Reading the number
How to read a hospital’s success rate, and what independent outcome data exists in India.
Someone has quoted you a figure, and you want to know whether it means anything. This page takes one apart, sets out which outcome datasets India holds as of September 2026, and says what to ask for in writing.
A hospital’s success rate is a claim until it comes with what it counts, out of how many patients over which dates, and who counted and checked it. Ask for all four in writing before comparing it with anything, including a figure from your own country.
India’s joint replacement and heart surgery registries depend on units choosing to contribute, and the surgeons who write about them say they capture only part of the picture. Its cancer registries cover part of the population. Our reading as of September 2026 found no hospital-level or surgeon-level outcome results from India’s joint replacement, cardiac surgery or cancer registries that a patient could use to compare hospitals.
This site’s own reading, on 25 September 2026, of the registry pages, reports, leaflets and papers cited on this page Checked 2026-09-25.
An absence finding from our own reading, not a statement by any registry. A registry may share results in ways we could not see. The joint registry’s own site links no annual report; the surgeon society’s site carries a public aggregate slide deck, updated in August 2026, of counts and breakdowns with no hospital or surgeon results.
Some registries elsewhere are voluntary or partial too, and the UK joint registry, the UK cardiac surgeons’ society and the US society, all of which publish results, each caution against reading them as a league table. How to read a hospital’s own success rate is set out below.
Medically reviewed
Reviewed by Dr Gunjan Patel, MBBS, registered with the Gujarat Medical Council (G-65059). Last reviewed 25 September 2026. Reviewed for accuracy only. This page is general information, not advice about your own case, and Dr Gunjan Patel does not diagnose, prescribe or treat.
A claim, taken apart
One sentence from a hospital page, and the five things it leaves out.
Any of these gaps can sit under a true number, and each one makes comparison unsafe until it is filled.
“Our cardiac team has a 99 percent success rate.”
An invented sentence in the style of a hospital web page, with no real hospital behind it.
What counts as success
Discharge, six weeks, five years, an implant that has not needed revision, a tumour that has not come back: each is a fair measure, and no two count the same thing. Even the word survival covers different things. One 2025 national adult cardiac surgery audit report for England, Wales and Northern Ireland gives different survival figures for the same service depending on which patients and which period it counts.
National Institute for Cardiovascular Outcomes Research, National Adult Cardiac Surgery Audit 2025, 2nd edition, annual report summary Checked 2026-09-25.
Read from the audit report on 25 September 2026. It shows how one report can hold several correct figures and is not a benchmark for any hospital.
Out of how many
Every rate has a top and a bottom. The top is what went well. The bottom is every operation that was counted. Emergency cases, revisions and the most complicated patients may or may not be inside it, and a figure can describe one surgeon or a whole unit without saying which.
A rate built on a small number of operations can swing widely from one year to the next. A registry’s own totals raise the same question. About 27,000 knee replacements were reported to the Indian Joint Registry for 2019, against a market-survey estimate of about 200,000 knee replacements done in India in 2020, and the authors of a 2025 editorial call this a small fraction of the total; the two figures come from different years, so no coverage rate should be read from them.
Vaidya et al., "India Joining the World of Hip and Knee Registries", Indian Journal of Orthopaedics, 55(Suppl 1):46-55, 2020, and a 2025 editorial that cites it Checked 2026-09-25.
The 2020 paper is by authors linked to the surgeon society that runs the registry and its software supplier. The 2025 editorial is by three orthopaedic surgeons who argue for a mandate. The 2019 count is given as around 27,000 and is not confirmed as a full calendar year, and the 200,000 is a market-research estimate for another year, so we give no coverage percentage.
How sick the patients were
Sicker patients make the arithmetic unfair. A unit that accepts the sickest patients can post a lower rate than one that refers them elsewhere, so a raw figure can reward the wrong thing. Observed death rates between 23 surgeons ranged from 0 to 3.7 percent in a 2003 BMJ study of 8,572 first-time isolated bypass operations at four NHS centres in north west England between 1999 and 2002, counting in-hospital deaths, and the authors concluded that crude comparisons of death rates can be misleading and may encourage risk-averse behaviour. Risk adjustment tries to correct for that, and it has limits of its own. An Indian cardiac surgeon adds a caution. EuroSCORE and the Society of Thoracic Surgeons risk models were built mainly on patients and practices outside India, a September 2026 editorial says, and while they usually fit reasonably, India has lacked its own data to check how well.
