IVF Cycle Workflow and Cryo Inventory: The Software Record

From the baseline scan onward, the record is a dated series of the same handful of measurements: follicle counts and diameters by ovary, endometrial thickness and pattern, and the hormone values from that morning's sample. The clinician then makes one decision per visit, which is continue, adjust dose, add antagonist, trigger, or cancel. Recording the decision alongside the numbers that produced it is what makes the chart readable to the consultant who covers on Sunday.

Follicle measurements should be structured, not typed into a note. Once they are structured you get the growth curve for free, the trigger decision has an audit trail, and a cancelled cycle can be reviewed properly at the clinical meeting instead of relitigated from memory. The cost is genuine: entering fifteen follicle diameters through a form is slower than writing them, and sonologists resent it. Mitigate with a grid entry screen designed for the scan room rather than a generic clinical form, and accept a short adoption dip.

The drug record is where cycles go wrong quietly. Gonadotrophin doses change mid-cycle, patients self-administer at home, and pharmacies dispense in vials that do not match the prescribed units. The record should carry prescribed dose, dispensed quantity, and administered dose as separate facts, because when a response is poor the first question is always whether the patient actually took what you think she took. A single free-text line saying the dose was increased answers none of it.

Retrieval day compresses several identity-critical steps into a couple of hours, and it is the first point where a mix-up becomes irreversible. The patient is identified in the theatre, the follicular fluid goes through a hatch to the embryology laboratory, the partner produces a semen sample somewhere else in the building, and the two are eventually brought together. Every one of those handovers needs a recorded identity check by two people, with the time and both names captured.

The trigger-to-retrieval interval should be a computed field, not a remembered one. Record the exact trigger time as administered, and the theatre record then shows the actual interval rather than the intended one. When retrieval lists slip, and they always slip, the interval is the first thing a review will ask about, and reconstructing it from a nursing note and an OT register is unreliable.

The andrology record deserves the same rigour as the oocyte record and rarely gets it. Sample production time, method, abstinence period, the container identifier, the preparation method, and the post-preparation count and motility all belong in the structured record, linked to the same cycle. Where donor semen is used, the bank's consignment identifier and screening documentation attach at this point, not later. Doing it later means doing it from memory.

The bench record is a sequence of timed observations on identified dishes: oocyte count and maturity at denudation, insemination or injection with the method and time, the fertilisation check the next morning, cleavage assessment, blastocyst grading, and the disposition of each embryo. The unit of record is the individual embryo, not the cycle, because a cycle may end with one embryo transferred and five frozen and each of those six needs its own history.

Grading schemes must be configured, not free-typed. Whether the clinic uses Gardner grading, a consensus alphanumeric scheme, or a time-lapse derived score, the vocabulary should be fixed so that two embryologists mean the same thing by the same string. Free text here destroys any possibility of comparing outcomes by grade, which is the single most useful internal analysis an embryology laboratory can run on itself.

Where a time-lapse incubator is in use, decide early whether the annotations live in the incubator's own software or in the clinic record. Both is worse than either, because they will diverge. If the device software is authoritative, the clinic record should carry the derived decision and a reference to the device record, and the interface between them should be documented as part of your equipment records rather than left as one embryologist's personal practice.

Witnessing means that at every step where gametes or embryos move between labelled vessels, a second competent person independently verifies patient identity on both source and destination before the move happens. Denudation, insemination, dish change, biopsy, vitrification, warming, and loading the transfer catheter are all witness points. The number of these in a single busy morning is larger than most clinics expect, which is why manual witnessing degrades under load precisely when it is most needed.

Electronic witnessing using RFID or barcode tags on every dish and straw removes the human bottleneck and produces a timestamped log, but it introduces its own failure modes. Tags fail, readers lose calibration, and staff learn to defeat an alarm they consider a nuisance. The correct posture is that electronic witnessing replaces the second pair of eyes for verification but not the culture that says an unwitnessed step does not happen. Buying the system and not changing the culture buys you an expensive log of the same errors.

Whichever route you take, the witnessing log must be queryable by cycle, by embryologist, and by date, and it must be retained for the same period as the clinical record. A log that exists only inside a device and cannot be exported is not evidence. HealUDoc activity logs can carry the witnessing entries alongside the cycle record so that the trail and the clinical narrative are retrievable together rather than from two systems with different retention rules.

Cryo storage is a warehouse with a legal obligation, and it should be modelled like one. Every stored item needs a unique identifier and a full address: tank, canister, cane or goblet, and position. Every movement of that item, including a temporary lift for another patient's straw, should generate a transaction with a timestamp and an operator. Clinics that model only the tank and rely on a paper map taped to the wall discover the gap the first time a cane is returned to the wrong canister.

The inventory must also carry the commercial and consent state of each item, because storage is not free and consent is not permanent. For each stored straw or device you need the owning patient or couple, the consent expiry, the paid-up storage period, the disposition instruction on expiry, and the current status. A quarterly reconciliation should list items whose consent has lapsed, whose storage fees are unpaid, or whose owners have not responded to contact attempts. That list is uncomfortable and it is exactly why it needs to be generated automatically rather than requested.

Physical vessel management belongs in the same system. Liquid nitrogen level checks, top-up records, alarm test records, temperature logging, and the vessel service history are equipment records that an assessor will ask for, and they are much easier to produce when they sit against the same tank object that holds the inventory. The trade-off is that this makes the embryology laboratory responsible for maintaining an asset register, which is real work nobody has been allocated.

A cycle is not closed at transfer. It closes at the beta hCG, the viability scan, and eventually the pregnancy outcome, and a clinic that stops recording at discharge has no denominator-correct success rate to quote to the next patient. Build the follow-up as scheduled tasks that fall due automatically after transfer, assign them to a named coordinator, and track the proportion of cycles with an unknown outcome as a quality measure in its own right. Above a few per cent, every success figure you publish is suspect.

The traceability test is the one to rehearse. Pick a stored straw at random and ask the system to tell you whose it is, which cycle it came from, which donor gametes if any contributed, who performed each bench step, who witnessed each transfer, what consent covers it today, and when that consent expires. Then run it in reverse: pick a patient and ask where everything belonging to her currently sits. Both directions should answer in one screen. Most clinics can do the first and not the second.

None of this is achievable by buying software alone, and it is worth saying so plainly. The record only holds if the embryologist enters the fertilisation check at the microscope rather than at four o'clock from a notepad, and if the coordinator chases the outcome that nobody is paying for. Budget the coordinator role explicitly. The traceability that the ART framework expects is a staffing decision wearing a software costume.