How we count
The NHS says the typical wait for planned surgery is about 14 weeks. We say 24. Both numbers are correct. They answer different questions, and only one of them is the question a patient is asking.
This page explains which figures we use, where they come from, and the choices we have made. It includes the choices that make our numbers look worse than the official ones, because those are the ones worth checking.
The two numbers
Every month NHS England publishes waiting time data in several forms. Two of them matter here.
The incomplete pathway figure counts everyone who is on a waiting list right now and asks how long each of them has waited so far. It is the number in the news, the number behind the 6.7 million waiting list total (NHS RTT, June 2026), and the basis of the 18 week standard. Across the specialties we cover it is currently about 14 weeks.
It is a snapshot of a queue. Nobody counted in it has finished waiting, so everyone in it will eventually wait longer than the figure says. And because new referrals join the list constantly, it is full of people who have barely waited at all, which pulls the middle down.
The admitted completed pathway figure counts people who were actually treated last month and asks how long each of them waited in total, from referral to operation. Across the three specialties we cover that is about 24 weeks.
These are finished journeys. Nobody in that number is still waiting.
We use the second one. The wait a patient wants to know about is the wait they will experience, not the current age of the queue. The NHS's own patient facing service, My Planned Care, does not use the incomplete figure either.
The admitted figure is longer than the incomplete figure at 116 of the 118 acute trusts that report orthopaedic data. It is not longer because we chose a pessimistic measure. It is longer because the incomplete figure is structurally biased low.
What the standard figure actually measures
This is the part we did not expect to find.
We tested how much of the difference between hospitals is explained by how quickly they treat people, and how much by something else. For each basis we take every trust's share of very short waits, meaning patients recorded at four weeks or under, and regress the trust's headline figure against it. On the admitted basis that share is a proxy for how much quick, minor work a hospital does. On the incomplete basis it is a proxy for how many recently referred patients are sitting on its list.
On the incomplete measure, the one everybody quotes, the correlation gives an r² between 0.73 and 0.84 depending on the specialty. Between three quarters and four fifths of the variation between hospitals is explained not by speed but by the composition of the list.
On the admitted measure it is lower in every specialty, and lower again at the upper end of the distribution where major surgery sits. In orthopaedics it falls to 0.32, and to 0.09 at the ninetieth percentile.
Put plainly: a hospital can look fast on the standard figure by having a rapidly growing waiting list, and slow by working through its oldest cases. League tables built on that figure are substantially ranking list churn rather than performance, and that is true in every specialty we have tested.
We have not seen this published anywhere. The script behind these figures is part of our codebase and runs against the published NHS file for any month. We will share it with anyone who wants to check it.
Nobody can tell you the wait for a hip replacement
Not us, not the NHS, not anyone.
Referral to treatment data is collected by treatment function, which means specialty. Trauma and Orthopaedics is one code covering hip replacements, knee replacements, carpal tunnel releases, arthroscopies and everything else an orthopaedic department does. There is no procedure level split, at any granularity, in any published NHS dataset.
We checked the alternatives. The National Joint Registry has volumes and outcomes but no waiting times. PROMs measures health gain, not waits. Hospital Episode Statistics carries a different, shorter clock and is not published as a wait by procedure. My Planned Care is specialty level.
So the finest sourced wait available to anyone is per trust, per specialty. When we show you a figure for a knee replacement, it is the orthopaedic figure at your hospital, and we say so. Nobody who tells you otherwise has a source.
This matters because the specialty figure mixes two populations with very different waits: day cases that are done in weeks, and major operations that take months. Where the data lets us separate them, we do. Where it does not, we say which population the number describes.
Where your wait actually sits
A wait of 40 weeks means one thing if you spend 35 of it waiting for a first appointment, and something completely different if you are seen in five weeks and then sit on a surgical list.
The published figure cannot tell them apart. It is one number from referral to treatment with no stages in it. The NHS stopped publishing milestone data around 2012.
We reconstruct the split by joining two separate NHS sources: the referral to treatment figure, which covers the whole pathway, and My Planned Care, which publishes the wait to a first outpatient appointment per trust per specialty. The difference between them is the part of the wait that comes after you are seen.