Bridgewater et al., BMJ, 327:13-17, 2003 Checked 2026-09-25.
The authors found their risk score predicted poorly in high-risk patients and recommended comparing low-risk cases, without further risk adjustment, as the benchmark. Full text read.
P. Narayan, editorial, Indian Journal of Thoracic and Cardiovascular Surgery, 42:1113-1114, published 14 September 2026 Checked 2026-09-25.
An author’s view about calibration data. It is not a finding that Indian outcomes are better or worse than outcomes anywhere else.
Over what period
One year, a multi-year average and a long follow-up are three different claims, and so is a figure counted at discharge against one counted months later. Without dates, a number can’t be checked against anything. The US Society of Thoracic Surgeons says a rating quoted publicly must state the dates it covers.
Who counted, and who checked
Results a hospital counts itself and posts on its own page are one thing. Cases entered into a registry and looked at by someone outside the unit are another. Under NABH, the national hospital accreditation scheme, the counting starts inside the hospital. The NABH Hospital Accreditation Standards, 6th Edition, require a hospital to identify and monitor key indicators of clinical structures, processes and outcomes, say its quality team verifies the data, and let each institution design its own method of collecting it.
NABH Hospital Accreditation Standards, 6th Edition, effective 1 January 2025, standard PSQ.3 and the indicator annex Checked 2026-09-25.
The annex gives hospital-wide examples such as unplanned return to theatre, surgical site infection and an intensive care mortality ratio, and specialty examples that are mostly checks on how care was delivered. As of September 2026 we found no bypass or joint replacement outcome indicator in the text, and nothing in it says the figures are published or checked outside the hospital. The guidebook that interprets each element was not read.
Where results are published, the publisher’s own rules shape what appears, so a missing or masked entry proves nothing either way. Before publication, the UK National Joint Registry gives consultants a review period in which they can ask for their data to be suppressed, a request that must be submitted by a set date to be considered, and it masks the numbers for any consultant in charge with fewer than five procedures.
National Joint Registry (UK), surgeon and hospital outcome publication notice for the 2025/26 data Checked 2026-09-25.
From the registry’s 2026 notice, for data refreshed on 11 August 2026. An earlier notice tied suppression requests to serious concerns about data accuracy, and the 2026 page states no ground. We found no source on how often requests are granted, and a request is not an opt-out right.
Which gap matters most depends on your operation and your health, and only your own surgeon can weigh that.
Ask in writing
Four questions to send to any centre that quotes you a success rate.
Send them by email, so the answers exist on paper and your own doctor can read them.
What does the success rate count, and what counts as a failure? Ask for the definition in plain words, the procedure it covers, and whether it describes one named surgeon or the whole unit.
Out of how many operations, over which dates, and for which patients? Request the number of operations behind the figure, the dates it covers, which cases were left out, and whether it was adjusted for how ill the patients were, and by what method.
Who counted it, and has anyone outside the unit checked it? Find out whether results go into a registry or an outside audit, which one, and whether the unit can show you what comes back to it. As of September 2026, the surgeons writing about the joint replacement registry describe contributing as voluntary, and the cardiac surgeons’ association invites units to submit data. The authors of the 2025 Indian Joint Registry editorial say that making registry reporting a criterion for NABH hospital accreditation, or a requirement of insurers for procedure cover, would ensure that centres routinely contribute data.
Attributed account, not established factAneja, Machaiah and Shyam, editorial, Journal of Orthopaedic Case Reports, 15(11), 2025 Checked 2026-09-25.
A recommendation, not current policy. As of September 2026 a text search of the NABH Hospital Accreditation Standards, 6th Edition found no mention of a registry; the NABH guidebook that interprets each element was not read.
How do you follow international patients after they fly home? Ask what the centre records after discharge and at what intervals, how many international patients it has heard back from, and whether the surgeon will speak to your doctor at home.
