The 47% No-Show Problem: Why Ortho, Ophthal, and Derm Need Different Automation Approaches

A 47% no-show rate is the worst tail in the data. The median across surgical-specialty outpatient visits sits closer to 18%. The driver is not consistent across specialties, and that is why a single automation pattern fails on two-thirds of the practices that buy it.

I should be precise about those numbers. The 47% figure is the worst-tail value I have personally seen in a South Florida specialty practice's own scheduling data; I am not citing it from a published benchmark. The 18% median is consistent with the directional ranges MGMA's practice operations cuts have shown for surgical specialties, but I will not pretend to a single canonical figure here. Treat the headline as the operator-grade observation it is, not as a peer-reviewed point estimate.

I have come to believe that the no-show conversation in specialty practice is broken at the diagnostic level. Vendors pitch reminders. Operators buy reminders. The rate does not move, or it moves for thirty days and then regresses.

There is a fair counterargument I want to address before going further. Published reminder studies (Hasvold and Wootton's meta-analysis, multiple JAMA pieces on SMS confirmation) document 15 to 30% no-show reductions from generic reminder systems. That is real lift at near-zero implementation cost, and any honest post on this topic has to concede it. The argument I am making is not that reminders do nothing. It is that they capture a floor effect (the patients who simply forgot) and then plateau. The remaining no-show volume in specialty practice is not a memory problem, and the next 5 to 10 percentage points of improvement come from workflow engineering tied to the actual driver, not from a better reminder.

What follows is a workflow-level breakdown of why ortho, ophthal, and derm each produce their own version of the problem, and what the automation actually needs to do in each case.

Post-op visit attendance in orthopedics

In my work with orthopedic practices, the highest-volume no-show category is not the new patient evaluation. It is the two-week and six-week post-operative visit. AAOS practice management resources have flagged post-op attendance as an operational concern for years; I am not aware of a single published benchmark that ranks no-show categories within ortho head-to-head, so I am stating the pattern as I have observed it across client books, not as a citation.

The mechanism is simple. The patient had surgery, tolerated it, reduced their pain medications, and now feels materially better than they did the day they scheduled the follow-up. The perceived benefit of attending has collapsed. The perceived cost (parking, time off work, a copay) has stayed constant.

This is a motivation problem. A reminder that says "your appointment is Wednesday at 10:00am" does little for that patient beyond the floor effect already discussed. They know. They are calculating whether to go.

The automation that works here is what I would call outcome anchoring at the re-confirmation touchpoint. The workflow pulls the procedure code and scheduled visit type from the EHR (this is a structured data pull, not inference), then generates a message tied to what that visit actually gates. A six-week post-ACL-reconstruction visit gates physical therapy clearance, return-to-sport authorization, and imaging orders.

If the automated outreach states that explicitly, 72 hours before the appointment, the calculus for the patient changes. This is not a reminder. It is a restatement of stakes.

Athenahealth's appointment reminder module does not do this natively. Neither does Cerner's patient messaging layer. Practices running ModMed with a custom n8n workflow between the PM system and their outbound messaging platform can configure something close, but it requires a human to write the template logic per visit type. Most don't.

Routine follow-up no-shows in ophthalmology

Ophthalmology has a structural problem that orthopedics does not. A significant share of ophthal follow-up volume requires dilation, and dilation visits create a patient cost that is invisible at the time of scheduling: you cannot drive yourself afterward, or you can drive and your vision is compromised for two to four hours. Patient satisfaction work in ophthalmology has consistently flagged transportation after dilation as one of the major cited barriers for solo travelers; I am hedging the language because I do not want to claim "most cited" without the AAO survey instrument in front of me.

Directionally, MGMA practice operations data tracks ophthal follow-up no-show rates higher than the surgical-specialty median, but I want to flag that I do not have a clean public DataDive cut to point to. If you are an MGMA member, the practice operations module is the place to verify against your own specialty cohort.

The driver cluster in ophthal is distinct from ortho: it is logistics failure more than motivation failure. The patient intended to come. They woke up with no ride, or their ride canceled, or they are in a county where rideshare density at 9:30am makes the trip impractical. Anecdotally, this is worse in Broward and Palm Beach than in Miami-Dade, based on what my client practices report; it is a workflow observation, not a sourced rideshare-density figure.