Two real hospitals, same specialty, same month:
| Wait to be seen | Then to surgery | Total | |
|---|---|---|---|
| Kingston and Richmond | 41 weeks | 8 weeks | 49 weeks |
| Queen Elizabeth, King's Lynn | 7 weeks | 50 weeks | 57 weeks |
The overall figures are within eight weeks of each other. The problem at each hospital is the opposite of the problem at the other, and the official number shows neither.
This changes the advice completely. At Kingston, paying privately to be seen skips most of the wait. At King's Lynn it skips almost none of it, and a patient who paid for a private consultation there would have spent money to save about seven weeks of a fifty seven week wait.
Across the 113 orthopaedic hospitals where we can compute the split, 40 are front loaded and 73 are back loaded. So for roughly two thirds of orthopaedic patients, paying to be seen sooner is not the answer, and we tell them that rather than selling them the option.
We can only do this for orthopaedics, and the reason is worth stating. In general surgery and ENT the wait to a first appointment comes out longer than the whole pathway wait for people who were treated. At one general surgery department it is 39 weeks to be seen against a 7 week pathway; at one ENT department, 63 against 28.
That is not an error in either figure. It is what happens when most referrals in a specialty do not end in one operation. The pathway figure describes the minority who reached surgery, many of them quick day cases. The first appointment figure describes everyone referred, including the larger group who are seen, investigated and sent on their way without an operation. Subtracting one from the other would produce a confident looking number that describes nobody.
So for those specialties we show the wait to a first appointment on its own, say what it measures, and do not pretend to know what follows it.
The figures on our home page
Three bars, three figures, all describing one patient: someone referred for orthopaedic treatment. We use orthopaedics because it is the specialty most people arrive here searching for, and because a chart that mixed specialties would be doing the thing this page argues against.
The NHS wait, 39 weeks. The point by which about two thirds of orthopaedic patients having planned surgery have been treated, measured from referral. Admitted completed pathways, acute trusts.
The 18 week mark on that bar is the standard the NHS is measured against, not a wait. It is there so you can see where the target sits inside the reality. About a third of orthopaedic patients are treated within it.
A faster NHS hospital, 16 weeks. The median wait at the ten fastest NHS trusts for orthopaedic surgery. We take the median of the ten fastest rather than the single fastest, so that no one hospital with an unusually high volume of quick day case work sets the figure.
Going private, 6 weeks. A private consultation lead of one to two weeks plus the scheduling window for the operation, averaged across the three kinds of orthopaedic surgery we cover. The range across those is three to ten weeks.
Each of the three has a written definition that reproduces the same way from the same file every month. They are illustrative of the gap rather than a prediction for anyone in particular, because at that point in the journey we do not know your postcode. Once you enter one, every figure we show you is your own hospital's.
The choices we make, including the unflattering ones
We do not understate. Where two honest measures disagree, we use the one that does not make the wait look shorter. A patient who plans for a longer wait and waits less is fine. A patient who plans for a shorter wait and waits longer has made a decision, sometimes a several thousand pound decision, on a false basis. This rule is why our headline figures are roughly double the ones in the news.
We say when we do not know. Where a figure cannot be computed honestly, we leave it out rather than estimating it. There are places on this site where a number is conspicuously absent. That is deliberate.
We removed our own estimates. An earlier version of this product estimated the front of the pathway as a fixed proportion of the whole. When we obtained the real per hospital figures, that assumption turned out to be wrong at 73 of 113 hospitals, and wrong in the same direction every time: it overstated how much paying privately would help. Those estimates are gone.
We exclude hospitals we cannot read. A small number of trusts do such a high volume of quick day case work that no percentile of their distribution describes major surgery. Rather than show a figure that flatters them, we leave them out of the comparison and say why.
We do not use averages of averages, or national figures dressed as local ones. Where we show you a national band because we do not hold a local one, we label it as national.
Our figures are two months old, and so is everyone's. NHS England publishes referral to treatment data roughly two months in arrears. The most recent month available is the most recent month we hold. My Planned Care is more current, updated weekly, and we refresh it daily.
The realised wait is backward looking. It describes people referred up to a year ago. If waits are rising it will understate what a patient referred today will experience; if they are falling it will overstate. Neither this figure nor the official one is a forecast. Nobody publishes one.
Sources
Everything on this site comes from published data. There are no surveys, no panels, no estimates dressed as measurements, and no data bought from anyone.
