How to read NHS RTT data without losing your mind

If you've ever tried to read NHS England's monthly Referral-to-Treatment (RTT) statistics directly, you've probably given up. The CSV is 80MB. The columns are obscure. The headline numbers in news coverage rarely match what's actually in the data.

This piece is what I wish I'd had as a primer when I first started analysing RTT data.

What "RTT" actually measures

The Referral-to-Treatment (RTT) clock starts the moment your GP refers you for a consultation. It runs through:

• Outpatient appointment(s)
• Diagnostics
• Treatment decision
• The treatment itself (if recommended)

It stops when treatment is complete — or when the pathway is closed (e.g., you decline treatment, the issue resolves, you're transferred to another service).

The headline NHS standard is 92% of patients should start treatment within 18 weeks of the referral. That's been the benchmark since 2008. As of early 2026, the actual figure has been hovering around 60% — significantly off-target — but the geographic + specialty variation is what most coverage misses.

The three numbers worth tracking

Most NHS coverage focuses on one figure: the headline waiting list size (currently ~7.22 million, January 2026). That's the right number for a national headline, but it hides where the action actually is.

The three numbers that matter for any specific question:

1. Median wait (weeks)

Half of patients on the list wait longer than this; half wait shorter. Per trust, per specialty.

This is what HospitalWaits surfaces because it's the most patient-relevant number. If your local trust's median for trauma & orthopaedics is 22 weeks, you have a 50% chance of waiting at least 22 weeks.

What median doesn't tell you: the long tail. A trust can have a 14-week median but a long tail where 5% of patients wait 80+ weeks. The full distribution matters.

2. % within 18 weeks

The proportion of incomplete pathways currently inside the 18-week target. This is the metric NHS England reports against, and the one that drives political attention.

A trust at 90% within 18 weeks is performing close to standard. A trust at 50% is significantly off.

3. Total waiting (list size)

The number of patients currently on the list. Useful for understanding scale + capacity pressure, less useful for predicting any individual patient's wait.

What "January 2026" means in the data

NHS England releases RTT data with a 6-week lag. The "January 2026" release (published mid-February 2026) reports the position at the end of January.

So if you read "the latest RTT data shows…" in February news, that's actually the end-of-January position. Real time in your local trust may already be different.

This is structural, not editorial. NHS Trusts submit the data; NHS England aggregates and publishes. The lag is the price of comprehensive coverage.

Specialty granularity

The published data covers 18 specialties (the Treatment Function Codes used for elective care):

• Trauma & Orthopaedics
• General Surgery
• Cardiology
• ENT (Ear Nose Throat)
• Ophthalmology
• Gynaecology
• Urology
• Plastic Surgery
• Cardiothoracic Surgery
• Dermatology
• Gastroenterology
• Geriatric Medicine
• Neurology
• Neurosurgery
• Oral Surgery
• Other (a residual bucket)
• Rheumatology
• Thoracic Medicine

The variation by specialty is enormous. Trauma & Orthopaedics + Ophthalmology together account for the largest share of long waits — they're high-volume elective specialties under capacity pressure.

If you're writing about NHS waits, treating "the wait" as one number across all specialties is misleading. Always break it down.

Trust-level variation is the real story

The same specialty can have median waits varying by 30+ weeks across trusts. Reasons:

• Local capacity (operating theatre time, bed availability)
• Workforce (vacancy rates among consultants)
• Catchment population (deprivation, age, demand)
• Historical investment patterns
• Recent industrial action / strike effects
• Referral patterns from local GPs

A trust at 9 weeks for cataract surgery and a trust 20 miles away at 38 weeks isn't unusual. This is exactly the variation Right to Choose lets patients exploit.

What the data doesn't tell you

The published RTT numbers are aggregates. They don't tell you:

• The identity of the consultant who'd see you
• Day-of-week or time-of-day variation
• Capacity for last-minute appointments
• Quality of care received
• Outcomes (separate dataset — see PHIN for private; NHS Outcomes Framework for NHS)

Don't over-interpret. The data tells you about queue length, not about the quality of the eventual treatment.

Common mistakes I see in news coverage

1. Treating list size as a wait. "7.22 million people are waiting" doesn't mean each is waiting forever. The list churns — patients enter, get treated, leave. Year-on-year growth in list size matters more than absolute number.
2. Comparing across years without inflation-adjusting for population growth or referral pattern changes. UK population grew. Referral thresholds changed. Compare apples to apples.
3. Quoting "the longest waits" without specifying specialty. "Some patients wait 2 years" — yes, in a specific specialty at a specific trust. Not the typical experience.
4. Confusing the 18-week standard with the 52-week standard. The 18-week standard is the formal target (92%). The 52-week standard is a separate hard limit (now legally a breach). Different things.
5. Conflating elective and urgent waits. RTT covers elective. The 2-week wait for suspected cancer is a separate, faster pathway. Don't mix.

How HospitalWaits processes the data

For full reproducibility, see our methodology page. Briefly:

• We ingest the full monthly CSV from NHS England the day after publication
• We compute median wait per trust × specialty using the published weekly band distribution (D-010 — see decisions log)
• We match Trust ODS codes to geographic location for postcode-distance queries
• We compute 12-month trends for each trust × specialty
• We expose the Doctor Data Wait Score™ — a 1-10 score normalising trust median against national median for the specialty

Every figure on the site links to its NHS source URL. Every calculation is documented.

Where to find the raw data

• NHS England RTT statistics page
• NHS Organisation Data Service (for trust metadata)
• The full CSV is published as "RTT [Month] [Year] full extract" — typically ~80MB

If you're a journalist, ICB analyst, or just curious, the raw data is open. The barrier is interpretation — which is what this piece + our methodology page exist to lower.

Disclaimers

This is general data-analysis information, not advice on individual NHS care decisions. NHS England is the canonical source; HospitalWaits processes their data but does not modify it. Every figure on this site is reproducible from the published CSVs.