The automation that addresses this is a logistics trigger, not a reminder. Two days before a dilation-flagged appointment, the workflow checks whether the patient has confirmed transportation. If not, it surfaces options: a ride-share link pre-loaded with the practice address, a telehealth alternative for the non-dilation component if one exists, or a reschedule prompt with available slots already populated. This requires the workflow to read the visit type flag from the EHR, which most ophthal-specific EHRs code reliably for dilation-required visits.

EyeMD EMR and Nextech both carry dilation flags at the visit-type level. The gap is that neither routes those flags outward to an automation layer without custom integration. Practices on Eyefinity have better API access here, but adoption of that integration remains low based on what I see in the South Florida market.

Cash-pay cosmetic schedule shifts in dermatology

Derm is the outlier. In the practices I work with, cosmetic derm accounts for a disproportionate share of no-show volume relative to its share of visit count, and the reason is not logistics or motivation. It is offer sensitivity. AAD operational guidance has noted the cash-pay dynamic in cosmetic derm; I am stating the disproportionality as a pattern from my own client data rather than citing a specific AAD benchmark figure.

A patient booked for a botox session in December who sees a Groupon for the same treatment on Tuesday will reschedule. A patient booked for a filler follow-up who gets a competitor's email on Monday will cancel. These are cash-pay patients in a competitive market. The practice is not a locked-in referral source for them. Miami-Dade and Broward have some of the highest concentrations of aesthetic derm practices in the country, which means the substitution options within a patient's drive radius are abundant.

No automation pattern fixes price competition directly. What it can address is perceived relationship value. The hypothesis I have tested with derm clients is that practices running a pre-visit engagement sequence (not a mass-blast, but a visit-specific sequence that references the patient's prior treatment, the outcome they discussed, the next step in their plan) see directional improvement in same-day cancel rates on cash-pay cosmetic slots. In the two practices where I have tracked this end to end, same-day cancel on cosmetic dropped from the low-20s to the low-teens over a quarter. Two practices is not a study. It is a pattern worth testing in your own book.

The mechanism is that the patient's mental frame shifts from "I have a commodity appointment" to "I have a continuation of a plan."

ModMed's Klara integration supports structured pre-visit messaging. Nextech's patient portal allows template-based outreach with merge fields from the prior visit note. The constraint is that the prior visit note has to have structured data worth merging (the patient's stated outcome goal, the treatment performed, the next milestone), and most derm notes are narrative-heavy and unstructured.

An extraction layer, running on a current-generation Claude Sonnet model at roughly $3 per million input tokens (verify Anthropic's current pricing page; the Sonnet line has iterated quickly), can parse the prior note and populate a structured summary that the outreach template can pull from. The cost per patient on that extraction pass is fractions of a cent. The operational lift to configure it is real, but one-time.

The single-vendor problem

Most practice management vendors pitch no-show reduction as a feature, not a workflow category. They show you aggregate no-show rate before and after their reminder product. They do not show you no-show rate by visit type or by payer class, because their product was not built to differentiate at that level. That product also captures the reminder floor effect I conceded earlier, which is exactly why the before-and-after slide looks good in the sales meeting and then plateaus in production.

The practices I have seen move their no-show rate materially (from the 22 to 28% range down to the 12 to 15% range, which is broadly where MGMA practice operations cuts put top-quartile specialty performance) did it by building different workflows for different visit categories, not by buying a better reminder system.

They mapped no-show data by visit type. They identified which category was driving volume. They diagnosed the actual barrier. Then they built the automation to address that specific barrier.

That is more work than buying a vendor module. It is also the only thing that has consistently worked beyond the floor.

Four questions to pressure-test a vendor offering no-show automation

Before you sign anything, run these:

1. Does your system differentiate outreach content by visit type, or does it send the same reminder regardless of whether the appointment is a post-op follow-up, a dilation visit, or a cash-pay cosmetic slot?

2. Can your system pull from prior visit data to personalize the pre-visit message, and where specifically does that data come from in our EHR?

3. Show me three reference accounts in my specialty where you can produce pre and post no-show rates segmented by visit type, and let me call them directly without you on the line.

4. If our no-show driver is logistics rather than memory, does your product have a pathway to surface transportation options or reschedule prompts at the moment of abandonment, and does that pathway require custom integration work on our side?

If the vendor cannot answer the first question with a yes backed by a product demo, the rest of the conversation is not worth your time. If they cannot answer the third without stalling, the meeting is over.