Obstetric Triage and Maternal Early Warning Documentation

Triage is a sequence, and the sequence should be the same every time regardless of how busy the unit is. A maternity nurse performs an initial assessment covering the presenting complaint, gestational age, and the observation set your protocol specifies. That produces a priority category, which determines how quickly a clinician sees the woman. The category, the time it was assigned, and the person who assigned it all belong in the record as discrete fields, not buried inside a note.

Gestational age deserves particular attention because it drives almost every subsequent decision and it is frequently entered wrongly. Capture it as a computed value from a recorded last menstrual period or a dating scan wherever possible, with the source noted, rather than as a number someone typed. An unbooked arrival with no records is a common scenario and the record should be able to say clearly that gestation is estimated and on what basis.

The triage record should also capture what did not happen. If the woman was assessed but declined admission, or left before review, or was referred elsewhere, that outcome needs to be a recorded disposition with a time and a reason. Units that only document admitted patients lose sight of an entire category of risk, and it is precisely the category that generates complaints and medico-legal correspondence.

Maternal early warning tools such as the modified early obstetric warning system exist because physiological changes in pregnancy make general adult early warning scores unreliable in this population. Which tool you use, which parameters it contains and what each trigger means are clinical decisions for your obstetric department, informed by current national and professional guidance. This article is about the records around the score, not about the score itself, and that distinction matters more here than almost anywhere else in hospital documentation.

From a systems point of view, the score should compute rather than be entered. A nurse enters the observations she has taken; the system derives the score and displays it. Entering a score directly invites arithmetic errors and, worse, invites a score entered without the underlying observations. Where the tool involves a colour-coded chart, reproduce the colour logic in the interface, because the visual pattern is a large part of why these charts work at all.

Keep the score attached to the observation set that produced it, with its own timestamp, so a later reviewer can see exactly which values generated which score. If observations were taken but one parameter could not be obtained, the record needs a way to say so explicitly rather than leaving a blank that a reader will interpret as normal. That single design detail prevents a good deal of retrospective confusion.

An escalation pathway is only real if it names a role, a contact method and a time expectation, and if the person in that role knows they are in it. Most maternity units have a pathway that reads well on paper and dissolves at two in the morning because the registrar is in theatre and the consultant on call is thirty minutes away. Write the pathway for the worst-staffed shift of the week, not the best, and test it against the roster.

In the record, escalation should be an event with structure: the trigger, the time, the role contacted, the contact method, the time of acknowledgement, and the time of attendance. Acknowledgement and attendance are separate and both matter. A registrar who answers the phone in ninety seconds and arrives twenty minutes later has produced two very different intervals, and collapsing them into one number hides the interval you can actually improve.

There is a cost to this that nobody mentions at the design stage. Every escalation event you make recordable is an event a nurse must record while managing a deteriorating patient. Keep the capture to a handful of taps, allow the detail to be completed afterwards with an honest entry time, and accept that during a genuine emergency the clinical care comes first and the record is completed in the debrief. A documentation standard that cannot survive an emergency is not a standard.

The interval worth measuring is from the observation that met the trigger to the arrival of the responder, and it is measurable only if both ends are timestamped independently. Where the trigger time is derived automatically from an entered observation and the arrival time is entered by the responder, the measurement is reasonably clean. Where both are entered by the same person afterwards, the number is a self-report and should be treated as one.

Report the distribution rather than the average. A mean response interval of eleven minutes conceals the three cases that took fifty. Use the median for the routine picture and look explicitly at the tail, because the tail is where harm lives. Reviewing the five longest intervals each month, individually and with the people involved, teaches you more than any dashboard.

Watch for the gaming pattern, which is universal and rarely malicious. When a response interval becomes a reported metric, people start recording arrival at the time they set off, or entering the trigger observation slightly late so the clock starts later. The counter is not surveillance but framing. Present the metric as a systems measure owned by the department rather than as an individual performance score, and pair it with a mechanism to record why a long interval happened.

The commonest failure is the reassessment gap. A woman is triaged, given a category, and then waits. Nobody reassesses because the initial category was low, and the record has no mechanism to require a second look. Build a reassessment interval into each priority category and make it visible in the same way an overdue observation is visible. This is the single change that most reliably catches deterioration in a waiting area.

The second is the postnatal blind spot. Triage processes are written around women in labour, and a woman who delivered ten days ago and returns unwell may be routed to general emergency, assessed by a team without obstetric input, and admitted under a non-obstetric speciality. Define a postnatal window in your routing rule and make sure the emergency department knows it exists. Discuss the window with your obstetric team and align it with prevailing national guidance.

The third is the handover into the labour room record. If triage runs in one system and the labour record in another, the observations taken at triage are re-entered or lost, and the timeline breaks at exactly the point where continuity matters. Keeping triage assessment, early warning scores and the intrapartum record in a single patient timeline is unglamorous integration work that pays back the first time a case is reviewed.

Once triage is structured, it becomes one of the more informative datasets a maternity unit owns. Arrivals by hour of day and day of week tell you whether your night staffing matches your night workload, which is a question most units answer by impression. Presenting complaint distribution tells you what your assessment room actually needs to be equipped for. Disposition mix tells you how many arrivals genuinely required admission, which is the honest denominator for any capacity conversation.

Use it carefully in the departmental meeting. The temptation is to present triage volumes as a workload argument and stop there. The more useful analysis pairs volume with intervals and outcomes: at which hours does the time from arrival to first assessment lengthen, and does anything follow from that. If nothing follows, say so; a measure that never changes a decision should be retired rather than reported indefinitely.

Keep the retention and access question in view as well. Triage records contain health data about people who may never have become patients of the hospital, including women who left before review. Under the DPDP Act 2023 that is still personal data you are accountable for, and it needs the same retention schedule, access control and audit trail as any other clinical record. Deciding that early is considerably easier than retrofitting it later.