Accurately predicting no-shows with advanced analytics to address patient access

Optimizing patient access by minimizing wait times for appointments and reducing no-shows has been a top initiative for many medical groups through the post-pandemic recovery.

According to the MGMA DataDive Practice Operations data set, time to third-next-available appointment for new patients, a key patient access metric, dropped by half from 2019 (10 days) to 2022 (five days) in primary care clinics.

That same data set also found that practice no-show and appointment cancellation rates have fallen below their pre-pandemic levels, with the single-specialty aggregate rate (combining primary care, surgical single specialty and nonsurgical single specialty practices) for no-shows haven dropped from 7% in 2019 to just 5% in 2022.

While the shift to telehealth and other pandemic-era flexibilities spurred several improvements, the road ahead — with in-person care demands rising — will require sustainable efforts to keep unused appointment slots at a minimum.

A Feb. 5, 2024, MGMA Stat poll found that 15% of medical groups use predictive analytics to improve no-shows or patient scheduling, compared to 85% that do not. The poll had 382 applicable responses.

All medical group leaders responding to the poll were asked what their biggest challenges with patient access are:

• Provider availability was the most frequently noted issue, either from simply not having enough doctors amid ongoing physician shortages or from blocks on physician schedules making it difficult to accommodate more patients in a timely fashion. “Template limitations prohibit certain types of appointments even if space is available,” one respondent told MGMA.
  • The increased use of team-based care and advanced practice providers (APPs) was cited by many practice leaders as a solution to ensure patients can be seen.
• Overall demand for care has been rising since the worst of the COVID-19 pandemic, according to many practice leaders, with one respondent noting that the inbound telephone call volumes “are greater than staffing capacity” this year.
• Many respondents noted they are struggling to build in more time for same-day appointments.
  • The 2023 MGMA DataDive Practice Operations survey report noted that surgical and nonsurgical specialties were the most limited on their same-day appointment availability in 2022, while primary care specialties managed an increase in their same-day rate in 2022 to above pre-pandemic levels reported in 2019.
• Ensuring an efficient check-in/registration process was noted as a growing challenge by many respondents, as many administrators have considered or adopted stricter policies about when patients should arrive for their appointments to prevent delays.
• The challenge of no-shows in general was mentioned by many practice leaders, with many noting that they have branched out into areas such as arranging transportation for patients to and from appointments to ensure they make it in to see their provider(s).
  • Several medical group leaders said that they have seen an uptick in patients not having a reliable cell phone number on file for appointment reminder calls.
  • Similarly, time and effort spent on payer authorizations and coverage checks was noted to be higher than it had in previous years, especially with many older patients adjusting to changes from traditional Medicare coverage to Medicare Advantage plans.

Leveraging predictive analytics to mitigate the loss of access and revenue

In a recent MGMA insight article, Sharon V. Nir, MBA, vice president of patient access optimization at Ardent Health Services, details how the group — which operates 30 hospitals and more than 200 care sites in six states — built upon its 2017 adoption of Epic as its sole EHR platform to add advanced predictive analytics features to address no-show rates in some markets that were as high as 18% in July 2022.

The Ardent team paired a no-show probability model for each appointment with a newly implemented strategy for appointment confirmation calls to patients; for appointment slots with a higher probability of no-shows, they could overbook in a targeted fashion. As Nir puts it, “the organization began focusing on the outcome rather than relying on patient behavior modification.”

To measure success with overbooking appointments with higher probability of no-shows, Ardent developed a Power BI report to summarize the opportunities and total volume of all overbooking, including random overbooking. In turn, they were able to shift the focus from measuring no-show prediction success “to quantifying its accuracy in identifying overbooking opportunities compared to random overbooking.”

Read the full article for Nir’s five key considerations for successfully implementing a no-show predictive model to optimize overbooking strategies.

More tips on deterring no-shows

Yuriy Kotlyar, co-founder and CEO of American Health Connection, offered these tips in his December 2023 MGMA insight article on addressing no-shows:

Optimize appointment reminders

• Remember that cancellations aren’t the problem: No-shows are the problem. Your automated reminders should make it easy for patients to cancel and release the appointment slot.
• Ensure your reminders offer personalized messages that include the specific details of their appointment, such as date, time, location and the provider’s name or service (e.g., MRI) being provided. With these details, the patient can determine, at a glance, whether they can come to the appointment or need to cancel and reschedule.
• Offer personal preference options to patients. These include the choice of channel (email, text message, phone call), language options and the ability to opt out of reminders.

Ensure operations support the reminder strategy

The next step is to ensure the contact center function has the workflows and capacity to support the patient reminder strategy. For example, based on one organization’s experience, about 15% of appointment reminders generate a call to the contact center. Even when there is a frictionless way to cancel the appointment through a text message reminder, some of the patients who cancel might want to speak to someone to reschedule the appointment.

The operational question is whether these patients will be able to get through to someone to cancel the original appointment and reschedule. If they cannot get through with a reasonable wait time, they are highly likely to become a no-show patient.

These calls must be accounted for in determining the right level of contact center capacity during evening hours. In weighing the costs of this capacity, consider the high value of preventing no-show appointments.

Additional resources

• “Restoring Balance to Patient Access and Medical Practice Operations” — This August 2023 data report details key benchmarks for value-based care, quality measures, appointment scheduling, no-shows, billing, collections and more.
• “Mirror, mirror on the wall, will the patient show at all?” — Sharon V. Nir shows how Ardent Health leveraged predictive analytics to mitigate the loss of access and revenue due to patient no-shows
• “Better patient communication to balance your approach to deterring no-show appointments” — Yuriy Kotlyar details how better communication offers opportunities to reduce no-shows and enhance patient experience.