At 9 p.m., your practice's phone rings. On the line: a post-op cataract patient with worsening pain, a parent whose toddler scratched his eye, a contact lens wearer whose eye went red this afternoon — or, statistically most often, someone who wants to book an exam and is calling when their own workday allows. After-hours coverage is the plan for all four, and most eye care practices assembled theirs by accident.
What actually calls at night
Sort the after-hours traffic honestly and it falls into three buckets. A small but critical stream of true urgencies: post-op complications, sudden vision changes, flashes and floaters, chemical exposure, trauma. A modest stream of clinical-but-waitable: the mild red eye, the lost contact lens, the broken glasses. And the largest stream by volume: administrative calls — booking, rescheduling, order questions — from people whose schedules don't match yours. Each bucket wants a different response, which is why a single voicemail greeting serves all three badly.
The non-negotiable: an urgent-symptom pathway
Whatever else you build, the urgent pathway comes first, because the stakes are clinical. The elements:
- A written triage protocol listing the red-flag symptoms — post-op pain or vision change, new flashes/floaters or curtain, sudden vision loss, chemical injury, trauma — and exactly what happens for each: page the on-call doctor, direct to emergency care, or both.
- A reachable human or reliable relay. Whether that's an answering service with paging instructions, a call-forward to the on-call OD, or a monitored line, the chain from symptom to doctor must work at 2 a.m., and it should be tested — actually called — quarterly.
- Voicemail that triages by itself. If voicemail is part of your stack, its greeting should lead with the urgent instructions ("if you are experiencing sudden vision loss…"), not your fax number.
Coverage options for the rest
Voicemail plus next-morning discipline. Legitimate for small practices — if someone actually works the voicemail queue first thing, every day. The failure mode isn't the voicemail; it's the Tuesday it doesn't get returned.
Traditional answering service. Solid for urgent relay, weak for everything else — messages, not resolutions. Priced per call, sensible as a safety layer.
AI phone agents. Increasingly reasonable for the administrative bucket at night: booking a routine exam at 9 p.m. is exactly the bounded task the technology handles. Keep it away from symptom triage — route anything clinical to the urgent pathway rather than letting software reassure someone about a retinal detachment.
Extended-hours human coverage. The premium tier: a virtual receptionist covering evening hours live — useful for practices in competitive markets where the 7 p.m. caller books with whoever answers. Most practices don't need it; practices fighting for new-patient volume sometimes find it pays for itself in captured bookings alone.
The stack most practices should run
A pragmatic combination: triage-first voicemail with an urgent relay that's tested quarterly; an AI or service layer for after-hours booking if volume justifies it; and — the piece most practices miss — a next-morning owner who works the overnight queue before the first patient arrives. That last role folds naturally into a virtual assistant's morning routine: voicemails returned, online bookings verified, urgent follow-ups confirmed with the doctor, all before 9 a.m. After-hours coverage, it turns out, is mostly won the next morning.
One closing audit worth doing this week: call your own practice tonight and listen to what your patients hear. If the greeting buries the urgent instructions, promises a callback that has no owner, or simply beeps — you now know exactly what to fix, and it costs almost nothing to fix it.
.png)
.png)
.png)
