Fertility Patient Follow-Up Automation
Outreach sequences your clinic writes and versions (text, call, wait, a consent question, a booking offer) running until the patient engages or opts out, with every touch on one record.
Follow-up in a fertility clinic fails for a structural reason rather than a technical one, and it is worth being precise about which parts of it should be automated and which should never be.
Where fertility follow-up actually breaks
Almost never at the first contact. Almost always at the third, when nobody is sure whether anyone tried.
A referral arrives and somebody calls. The patient does not pick up. Somebody leaves a voicemail, writes a note, and the note lives in whichever system that person works in. Two days later a different person looks at the same referral and cannot tell whether it has been chased, so either it gets chased twice or it does not get chased at all. Multiply that by a fax machine that never stops and the failure is not a missing feature, it is a missing shared record of what has been tried. The fix is less about automation than about every touch across every channel landing on one record where the next person can see it.
How a sequence works
Your clinic writes the sequence and versions it. The platform runs it and fixes what the answers mean.
You author the steps
Not a built-in cadence everyone shares
A sequence is an ordered list of steps your clinic writes, versions, and turns on or off. It is not a single cadence shipped to every clinic, which matters because the right number of attempts before backing off is a clinical and cultural judgement, not a default.
It starts from a confirmed referral
One trigger, deliberately
Today a sequence is started by a confirmed referral. Naming the one trigger is more useful than implying there are many: a clinic can plan around one it can see.
Steps: text, call, wait
The ordinary ones
A step sends an SMS, places an automated call, or pauses for a set time before the next one. Every system-sent text is logged automatically against the patient's record, so the timeline shows what went out and when.
Question steps park and wait
The important difference
A consent question or a preference question does not move on a timer. After sending, the sequence parks until the patient answers. That is the behavior that stops a clinic texting somebody who has already said stop.
The booking offer ends it
One terminal step
A booking offer invites the patient to book, and always ends the sequence. Booking is the moment a prospect becomes an active patient and the one point where outreach reaches into the clinical record.
You write the words, not the meanings
A guardrail worth knowing about
You can reword what a question step sends. What the answers mean (which reply selects text, email, phone, or no longer wishing to proceed) is fixed by the platform, so a reworded menu cannot accidentally change what a patient's reply is taken to say.
What has to stay human, and what consent does to the design
Automation that cannot hear a patient asking it to stop is not a feature, it is a liability.
Consent is not a checkbox on the way in; it is a state the sequence has to respect while it runs. That is why the question steps park rather than continue, and why an opt-out ends the sequence rather than pausing it. Everything with clinical content in it stays human: a result, a protocol change, bad news, and any reply where the patient is distressed. What automation is genuinely good at is the part that is currently done badly because it is tedious: the second and third attempt to reach somebody, the reminder, the confirmation that a message was read rather than merely delivered. A clinic that automates the first of those and leaves the second to people has it exactly backwards.
Questions clinics ask about follow-up automation
The ones that decide whether this is safe to turn on.
What can be automated, and what should not be?
Reaching out to a referred patient, reminding them, asking how they would like to be contacted, and offering a booking are all reasonable to automate under consent. Anything carrying clinical content (a result, a protocol change, bad news) should not be. Neither should any reply where a patient is distressed; those need a person, quickly.
How does a patient opt out?
A preference question includes an option for no longer wishing to proceed, and what that reply means is fixed by the platform rather than by the wording a clinic chooses. A sequence that receives it ends rather than pausing.
Does it text people who never consented?
Consent is captured as part of the sequence itself, through a consent step that parks and waits for the patient's reply rather than moving on a timer. The channel a patient chooses is recorded on their record and applies to what follows.
Can we see what was sent?
Yes. Every touch across every channel lands on one prospect record as a timeline, with system-sent texts logged automatically, and it shows when a patient read a secure message, so an unanswered message can be told apart from an unopened one.
Can we run different sequences for different situations?
Sequences are enabled independently of one another, so you can build and test one while another is live. Each is authored and versioned by your clinic.
What starts a sequence?
A confirmed referral, today. If your clinic wants a different trigger, say so in the demo. It is a reasonable thing to ask for and a better conversation to have before you commit than after.
See a sequence run end to end
Including the part where a patient says stop.
Book a demo and we will run a sequence against a synthetic patient, including the consent step, the park-and-wait behavior and the opt-out. It is a short demo and it is the one that tells you whether this is safe for your clinic.