How the 18-week RTT clock works: starts, stops and new clocks

MedX Services · 13 October 2026 · 7 min read

Almost every elective care decision in an English hospital is measured against one number: 18 weeks. Under the NHS Constitution, patients have the right to start consultant-led treatment within a maximum of 18 weeks from referral, unless they choose to wait longer or it is clinically appropriate to do so. The operational standard is that 92% of patients waiting should have been waiting for less than 18 weeks.

Behind that standard sits a detailed set of national rules about when a patient's waiting time "clock" starts and stops. Getting those rules right is the core of pathway coordination and validation work. Getting them wrong produces waiting lists that misstate how long patients have actually waited.

When the clock starts

A referral to treatment (RTT) clock starts when a permitted professional refers a patient to a consultant-led service, or to an interface or referral management service that may pass the patient on to one. The start date is the date the provider receives the referral. For referrals booked through the NHS e-Referral Service, it is the date the patient converts their unique booking reference number (UBRN).

Direct-access referrals to services that are not consultant-led, such as some physiotherapy or audiology services, do not start an RTT clock. Self-referrals start one only where local commissioners and providers have agreed that pathway.

What stops the clock

Treatment

The clock stops when first definitive treatment starts. The rules define this as "an intervention intended to manage a patient's disease, condition or injury and avoid further intervention". Admission for a day case or inpatient procedure stops the clock. An outpatient appointment does not, unless treatment is actually given at it.

Non-treatment decisions

A clock can also stop without treatment, once the decision has been communicated to the patient (and then to their GP or referrer without undue delay):

• Active monitoring, where the clinician decides monitoring is the most appropriate option, or the patient chooses it.
• Decision not to treat, made by the clinician.
• Patient declines treatment after it has been offered. Simply wanting to delay treatment does not count.
• Return to primary care for treatment that is not consultant-led.

Did not attend (DNA) rules

The rules treat a missed first appointment differently from a missed later one:

• First appointment DNA: the clock is nullified, meaning it is removed from waiting time measurement. This applies only if the provider can show the appointment was clearly communicated to the patient. If a new appointment is booked, a new clock starts on the date the patient agrees it.
• Later DNA: the clock is not nullified. It stops only if the patient is discharged back to their GP, the discharge is not against their clinical interests, and it follows a published, clinically agreed local policy that protects vulnerable patients such as children. If the provider keeps clinical responsibility, the clock keeps running.

A patient who cancels before an appointment, however late, has not DNA'd.

There are no clock pauses

Since 1 October 2015, there has been no provision to pause or suspend an RTT clock under any circumstances. Waits caused by patient choice are managed through the rules on declined treatment, active monitoring and patient-initiated delay, not by stopping the clock temporarily.

When a new clock starts

A new clock starts for a substantively new or different treatment not already in the agreed care plan, for the second side of a bilateral procedure once the patient is fit and ready, on a new referral, and when a decision to treat is made after a period of active monitoring.

Where validation usually finds errors

From day-to-day validation work, the same patterns recur. These are the checks worth building into practice:

1. Clocks stopped at outpatient attendance where no definitive treatment was given and no non-treatment decision was recorded.
2. Active monitoring recorded without the decision being communicated to the patient. The clock should stop on the date it is communicated, not the date the clinic took place.
3. First-appointment DNAs nullified without evidence that the appointment was clearly communicated.
4. Duplicate open pathways after a re-referral or a transfer between specialties.
5. Missed new clock starts after active monitoring ends with a decision to treat.

Practising these decisions on fictional pathways, before working on a live waiting list, is exactly what the MedX Simulator and our RTT & Waiting List Management training are designed for.