How Initial Booking Channels Shape Cancellations and No-Shows

During our discussions with practice leaders, MDfit is often ask a simple question:

Does the way a patient initially books an appointment change the chance that they cancel or no-show later?

The short answer is yes — but not in the way you might expect.

The more precise answer found in both literature and MDfit's own data, is that there are a consistent set of mechanisms that drive cancellations and no-shows, regardless of the scheduling pathway. It's the variation in these mechanisms by pathway which leads to differences in cancellations and no-shows: lead time, booking friction, patient certainty, and whether the appointment scheduling was fully completed at the first touch.^[1][2][3]

For example, when a phone call to the practice results in the patient discovering scheduling availability that is not shown online or via a portal, that variation in lead time can be the true driver of a future cancellation or no-show.

The four booking modes in this comparison

This post compares these pathways:

1. Existing patient appointment scheduled as a follow-up in the office during an appointment
2. New patient appointment scheduled by calling the practice
3. New patient appointment scheduled via an online portal (real-time booking)
4. New patient appointment scheduled via an online request form (asynchronous request, then staff processing)

It also separates no-shows from cancellations/reschedules, because of the different operational outcome. For most practices, a channel that produces more early cancellations but fewer no-shows is actually the better channel, because an early cancellation creates recoverable capacity while a no-show usually does not.^[2][4]

A quick note before the comparison

Two cautions are important here.

First, specialty matters. The same booking channel can behave differently in dermatology, orthopedics, GI, ophthalmology, primary care, and tertiary referral clinics.

Second, implementation matters at least as much as channel. A direct-booking portal behaves differently from a web request form that simply captures the patient request and drops it into a staff queue. Treating those as the same "online scheduling" channel is one of the biggest sources of confusion in access analytics.^[5]

The comparison at a glance

1) Existing patient appointment scheduled as a follow-up in the office during an appointment

This is the lowest-friction booking mode in everyday ambulatory practice. Why? Because the patient is already physically present, the relationship already exists, the clinician has just reinforced the need for follow-up, and the patient (and staff) do not need to remember to call back later. That combination removes several failure points all at once.

Direct outpatient studies that isolate checkout-scheduled existing follow-up visits and publish clean no-show benchmarks are surprisingly sparse. But the follow-up attendance literature points in the same direction. A Dartmouth-Hitchcock ophthalmology study found that when patients who had not scheduled on the way out were given a prescheduled follow-up appointment, substantially more follow-up appointments were completed than with a postcard reminder system (74% vs 54%).^[6] In a randomized trial of emergency department patients, follow-up compliance was 59% when the appointment was made before discharge versus 37% when patients were told to arrange it themselves later.^[7] In another study of post-discharge primary care follow-up, having an appointment booked before discharge was associated with higher attendance (adjusted OR 2.14, 95% CI 1.07–4.40).^[8]

Taken together these studies support the same operational principle: when the next appointment is booked while the patient is still present, follow-up completion improves.

So if you are trying to benchmark your own practice, this category is best treated as the lowest expected non-completion risk pathway of the four. This is especially true when the follow-up interval is not excessively long, as when the length between visits grow, both no-shows and cancellations rise.^[1][3][4]

2) New patient appointment scheduled by calling the practice

Many specialty practices assume phone scheduling is the "safe" or "higher commitment" path for new patients. Sometimes it is. But the literature does not support the idea that staff scheduling by phone is automatically the best-performing channel for no-show prevention. Phone scheduling by staff does have real advantages. It allows staff to clarify symptoms, steer patients to the right subspecialty, check obvious prerequisites, and reduce some types of mis-booking. But it can also introduces its own friction, with hold times, callbacks, abandoned calls, limited business hours, and a longer elapsed time between patient intent and confirmed appointment. That scheduling interval is one of the most replicated predictors of both no-shows and cancellations.^[1][2][3]

When staff scheduling has been compared directly with online booking, the results are mixed rather than uniformly favorable for either. In a large multisubspecialty orthopedic practice, Kachooei and colleagues found no significant difference in no-show rates between online-scheduled and traditionally scheduled appointments for either new or follow-up visits.^[9] In contrast, an ophthalmology practice study found that offline-booked appointments had a higher no-show rate than online-booked appointments (5.9% vs 1.8%), although online-booked visits were more often canceled or rescheduled before the visit.^[2]

These studies suggests that phone/staff scheduling may sometimes produce fewer visible pre-visit changes, but not necessarily lower final no-show risk. Phone based scheduling is still valuable for complex new-patient routing, but it should not be assumed to be the lowest no-show channel by default.