A reply is more information, not a guarantee. A refusal to answer in writing is also an answer, and it’s worth keeping. Your own surgeon, and your doctor at home, can tell you what any figure means for your operation.
What exists in India, as of September 2026
Four sources of outcome data, and what each one can and cannot show.
All four are real and in use. The surgeons who write about the first two say in print that more units should take part.
| Source | What it covers | What it does not show | Who can see results |
|---|---|---|---|
| Indian Joint Registry, for hip and knee replacement | Operations reported by participating surgeons and hospitals; taking part is voluntary. | Surgeons who write about the registry call what it captures a small fraction of the total. | Surgeons and hospitals see their own results, and reports are aggregated. |
| National cardiac surgery database, from the cardiac surgeons’ association | Operations entered by participating Indian units over six years. | A national picture. A September 2026 editorial calls it a very small minority of Indian cardiac surgery. | The association links an aggregate report (the August 2026 edition is public) and offers analysed data to units with enough entries. We found no public unit-level or surgeon-level results. |
| National Cancer Registry Programme, run by the Indian Council of Medical Research | Cancer cases recorded by area-based and hospital-based registries, including incidence and mortality for defined areas. | Any single hospital’s treatment results. The 2020 report pools hospital-based data, and area figures are not a hospital success rate. | Anyone, through published reports and papers. |
| NABH accreditation indicators | Indicators the standards list and a hospital monitors, such as infections, unplanned returns to theatre and an intensive care mortality ratio. | No bypass or joint replacement outcome indicator found in the standard text as of September 2026, and hospitals may design their own collection method. | The text does not say who sees the figures. Separately, it asks governance to inform the public of quality and performance. |
This page covers joint replacement, heart surgery and cancer care, plus NABH. IVF success rates have their own definitions and sit on the fertility treatment page. Transplant is not covered on this page.
Joint replacement. The Indian Joint Registry for hip and knee replacement was started and is maintained by professional bodies of Indian orthopaedic surgeons, and participation remains voluntary; a November 2025 editorial by three Indian orthopaedic surgeons says a registry needs at least 90 percent of arthroplasty surgeons taking part to produce valid data, and that current voluntary reporting falls well short of that. Who sees results is a separate matter. The Indian Joint Registry’s own leaflet says results are published in aggregated form and that surgeon-level and hospital-level data go only to the individual participating surgeon or the nominated hospital user, while a May 2025 peer-reviewed review lists a 2019 newsletter and a 2021 slide deck as the registry’s annual-report material and says the number of contributing hospitals is not given.
Aneja, Machaiah and Shyam, editorial, Journal of Orthopaedic Case Reports, 15(11):439-445, 2025 Checked 2026-09-25.
The authors argue for making participation mandatory and declare no conflict of interest. The editorial does not say whether any of them help run the registry.
Indian Joint Registry, "Information for Surgeons and Hospitals" leaflet; Shum et al., EFORT Open Reviews, 10(5):250-257, 2025 Checked 2026-09-25.
The leaflet is undated, with file details pointing to 2021, and says surgeons and hospitals are encouraged to take part; the word voluntary comes from the surgeon-authored papers. As of 25 September 2026 the registry’s own site links no annual report. The surgeon society’s site carries a public aggregate slide deck updated on 24 August 2026, with data to March 2026, showing counts and age and indication breakdowns and no hospital or surgeon results; the May 2025 review described an earlier version.
Heart surgery. Close to 50,000 operations had been entered by Indian units over 6 years into the cardiac surgeons’ association’s database, which an Indian cardiac surgeon writing in September 2026 calls a very small minority of Indian cardiac surgical volume and not a national picture, adding that the nation has to participate. The Indian Association of Cardiovascular Thoracic Surgeons’ database page links a public August 2026 report of national case counts and outcome slides, with no unit-level or surgeon-level results in its text, and invites units to submit data, saying it is free, that privacy and confidentiality are assured and that a unit with more than 1000 entries gets analysed data. The risk scores Indian cardiac surgeons use are covered in the decoder above, under how sick the patients were.
P. Narayan, "Guidelines do not come out of thin air", editorial, Indian Journal of Thoracic and Cardiovascular Surgery, 42:1113-1114, published 14 September 2026 Checked 2026-09-25.
