Listen to this article

Patients who fail to show up for scheduled appointments have long been a costly problem for health care providers.

Data has shown that missed appointments cost the U.S. health care system more than $1.5 billion annually. For individual physicians, it averages out to about $200 per unused time slot.

Besides lost revenue for practices, no-show patients can potentially delay important medical treatments for themselves. It can lead to interrupted continuity of care, poorer clinical outcomes, unmonitored medication use and delayed preventative screenings.

Skipping an appointment without notifying staff also creates more administrative work and prevents other individuals from getting the care they need.

There are a range of reasons why patients don’t show up. Major factors include language barriers, economic issues, transportation and mental illness, a 2024 Penn State Health study found. Scheduling conflicts, lack of reminders and forgetfulness also come into play for some people.

While practices often implement strategies to mitigate missed appointments – like reminder emails, texts and phone calls – no single approach is entirely effective. For providers searching for solutions to reduce no-shows, it’s essential to understand why patients are missing their appointments.

That’s where Millburn-based Predictive Health Solutions comes in.

An NJBIZ 2024 finalist for Emerging Business of the Year, PHS is leveraging the power of predictive analytics to help health care facilities gain insight into the root causes of no-shows so they can develop targeted interventions and optimize scheduling.

PHS – a joint venture between the Center for Discovery, Innovation and Development at Children’s Specialized Hospital in New Brunswick and specialized software company Pinnacle Solutions Inc. – is rolling out a tool that aims to reduce patient backlogs and long waitlists; improve operational efficiencies; promote better patient health outcomes; increase potential revenues; enhance the care experience; and minimize caregiver burnout.

Using a combination of patient data and external sources, along with advanced machine learning and AI capabilities, the Patient No-Show Predictor identifies high-risk patients as well as potential factors affecting their ability to keep an appointment, such as weather, social determinants of health or many other components.

Then, staff can quickly make informed decisions regarding intervention strategies that directly impact a patient’s likelihood of attending their appointment and receiving the care they need.

For example, staff may choose to offer options to those with unreliable transportation or suggest afternoon appointments for those whose morning commute would be affected by traffic delays.

Teams may also choose to use the intelligence gathered to customize individual reminder protocols to maximize the impact and value of those existing tools. Data can also be leveraged for strategic overbooking, as opposed to, for example, blindly scheduling same-day appointments, PHS said.

Filling gaps

After announcing its launch in December 2022, PHS began piloting its Patient No-Show Predictor at Children’s Specialized Hospital, an affiliate member of RWJBarnabas Health and a member of Children’s Miracle Network Hospitals.

Created by Children’s Specialized Hospital in 2018, CDID focuses on improving the lives of children, adolescents and young adults with special health care needs through the discovery, development, growth and commercialization of innovative care, products, services, collaborations and ideas.

Just a few of the emerging technologies the Center for Discovery, Innovation and Development has explored via projects and partnerships include the use of virtual reality in behavioral health treatment and a nutrition delivery device to assist enteral feeding patients.

Like many health care facilities, no-show appointments are a problem at Children’s Specialized Hospital, according to CDID Director and PHS Chief Operations Officer Victoria Porterfield Gregorio.

Commenting on the industrywide trend, Gregorio said, “Over the last couple of years, wait lists for every service are growing longer and longer and longer. And that’s because of provider shortages. Of course, the best thing to do would be to find more providers who can see more patients.

“But, alternatively, identifying no-shows and where doctors are sitting idle and they could have seen an appointment … could fill in those gaps to help reduce those really long wait periods that we are seeing, especially in New Jersey,” she said.

When it came to selecting a partner to co-develop the tool, Gregorio said Indianapolis, Ind.-headquartered Pinnacle Solutions was the top choice.

“They are experts in data analysis and data science. I think they also realized our uniqueness in how we developed operational actions from the data that were being presented. And I think they recognized and see that at the Center for Discovery, Innovation and Development, our expertise is in being able to craft operational solutions to make it tangible for other organizations,” she said.

“You’ll see a lot of data analytics companies out there, but you don’t necessarily see operational health care expertise attached to the AI component of data analytics companies. So that’s what I think makes us unique and that’s where the joint venture just made sense,” Gregorio said.

