard exclusion1 case [7]UPHELD. The doctor personally bore roughly ₹3.1 lakh [7]
Wrong policy type — legal liability, not professional indemnity1 case [8]UPHELD. A ₹20 lakh claim failed and the hospital bore the award itself
Claim event fell outside the policy period1 case [12]Raised successfully — earlier consultations were inside the period but the treatment triggering the claim was not
Territorial limitation — cover only at the doctor's own clinic1 case [13]REJECTED. The policy indemnified practice in any place in India
Rateable proportion — push liability to a co-insurer1 case [14]Failed. The insurer was still directed to pay ₹20 lakh

1. Why this dataset exists

We would rather have published claims data. It does not exist.

The phrase "professional indemnity" does not occur in the IRDAI Annual Report, the Handbook on Indian Insurance Statistics, the GI Council Yearbook or the Council for Insurance Ombudsmen's annual report. We text-searched each document to confirm the absence rather than inferring it [1,2,3,4]. The only insurer-wise claims ratio IRDAI publishes stops at five segments — Fire, Marine, Motor, Health and "Other Segments" — with liability folded into that last bucket alongside crop, credit, engineering and aviation [2,1]. Even the GI Council's finest published split stops one level above doctors' PI, in a ₹4,187.8 crore "other liability covers" line shared with D&O, E&O and cyber [15].

So when a seller quotes you a claim settlement ratio for doctors' indemnity, they have made it up. There is no source they could be quoting.

What the regulator publishes instead, and precisely what each figure can and cannot support.

Published figureWhat it cannot tell you
No PI-specific claim data exists in any regulator publication [1,2,3,4]Anything at all. This is the baseline finding that makes the judgment sweep necessary
Liability segment net claims ratio 31.4% to 51.0% over ten years, 46.2% in FY2023-24 [3][3]Nothing doctor-specific — liability aggregates workmen's compensation, D&O, E&O, cyber, product and commercial general liability
Insurer-wise ratios stop at five segments [2,1]Any per-insurer liability performance — the bucket is blended with crop, credit, engineering and aviation
52.6% of liability premium ceded to reinsurers [1]How the gross claim in front of the policyholder was handled. Published ratios are net
₹1,328.5 crore of liability claims paid in FY2024-25, about 1.2% of non-health claims [1]The doctors' PI share, which is not separately reported even by the GI Council [15]
1,04,643 liability claims reported and 91,434 paid in FY2023-24, about 87% by count [3]Whether doctors' PI claims behave like the segment average. This is all liability lines together
86.7% of entertainable general-stream ombudsman complaints concern repudiation [4]PI specifically — the Ombudsman publishes no line-of-business split within general insurance

What does exist is the litigation record. We ran a full-text sweep of Indian court and consumer-forum decisions for doctors' professional indemnity disputes and read 25 judgments in full, coding insurer conduct case by case. That is this page.

2. The bias warning, stated before the findings rather than after

A catalogue of judgments over-represents conflict by construction. An insurer that pays a claim promptly generates no public record at all, so every case in this sample is, by definition, one that went wrong for somebody. The sample also skews toward the four public-sector insurers, because they wrote the bulk of Indian doctors' PI over the study window [5].

This means two specific things you should hold onto while reading. First, the "contested in 22 of 25" figure is not a rejection rate; it is a description of what litigation looks like. Second, the insurer appearance counts are a map of market share and policy vintage, not a league table of behaviour.

3. What insurers actually argued

The hero table above is the complete list of grounds run across the sample. Four patterns deserve drawing out.

Late notification is the claim-killer. It was the most litigated single ground, appearing in 4 of the 25 cases [5], and it produced the only NCDRC-upheld repudiation against a doctor in the entire set. An orthopaedic surgeon lost a negligence case in which the complaint was filed in 1999; he told his insurer in 2008, when execution proceedings began. The NCDRC upheld the repudiation and held that the obligation to give written notice as soon as reasonably practicable and to forward every summons immediately was substantive, not a formality [6]. It is also, notably, the ground most completely within a doctor's own control.

The merits defence is the product working. In 10 of the 25 cases the insurer ran or adopted the defence against the patient's negligence claim [5]. The clearest example runs the opposite way to the cynical narrative: New India appealed to the NCDRC after a State Commission had awarded roughly ₹9 lakh against a hospital, argued there was neither deficiency in the treatment nor negligence by the treating doctors, and got the award set aside entirely [9]. The insurer kept fighting after its insured had stopped.