An editorial that argues for more units to take part, not a registry report. It gives no completeness figure and no procedure mix.
Indian Association of Cardiovascular Thoracic Surgeons, database page and "IACTS National Database Report August, 2026", file created 21 September 2026 Checked 2026-09-25.
Read on 25 September 2026, the report through its text layer, so its charts could not be read. Its adult in-hospital mortality slide notes that 15 percent of that field was left blank. We did not see what a contributing unit receives.
Cancer. Area figures exist, and they describe places and their record-keeping. They cannot stand in for a hospital’s success rate. Around 11 percent of India’s population was covered by the 38 population-based cancer registries of the National Cancer Registry Programme, according to a 2021 paper by authors from the programme’s own institute, which also counted 269 hospital-based registries. The National Cancer Registry Programme report published in 2020, covering data from 2012 to 2016, pools and analyses the data of the hospital-based registries it draws on, and lists the hospitals and their new-case counts but gives no treatment or outcome results for any single hospital.
Nath and Mathur, Indian Journal of Surgical Oncology, 13(Suppl 1):2-7, 2021 Checked 2026-09-25.
Coverage figures differ between papers from the same institute, because they count different sets of registries in different years: a 2025 paper by the programme’s investigators counts 43 population-based registries covering approximately 18 percent, and a 2023 paper gives 16.4 percent. None of them measures any hospital.
ICMR National Centre for Disease Informatics and Research, Report of National Cancer Registry Programme 2020 Checked 2026-09-25.
Read directly, with a text search of the report. As of 25 September 2026 the institute’s reports page lists it as the latest main programme report. We found no hospital-level outcome table in its text layer.
NABH indicators. The monitoring described in the decoder above is the hospital’s own. The NABH 6th Edition governance standard ROM.1 includes an element that those responsible for governance inform the public of the quality and performance of services, and leaves what that means to a separate guidebook that we did not read. What a hospital publishes under that governance element is a question for the hospital.
NABH Hospital Accreditation Standards, 6th Edition, effective 1 January 2025, standard ROM.1 Checked 2026-09-25.
Printed as a commitment-level element, which the standards say is assessed at the final assessment. The text prescribes no measure or format, so it neither names a published outcome figure nor rules one out.
Registries elsewhere
What the UK, the US and Australia publish.
Registries elsewhere differ from India's in what they publish, whether taking part is compulsory and how complete they are. None of that says which country's surgeons do better.
Submitting joint replacement data to the UK National Joint Registry is mandatory for NHS hospitals and independent-sector hospitals in England and Wales and for Northern Ireland hospitals doing publicly funded procedures, and the registry publishes a public profile for most consultants in charge of joint replacements, showing operation numbers and, for hip and knee replacements, mortality within 90 days of surgery, with small numbers masked and some profiles withheld. Its own pages are careful about how far that goes. The registry says its statistics should not on their own be taken as a guide to the standards of a hospital, that its adjustment for how ill patients were is not precise, that being flagged as a potential outlier does not mean performance is poor, and that it does not think it helpful to portray one surgeon as better than another.
National Joint Registry (UK), data completeness and quality page, and Surgeon and Hospital Profile pages Checked 2026-09-25.
The registry’s own data-completeness page reports 96.98 percent compliance for 1 April 2025 to 31 March 2026. The profile site’s FAQ says the registry cannot cross-check hospitals in some jurisdictions or in the independent sector, although its data-completeness page describes an audit of both NHS and independent hospitals.
National Joint Registry (UK), Surgeon and Hospital Profile FAQ and patient information pages Checked 2026-09-25.
The registry’s own words about its own joint replacement statistics, read from its public pages in 2026.
Heart surgery in the UK went another way. The Society for Cardiothoracic Surgery in Great Britain and Ireland says individual surgeon outcome data for adult cardiac surgery is no longer published, that data from its quality assurance programme put in the public domain will be at unit level only, and that the data do not easily lend themselves to league tables.
Society for Cardiothoracic Surgery in Great Britain and Ireland, unit and outcome data page Checked 2026-09-25.
The page is undated, but its quality-assurance text describes a plan to start data collection in April 2021 and reach a final stage in 2024, so the unit-level-only statement dates from about 2020 to 2021.
