The Fourth Analyte: Time Assurance for UK Pathology Systems

A pathology result is rarely just a number. It is a number attached to a patient, a sample, an analyser and a moment. The first three are usually governed by explicit controls: positive patient identification, sample acceptance criteria, analyser maintenance records. The fourth, the moment, is often assumed. That assumption is becoming harder to defend.

Why a timestamp is not mere metadata

In a modern UK laboratory, the timestamp is part of the diagnostic record. It determines whether a result falls inside a clinically useful turnaround window. It sequences samples for delta checks. It anchors quality control events, maintenance logs and incident chronologies. In transfusion science and acute medicine, a disputed timestamp can change the story of what was known, and when.

Middleware adds another layer. Rules that flag a result as materially different from a previous value depend on the order of those results. If an analyser clock drifts backwards, a later sample can appear earlier than its predecessor. The arithmetic may still run, but the clinical meaning may not. The same applies to audit trails: an entry that cannot be placed reliably in sequence is weaker evidence than it appears.

Where time goes wrong

Clock error is rarely dramatic. It is usually mundane. An analyser oscillator drifts by seconds over months. A network time protocol client fails silently after a switch upgrade. A virtual machine inherits the wrong time zone from a template. A middleware server is set to local time while the laboratory information management system stores UTC, and a report displays one without the other.

British Summer Time introduces predictable ambiguity. When clocks go back in October, the hour from 01:00 to 02:00 occurs twice. When they go forward in March, an hour disappears. Systems that store local timestamps without an offset cannot distinguish the first 01:30 from the second. Systems that convert incorrectly can place samples in the wrong order or generate apparent turnaround breaches that never happened. The problem is not the change itself; it is the absence of a tested convention for handling it.

Manual edits are another risk. An operator may correct a clock to match a wall clock, or a vendor engineer may adjust a server during maintenance. If the change is not logged with a reason, authorised by a second person and reflected in the audit trail, the laboratory loses the ability to reconstruct its own timeline.

The assurance question

Timestamp integrity sits awkwardly between disciplines. It is partly an IT issue, partly an instrument issue, partly a quality issue. That is precisely why it deserves an explicit place in the quality management system rather than being left to the good intentions of individual sections.

The assurance question is simple: can the laboratory demonstrate that every system contributing to a result record agrees on a common time reference, and that any deviation is known, bounded and documented? That question may arise during a UKAS assessment against ISO 15189, an MHRA inspection connected to an IVD, or an internal serious incident review. In each case, the answer should not depend on a single member of staff remembering how a particular analyser was configured.

A practical routine

A time assurance routine need not be elaborate, but it should be repeatable and evidenced.

Inventory the time sources. List every analyser, middleware server, LIMS interface, virtual machine and network device that contributes a timestamp to a result or audit record. Record its time source, time zone setting and whether it displays local time, UTC or both.

Define the hierarchy. Identify the authoritative time source for the laboratory and the expected synchronisation path. Document what happens when that source is unavailable. A system that quietly falls back to its own clock is not synchronised; it is merely waiting.

Measure drift. Compare each device with a reference at defined intervals. Record the offset and set a threshold for action. A few seconds may be acceptable for some purposes; it may not be for others. The threshold should be justified, not inherited.

Test the transitions. Before each British Summer Time change, verify in a non-production environment how systems handle the duplicate and missing hours. Confirm that stored timestamps include an offset or are held in UTC, and that displayed times are converted consistently.

Control manual changes. Restrict who can alter system clocks. Require a reason, a second check and an audit entry. Treat an unlogged clock change as a data integrity event, not a housekeeping task.

Monitor and review. Use automated alerts where available. Include time synchronisation status in routine quality review. When an incident occurs, reconstruct a short timeline across analyser, middleware and LIMS as a drill. If the sequence cannot be established, the routine has found a gap before an inspector does.

Making time visible

The strongest control is cultural as much as technical. Staff should know which clock is authoritative, how to report a suspected discrepancy and why a manual adjustment is not a neutral act. Suppliers should be asked to confirm time-handling behaviour, including daylight-saving transitions and audit-trail treatment, as part of procurement and change control.

None of this makes pathology more exciting. It makes it more defensible. The instrument clock is not a background setting. It is a variable that shapes the record, and in a quality system that already treats reagents, calibration and competence as controlled inputs, time deserves the same discipline.