Its performance depends heavily on hold times, callback design, lead time, and how much uncertainty staff resolve during the call.

3) New patient appointment scheduled via an online portal

This category refers to real-time direct booking — a patient is presented available slots and confirms an appointment inside a portal or similar scheduling interface.

This is where some specialty practices become skeptical, because online booking can look "too easy" and the clinicians become concerned of misbookings and loss of control. The evidence suggests a more nuanced story here.

In many real-world settings, direct online booking is associated with lower no-show rates than phone assisted staff booking. See again, the ophthalmology practice study mentioned above with lower no-show rates for online-booked appointments (1.8% vs 5.9%).^[2]

At the same time, that study found that online-booked visits were more often canceled or rescheduled than phone assisted staff booked visits.^[2] Depending on your practice, that cancel/reschedule "churn" should not automatically be read as a failure or problem. For many organizations, a patient who cancels or reschedules early is operationally preferable to a patient who simply disappears at their appointment time.

Other evidence is more conservative. In the orthopedic study by Kachooei et al., no-show rates for online-scheduled versus traditionally scheduled appointments were not significantly different for new patients.^[9]

Put those two together and the interpretation becomes: real-time portal self-scheduling is not a guaranteed no-show win, but it can match or outperform phone/staff scheduling on no-show in some settings — and it often shifts some "churn" earlier, into cancellations/reschedules that are easier to recover operationally.

So, a online/portal channel may look worse if you only count cancellations, but is actually better if you care about same-day capacity loss.

4) New patient appointment scheduled via an online request form (asynchronous request, then staff processing)

This category needs special treatment because it is often mislabeled during our discussions with prospective customers. A web form that says "Request an appointment" is not the same thing as "Patient self-scheduling".

As Zhao and colleagues noted in their systematic review of web-based appointment systems, an asynchronous system allows patients to submit a request through email or an electronic form, after which staff manually process it later. In that model, requests are often placed in the same queue as phone requests and are limited by staff backlog and office hours.^[5]

That matters because it changes the mechanisms that drive non-completion. A direct-booking portal lets the patient finish the task immediately. A request form does not. It inserts lag, queue time, and additional staff mediation.

The no-show literature is highly consistent here: longer scheduling interval is one of the strongest and most replicated predictors of missed appointments and cancellations.^[1][3][4] In VA data, Whittle et al. found that missed appointment rates rose from 12.0% at day 1 to 20.3% by day 13, while cancellation rates increased from 19% during month 1 to 50% by month 12.^[4] In ophthalmology, McMullen and Netland found that no-show rates increased as lead time increased, with especially large effects in the higher-no-show resident clinic population.^[3]

The best directly relevant empirical example for request-form scheduling comes from the university hospital arm of the Freiburg ophthalmology study. There, patients submitted appointment requests through a web-based contact form, requests were triaged internally, and processing could take up to 7 business days. In that setting, the no-show rate for online-requested appointments was 14.3%, compared with 11.2% for appointments scheduled by specialists.^[2]

The authors themselves note that the hospital workflow was a request/triage system, not a direct-booking interface, and explicitly interpret the higher no-show rate in light of those implementation differences.^[2]

So the takeaway for this channel is: online request forms behave more like queued manual scheduling than like true self-scheduling. Their no-show and cancellation performance is usually driven less by "digital vs phone" and more by the lag and friction the request creates before the appointment is finalized.

In other words, if your "Schedule Online" button is really a request form, you should evaluate it as an intake workflow and not a self-scheduling channel.

Bottom line

• Existing follow-up booked while the patient is still in the office is the lowest-risk pathway for noncompletion.
• New patient phone scheduling is not automatically the best no-show channel; its performance is mixed and highly implementation-dependent.
• New patient direct portal scheduling can match or beat phone scheduling on no-show in many practice settings, while often increasing earlier cancellations/reschedules.
• New patient online request forms should be treated as asynchronous intake, not true self-scheduling, and they are often the highest-risk channel for lag-driven non-completion when response time is slow.

The bottom line is the channel that gives patients the fastest clear path to a correct, confirmed appointment — with the shortest avoidable lead time — will give you the best attendance pattern.