NHS England Referral to Treatment (RTT)
The monthly national collection of waiting times, published by provider and treatment function.
We ingest all three published series: incomplete pathways, admitted completed pathways and non admitted completed pathways. The admitted series drives every wait figure on the site. The others are held for comparison and for the analysis on this page.
Waits, means and percentiles are derived from the published week band histogram using linear interpolation within bands, with a fixed assumption for the open ended band above 104 weeks. Every derived median is checked against the median NHS England publishes for the same row, and the ingest fails if the two disagree by more than a tenth of a week. On the current month they agree to three decimal places.
Coverage: 118 acute trusts with confident orthopaedic data, 108 for general surgery, 107 for ENT. These counts are on the same basis as every wait figure we publish, meaning acute trusts with enough completed pathways in the month to report confidently. Reporting month is printed on every page that carries a figure.
NHS My Planned Care
The NHS's own patient facing service, publishing the wait to a first outpatient appointment by trust and by specialty, refreshed weekly from the Waiting List Minimum Data Set.
We hold 778 trust and specialty rows from 441 provider pages, covering orthopaedics, general surgery, ENT and spinal surgery. Of those, 750 carry a published figure and 28 are published by the trust as currently unavailable, which we record as a gap rather than filling in.
This is the source that lets us split a wait into getting seen and everything after. It is a different collection from RTT, using a mean rather than a median, so we never pool the two. It sits alongside our other figures, labelled.
NHS England Diagnostic Waiting Times (DM01)
Monthly diagnostic waits by provider across fifteen test types. We use it for the scan queues that sit inside a pathway: MRI, CT, ultrasound and others.
Current national typical waits are around two weeks for MRI, CT and ultrasound. This is why we do not claim that paying for a private scan saves a patient months. For most people right now, it does not.
Care Quality Commission
The provider register, used to identify and locate independent hospitals. 2,687 acute independent providers, geocoded.
NHS Organisation Data Service
Trust codes and locations. 248 located providers, of which 134 are classified acute. Only acute trusts appear in our comparisons, because a community trust's musculoskeletal triage wait is not a surgical wait and ranking one against the other would be misleading.
PHIN, the Private Healthcare Information Network
The statutory source for private consultant fees and patient satisfaction. Queried live, per postcode, for the paid plan.
Private surgery prices
Taken from providers' own published self pay price lists, hospital by hospital. We record the verbatim sentence describing what each price includes, alongside the URL and the date we read it. National "from" prices are never used to anchor a range, because they are marketing floors rather than prices anyone pays.
Where a provider's wording suggests something is excluded, we record the wording and do not classify it ourselves.
postcodes.io
Geocoding. Also the source of the country check that stops us showing English figures to patients in Scotland, Wales or Northern Ireland, whose systems publish differently and whose right to choose a hospital is not the same.
How the data stays current
The RTT and diagnostic pipelines run daily. My Planned Care runs daily against a source that updates weekly.
Each pipeline compares what NHS England currently offers against what we hold, and fails loudly if we fall more than one publication cycle behind. That check deliberately looks for available files in a different way from the importer that reads them, so that a renamed file cannot cause the importer to quietly skip a month while reporting success. It caught exactly that within minutes of being built: NHS England had renamed a diagnostics file and our importer had stopped seeing it.
There is also a public health endpoint reporting what we hold against what is available, monitored externally, so a pipeline that stops running cannot go unnoticed.
What we do not know
We do not know your diagnosis, and we do not ask for it in a way that would let us give you clinical advice. We are not clinicians and this is not medical advice.
We do not know how long you personally will wait. Nobody does. Every figure here describes a group of patients like you at a hospital like yours.
We do not hold procedure level waits, because they do not exist.
We do not hold data for Scotland, Wales or Northern Ireland.
We do not measure how long a private consultation takes to arrange. We use one to two weeks, based on what private providers publish about their own booking times, which is marketing rather than measurement. We use the slower end of what they claim. Across the range of NHS waits this makes a difference of at most a week, so we have not pursued it further.
Checking our workings
If you want to test any figure on this site, the underlying files are public. NHS England publishes RTT and DM01 monthly, My Planned Care publishes per trust, and every private price we quote is on a provider's own website with the date we read it recorded.
If you find something wrong, tell us and we will correct it and say what changed.