Mother and Baby Identification Controls in Maternity Units

Standard practice in most well-run maternity units is two bands on the newborn, applied before the baby leaves the delivery area, plus a band on the mother carrying the matching identifier. Two on the baby is not excess caution; neonatal limbs are small, bands loosen, and a single band that comes off leaves an unidentified infant. Where a band must be removed for a procedure or a cannula, the policy should require a replacement to be applied before the original is cut off, and the swap should be recorded.

The content matters as much as the count. A band carrying only a name is inadequate in a unit where two women share a surname, which happens more often than intuition suggests. Carry a unique identifier that ties to the hospital record, along with mother identifiers and the delivery details, and print it rather than handwriting it. Handwritten bands smudge, and a smudged digit is the beginning of a near miss.

Decide in advance what happens when there is no name yet. Newborns are frequently unnamed for days, and the naming convention you adopt has to be unambiguous, machine-readable and stable across multiple births. The convention should also handle multiples explicitly, with the birth order recorded on the band, because twins in the same nursery are the single highest-risk identification scenario a maternity unit routinely faces.

The verification itself should be a spoken, two-person act rather than a glance. The person handing over reads the identifier aloud from the baby's band; the person receiving reads it from the mother's band or from the record and confirms. It takes a few seconds and it catches the specific error that pure visual checking never catches, which is the assumption that the baby you are holding is the baby you think you are holding.

Feeds are the highest-frequency verification point and the one most often skipped. In a rooming-in ward where the mother feeds her own baby, the risk is low and the check is quick. In any situation where staff bring a baby to a mother, or where expressed milk is administered, a full match check is warranted every single time. This is one of the few places where insisting on a check at every occurrence, however repetitive, is genuinely proportionate.

Record the verification where it is operationally cheap to do so and accept that not every check can be recorded. A scan of both bands at a feed or a transfer creates a timestamped matched pair with almost no extra effort, and HealUDoc activity logs can hold that trail against the encounter for both mother and baby. A written signature per verbal check, on the other hand, will be back-filled within a month and is worse than no record at all.

Almost every reported newborn identification near miss involves a transfer. A baby moves from labour room to nursery, or from nursery to a special care unit, or from a special care unit back to the postnatal ward, and somewhere in that movement the paperwork and the infant travel separately. The receiving unit accepts the baby, creates or opens a record, and a mismatch that will only be discovered later has been created.

The fix is procedural and unpopular: nothing moves without a completed transfer record, and the receiving nurse performs the identity check before accepting custody rather than after settling the baby. Building a hard stop into the system, so a receiving unit cannot open a newborn record without confirming the matched identifier, converts a policy into a control. It will occasionally be inconvenient at three in the morning, which is precisely when the control is doing its job.

Pay particular attention to transfers out of the facility. A neonate referred to another hospital leaves your identification system entirely, and the referral documentation is the only thing that carries identity with them. Make the referral note carry the mother's identifiers, the delivery record reference and a contact number, and keep a copy that ties to both records. This overlaps with your wider neonatal unit documentation and should be designed alongside it rather than separately.

Handling expressed breast milk turns identification into something closer to a laboratory specimen problem. Milk is expressed, labelled, stored, sometimes fortified, retrieved and administered, often hours later and by someone who was not present at expression. Each of those steps is a place where a container can be labelled wrongly or picked up wrongly, and the consequence of a mismatch is an infant receiving another mother's milk, which is a reportable adverse event with real clinical and family implications.

Treat the container the way a laboratory treats a sample: labelled at the point of expression, never before, with the mother identifier, the infant identifier, the date and time of expression, and any additive recorded. Storage location should be assigned rather than chosen, and retrieval should be verified against the infant's identifiers at the cot side rather than at the refrigerator. The gap between retrieval and administration is where most errors are introduced.

If your unit operates or draws from a lactation management centre or a human milk bank, the traceability requirements are stricter and follow the operational guidelines for those facilities. Donor milk needs a batch identity, a pasteurisation record and a link from batch to recipient that survives audit. That is a documented supply chain, and it should be designed with your quality team rather than improvised by the nursery.

Identification incidents are almost never reported as identification incidents. They surface as a band found on the floor, a baby brought to the wrong bay and returned, a milk container with an unreadable label, a transfer note that arrived late. Staff classify these as minor annoyances rather than as safety events, and so the unit never sees the pattern that precedes an actual mismatch. Naming these explicitly in your incident categories is the single most useful reporting change a maternity unit can make.

Make reporting them cost thirty seconds and produce no blame. A short form with a category, a free-text line and an optional name, reviewed weekly by the unit in-charge, will surface more useful data than a formal incident process that requires a manager's countersignature. The measure of success in year one is that reports go up, and this needs to be explained to management before the programme starts or the rising numbers will be read as deteriorating safety.

Review the reports as a group rather than case by case. Ten reports of bands coming loose point at a procurement decision, not at ten individual lapses. Six reports of babies arriving before transfer notes point at a workflow gap between two units. The analytic value is in the aggregate, and it is entirely lost if each report is closed individually with a note saying the staff member was counselled.

The last check is the one that catches everything the chain missed, and it should be performed by someone who was not involved in the delivery or the ward care. At discharge, the mother's band and both of the baby's bands are read against the record together, in the presence of the mother, and only then are the bands removed and the gate pass issued. Removing bands at the ward and issuing the pass at the desk breaks the chain in its final metre.

Physical control matters alongside the record. Most maternity units in India operate an attendant pass system, restrict nursery access and require a named person for infant movement. Whether or not you use electronic tagging, decide who is permitted to carry a newborn within the facility and write it down. Units that permit family members to carry infants between wards have effectively removed the transfer control described earlier, whatever their policy says on paper.

Audit the discharge check the way you would audit anything else: sample the records monthly, confirm that the check is timestamped and attributed, and reconcile the count of discharged newborns against the count of deliveries and admissions for the month. That reconciliation is dull, it takes an hour, and it is the only thing that will tell you whether a newborn record exists that never received a discharge event.