Insurers increasingly resist being in the case at all. The NCDRC held in 2024 that PI insurers cannot be directed to pay the complainant directly at the consumer-forum stage [11], and insurers have run a locus argument that only the policyholder is entitled to proceed against them [10]. The consequence for a doctor is a cash-flow problem: you pay the award, then recover.

Three repudiations were upheld, on three different grounds. Nine-year-late notification [6]; the "deliberate, conscious or intentional disregard" exclusion, after which the doctor personally bore roughly ₹3.1 lakh [7][7]; and a hospital that had bought a legal-liability policy instead of professional indemnity, whose ₹20 lakh claim failed completely [8]. None of the three was about whether the doctor was negligent [5,6,7,8].

4. Insurer by insurer, with the caveat attached

Appearances in the 25-judgment catalogue. Read this as a map of who wrote the policies, not a league table of who pays.

InsurerAppearancesPattern in the cases we read
New India Assurance13 of 25 [5]Contests nearly everything, including appellate fights after its insured has stopped [9]; twice denied a policy existed at all [5]. Consistent with the largest historical book, not a verdict on payment
United India8 [5]Repeatedly paid only up to small sums insured, leaving the doctor to fund the gap — ₹10 lakh awarded against a ₹4.75 lakh limit in one case [16]
Oriental4 [5]Fought allocation and intimation points but paid when negligence stuck, including its full ₹5 lakh limit in a 2025 NCDRC matter
National Insurance2 [5]Small sample, but holds the only upheld exclusion-based repudiation in the set [7]
ICICI Lombard1 [5]Resisted being impleaded mid-case; private-insurer PI disputes have barely reached the fora despite ICICI being the largest liability writer by premium [15]
Reliance General (now IndusInd)1 [5]Raised non-notification

Two defences worth noting because they failed. An insurer argued that cover extended only to the doctor's own clinic and not to a temporary arrangement at another hospital; the forum rejected it, because the policy indemnified practice "in any place in India" [13] — useful precedent for any doctor who rotates or moonlights. And an insurer invoked the rateable-proportion clause to push liability onto the doctors' personal PI insurer where more than one cover existed, and was still directed to pay ₹20 lakh [14].

One defence that succeeded is worth equal attention. In a 2026 Uttarakhand matter the insurer argued it need not indemnify because the treatment triggering the claim fell outside the annual policy period, even though earlier consultations fell inside it [12]. On claims-made annual policies, the dates around treatment decide period disputes — which is the practical argument for never letting cover lapse, examined in claims-made policies and retroactive dates.

5. The loss nobody calls a rejection

The single most common way an Indian doctor loses money on this product does not appear in the repudiation table at all, because it is not a repudiation. It is the limit running out.

In a Kerala case, ₹10 lakh was awarded for a botched laparoscopic hysterectomy against a policy limit of ₹4.75 lakh; the insurer paid its limit and the hospital and surgeon personally funded the ₹5.25 lakh gap [16]. The policy performed exactly as written. United India shows this pattern repeatedly across the sample, paying to small sums insured and leaving the doctor to fund the difference [5].

The structural reason is that limits were set in a different era. The modal sum insured across the 2000-2015 cases in this sweep was ₹10 lakh [5], against a Supreme Court quantum regime that reached roughly ₹1 crore in 2009 [17] and ₹6.08 crore plus interest by 2013 [18]. The award data is tracked separately at the negligence award tracker.

6. What the analysis actually supports

Six conclusions survive the bias warning.

  • Notification discipline is the highest-value habit available to a doctor. It is the top litigated ground [5], the only upheld repudiation against a doctor at NCDRC level [6], and free.
  • Keep every policy schedule forever. Insurers have denied a policy existed [5], and claims arrive up to sixteen years after treatment [19].
  • Buy the right product. A legal-liability policy is not professional indemnity, and the difference destroyed a ₹20 lakh claim [8].
  • Buy an adequate limit. Under-insurance is the commonest real loss and is invisible in every rejection statistic [16].
  • Expect to fund the award first [11].
  • Do not choose a carrier on this data. Choose on the wording — the excess, the reporting window, the hammer clause — which is what the best and worst ranking scores, and which is the only dimension where the evidence is complete rather than sampled.