The US pattern is different again. The US Society of Thoracic Surgeons says its national database includes more than 95 percent of adult cardiac surgery procedures performed in the US, that public reporting of its star ratings is completely voluntary, and that ratings quoted publicly must carry the surgery dates and are not intended for direct comparison to other participants.
Society of Thoracic Surgeons, frequently asked questions on the national database and public reporting, updated July 2026 Checked 2026-09-25.
The 95 percent figure is the society’s own statement about its database. Participants are usually hospital surgery departments or groups, not individual surgeons. As of September 2026 we found no figure for what share of participants report publicly.
Australia’s national joint replacement registry says it began collecting data in 1999, became fully national during 2002, receives data from all public and private hospitals that do joint replacement, and treats revision surgery as its principal measure of outcome, while a federal ministerial quality-assurance declaration prohibits disclosure of information that identifies individual patients or health care providers and is known solely because of the registry.
Australian Orthopaedic Association National Joint Replacement Registry, background, hospitals and governance pages Checked 2026-09-25.
The registry describing itself, on undated pages, and we did not audit its completeness claim. The declaration covers information that identifies providers; it is not a statement about every report the registry publishes.
One gap matters to readers of this page in particular. As of September 2026, none of the registry descriptions read for this page says how a patient who has an operation in one country and then flies home to another is recorded or followed. So question four goes to the centre.
This site’s own reading, on 25 September 2026, of the registry pages, leaflets and papers cited on this page Checked 2026-09-25.
An absence in the documents read, not evidence that no arrangement exists. It is offered as a reason to ask the centre, and no more than that.
What we do with this
Where we fit, and where we do not.
You can send these four questions to any centre yourself, by email, and nothing on this page needs us. If you go on to full case management, getting the named surgeon, the implant and the exclusion list confirmed in writing is already on our task list.
The free 30 minute consultation is where you tell us which figure you were quoted and how it was worded. At the end of it we tell you honestly whether an independent opinion would help you. What happens once you are home is covered on the aftercare page.
We hold no outcome data of our own yet, so we cannot rank surgeons by results, and the public sources above would not support a ranking if we tried. We do not diagnose, prescribe or treat. What we do not claim, in full.
Questions
What people ask about a hospital's success rate.
Is a hospital success rate of 98 or 99 percent believable?
It can be true and still tell you very little. Whether it is believable depends on what was counted, out of how many, over which dates and by whom, and a figure that arrives without those can’t be tested. Once they arrive in writing, you have something to put in front of your own doctor.
A coordinator emailed me a PDF with a success rate and nothing else. What now?
Reply with the four written questions and keep the PDF. Ask for the answers as text in an email rather than on a call, so they can be read later and shown to your own surgeon. If some are refused, that tells you how much weight the number can bear. You can also tell us how the figure was worded on the free 30 minute consultation, and we will say honestly whether an independent opinion would help.
Can I compare an Indian hospital's figure with one from my own country?
Only if both count the same thing, over the same period, for patients who were equally ill, and a like-for-like match is hard to find. Even where registries publish results, in the UK and the US, their publishers caution that the figures are not a league table: the UK joint registry says its statistics should not on their own be taken as a guide to a hospital’s standards, and the US society says its ratings are not intended for direct comparison with other participants.
IVF has its own version of the problem, set out on the fertility treatment page.
What does NABH accreditation say about a hospital's results?
The standard text we read asks a hospital to monitor a set of indicators that the standards define, and lets it design its own method for collecting the data. It says the hospital’s quality team verifies the data, and it does not say the figures are published or checked by anyone outside the hospital. As of September 2026 we found no bypass or joint replacement outcome indicator in it. The same standards ask those responsible for governance to inform the public of quality and performance, and the standards say the interpretation of each element is in NABH’s separate guidebook, which we have not read, so ask the hospital what was reviewed, by whom and when. The page on NABH and the medical visa says more.
Bring the number you were quoted to a free conversation.
A free 30 minute consultation. Tell us which figure a centre gave you and how it was worded, and we will say honestly whether an independent opinion would help. Please do not send records or scans until we have spoken.