The grounds-by-ground breakdown for a doctor facing a live dispute is at why indemnity claims get rejected, and the sector-level comparison this data is often misused to settle is at PSU vs private insurers.

Frequently asked questions

What is the claim settlement ratio for professional indemnity insurance in India?
There isn't one, and nobody can produce one. The phrase "professional indemnity" does not occur in the IRDAI Annual Report, the Handbook on Indian Insurance Statistics, the GI Council Yearbook or the Insurance Ombudsman's annual report — we text-searched each to confirm it [1,2,3,4] — and IRDAI's insurer-wise claims tables fold liability into an "Other Segments" bucket alongside crop, credit and engineering [2,1]. Any figure presented as a PI settlement ratio is invented.
Do indemnity insurers actually pay doctors’ claims in India?
The litigated record shows insurers contesting in roughly 22 of 25 cases and paying voluntarily in none [5] — but that sample is built from disputes, and claims paid without a fight never generate a judgment. The more useful finding is that in 10 of the 25 the insurer ran or adopted the doctor's own defence against the patient [5], and only three outright repudiations were upheld [5,6,7,8].
What is the most common reason an indemnity claim fails in India?
Late or missing notification. It was the most litigated ground in our sample at 4 of 25 cases [5], and the anchor case is an orthopaedic surgeon whose insurer heard about a 1999 complaint only in 2008, when execution began — the NCDRC upheld the repudiation and held that notice conditions are substantive obligations, not formalities [6].
Which insurer fights doctors’ claims the most?
The counts do not support that question. New India appears in 13 of the 25 judgments, more than any other insurer, but that reflects its historically largest book in the line and the age of those policies rather than its claims behaviour [5]; United India appears 8 times, Oriental 4, National twice, and ICICI Lombard and Reliance once each [5]. Private-insurer PI disputes have barely reached the fora.
Will my insurer pay the patient directly if I lose?
Generally not at the consumer-forum stage. The NCDRC held in 2024 that PI insurers cannot be directed to pay the complainant directly [11], and insurers have separately resisted being joined at all, arguing that only the policyholder is entitled to proceed against them [10]. Expect to fund the award and recover afterwards.
Can an insurer deny that my policy ever existed?
It has happened, in 3 of the 25 judgments we catalogued [5]. None of those denials was conclusively tested because the complaints failed on other grounds, but the practical lesson is unambiguous: keep every policy schedule from every year forever, including from jobs you have left, because a claim can arrive sixteen years after the treatment [19].
Do insurers ever defend the doctor rather than fight them?
Routinely, and this is the part the pessimistic reading misses. In 10 of the 25 cases the insurer ran or adopted the merits defence against the patient's negligence claim [5], and in one, New India appealed to the NCDRC after its insured had stopped and got a roughly ₹9 lakh award set aside entirely, arguing there was neither deficiency in the treatment nor negligence by the treating doctors [9].
Why do you keep saying this is not a settlement ratio?
Because it structurally cannot be one. Judgments exist only where there was a dispute — an insurer that pays a claim promptly generates no public record at all — so a catalogue of judgments over-represents conflict by construction, and it skews toward the public-sector insurers that wrote most doctors' policies over the study window [5][5]. It is a map of how these policies behave when tested in court, and nothing more.

Sources

Every figure on this page is footnoted to one of the primary documents below. Reliability tiers: A = regulator, court or filed document; B = reputable publisher or carrier official page; C = user-generated (reported by users).

  1. 1.IRDAI — IRDAI Annual Report 2024-25. https://irdai.gov.in/annual-reportsTier A · Regulator / court / filed document · Published 2025-12-01 · Accessed 2026-07-17
  2. 2.IRDAI — Handbook on Indian Insurance Statistics 2024-25. https://irdai.gov.in/handbook-of-indian-insuranceTier A · Regulator / court / filed document · Published 2026-02-01 · Accessed 2026-07-17
  3. 3.General Insurance Council — GI Council Yearbook 2023-24. https://www.gicouncil.in/yearbook/2023-24/Tier A · Regulator / court / filed document · Accessed 2026-07-17
  4. 4.Council for Insurance Ombudsmen — Insurance Ombudsman Annual Report 2023-24. https://cioins.co.in/annualreports/AnnualReport2023-2024.pdfTier A · Regulator / court / filed document · Published 2024-09-01 · Accessed 2026-07-17
  5. 5.IndianKanoon (catalog compiled by Best Worst Insurance) — Doctors' PI judgments sweep — 25-case catalog of insurer conduct (consumer fora, High Courts, Supreme Court). https://indiankanoon.org/search/?formInput=%22professional%20indemnity%22%20doctorTier A · Regulator / court / filed document · Accessed 2026-07-17
  6. 6.NCDRC (via IndianKanoon) — Dr. Tarunjit Dutta Roy v. New India Assurance (NCDRC, 2013). https://indiankanoon.org/doc/194587303/Tier A · Regulator / court / filed document · Accessed 2026-07-17
  7. 7.West Bengal SCDRC (via IndianKanoon) — Dr. Shymal Baran Mondal v. National Insurance (WB SCDRC, 2019). https://indiankanoon.org/doc/148807510/Tier A · Regulator / court / filed document · Accessed 2026-07-17
  8. 8.Gujarat SCDRC (via IndianKanoon) — K.M.G General Hospital v. United India Insurance (Gujarat SCDRC, 2022). https://indiankanoon.org/doc/108725571/Tier A · Regulator / court / filed document · Accessed 2026-07-17
  9. 9.NCDRC (via IndianKanoon) — New India Assurance v. Palagiri Kumari (NCDRC, 2020). https://indiankanoon.org/doc/63936975/Tier A · Regulator / court / filed document · Accessed 2026-07-17
  10. 10.Chhattisgarh SCDRC (via IndianKanoon) — Smt. Nisha Rathore v. We Care Super Speciality Hospital (Chhattisgarh SCDRC, 2023). https://indiankanoon.org/doc/192504399/Tier A · Regulator / court / filed document · Accessed 2026-07-17
  11. 11.NCDRC (via IndianKanoon) — Gopal Aggarwal v. Sarvodya Hospital (NCDRC, 2024). https://indiankanoon.org/doc/73878019/Tier A · Regulator / court / filed document · Accessed 2026-07-17
  12. 12.Uttarakhand SCDRC (via IndianKanoon) — United India Insurance v. Smt. Vijay Sharma (Uttarakhand SCDRC, 2026). https://indiankanoon.org/doc/161990819/Tier A · Regulator / court / filed document · Accessed 2026-07-17
  13. 13.Kerala SCDRC (via IndianKanoon) — New India Assurance v. Smt. Parvathikutty K. (Kerala SCDRC, 2011). https://indiankanoon.org/doc/176922002/Tier A · Regulator / court / filed document · Accessed 2026-07-17
  14. 14.NCDRC (via IndianKanoon) — Deep Nursing Home v. Manmeet Singh Mattewal (NCDRC, 2012). https://indiankanoon.org/doc/145604686/Tier A · Regulator / court / filed document · Accessed 2026-07-17
  15. 15.General Insurance Council — GI Council segment-wise premium report FY2024-25 (final, provisional & unaudited). https://www.gicouncil.in/media/4513/final-segment-24-25.xlsxTier A · Regulator / court / filed document · Published 2025-05-01 · Accessed 2026-07-17
  16. 16.Kerala SCDRC (via IndianKanoon) — United India Insurance v. Lathika (Kerala SCDRC, 2015). https://indiankanoon.org/doc/4522398/Tier A · Regulator / court / filed document · Accessed 2026-07-17
  17. 17.Supreme Court of India (via IndianKanoon) — Nizam's Institute of Medical Sciences v. Prasanth S. Dhananka (SC, 2009). https://indiankanoon.org/doc/1110505/Tier A · Regulator / court / filed document · Accessed 2026-07-17
  18. 18.Supreme Court of India (via IndianKanoon) — Balram Prasad v. Kunal Saha (SC, 2013) — AMRI / Anuradha Saha. https://indiankanoon.org/doc/35346928/Tier A · Regulator / court / filed document · Published 2013-10-24 · Accessed 2026-07-17
  19. 19.DailyRounds (reporting NCDRC) — NCDRC orders ₹1 crore against Indraprastha Apollo Hospitals and Dr Sohini Verma (order dated 22 April 2015; delivery 10 June 1999). https://www.dailyrounds.org/blog/15-years-after-delivery-ncdrc-orders-rs-1-crore-as-compensation-for-medical-negligence/Tier B · Reputable publisher / carrier official · Published 2015-04-22 · Accessed 2026-07-25