Drawing upon their combined expertise, CDID and Pinnacle set out to create a solution that empowers providers to reduce missed appointments, enhance operational efficiency and ultimately deliver improved care experiences for patients.

Gregorio said, “The idea came from the tools and the power behind AI, data science and data analytics. And then there’s the operations of what do you do with that information to impact change … What we’ve learned and been able to effect change in is the impact of tailored solutions.”

Effecting change

PHS believes the partnership marks a big step forward in health care’s future by decreasing costs while keeping patients at the center of care.

Since the Patient No-Show Predictor was introduced at Children’s Specialized Hospital in 2023, it has resulted in a drop in no-show rates and less waiting time for scheduled appointments. It’s also led to more time allocated to providing quality care and increased patient satisfaction.

According to PHS, Children’s Specialized Hospital achieved 93% accuracy in predicting patient no-shows and decreased missed appointments by 60% at one of 10 outpatient locations using the tool.

“You hear about patient experience, and you hear about those things being important, but it was really interesting to see how that change impacted the show rate in terms of data,” Gregorio said.

She noted that the hospital serves youngsters with unique challenges. So, for many parents, coming to Children’s Specialized Hospital is a brand-new experience. “There may be anxieties built in. So, developing a very tailored approach for welcoming new patients and having that really warm welcome is important,” Gregorio explained.

Following a successful test run at Children’s Specialized Hospital, PHS is targeting hospitals, clinics, large practices, medical service organizations, dental service organizations and diagnostic centers.

PHS Vice President Of Sales Bert Fajardo said, “In speaking with a lot of organizations, they sort of take it [patient no-shows] as a pain point that everyone needs to deal with. And they don’t realize there’s a huge ripple effect with regard to the impact.”

“And oftentimes folks that are involved with making sure that people confirm their appointments, it is a side of desk activity. It’s very manually intensive. And oftentimes in many organizations, folks are told, ‘Look, here’s who’s coming tomorrow or two days from now, make sure you call everyone and get a confirmation,’” Fajardo said.

“And then you throw in these messaging systems and what ends up happening becomes a bit more chaos because the patients are getting inundated with messages. The staff that are assigned to work on this are juggling three different, four different jobs that they’re responsible for. And no one is really leveraging the data to make it a data-driven approach that says, OK, now let’s focus on the patients that in all probability are more likely to not show up or cancel,’” he said.

Fajardo went on, “The problem is that organizations need to zero on the patients that are more than likely not going to show up. And that frees them up from a lot of the manual effort reaching out to patients that are coming.”

Gregorio said, “A lot of organizations will do a blanket reminder of a booking. But it doesn’t take into effect someone’s likelihood of not showing up or what’s the underlying reason for that. For example, Burt’s reason for not showing up for his appointment may be very different than my reason for not showing up for an appointment.”

“I think that is why it’s a pervasive challenge, is because while they’re solutions, they tend to be more for everybody. And that’s not the challenge. The challenge is unique to individuals. So, what our solution tries to do is to bring to light what are more unique challenges at the individual level so that organizations can make smart decision making and how they affect change,” Gregorio said.

A ‘Shore’ solution

As for the data and sources used to help inform the no-show rate, Gregorio said, “We really threw everything into it.” The solution leverages data, including ZIP code and family demographics, along with other enriched information, such as weather, traffic, construction, demographic, clinic specialties, retention, ratings and other factors to predict the likelihood of a patient showing up for their scheduled appointments.

Additional information like previous appointments, diagnosis codes, demographic characteristics, distance away from the practice and many other attributes can further improve the model results.

“What’s beautiful about machine learning is that you can throw anything and everything in the model and the model will determine what’s important,” Gregorio explained. “The tool itself probably took a couple of weeks to build, but we’re always refining it. We’re always finding new things to really make it better … That is a continuous process to be quite honest,” she said.

Fajardo added, “The more data we throw at it, the more intelligent the model becomes, the more information we get that we’re able to react to.”

“Predictive analytics is a tool. But the overarching offering from Predictive Health is basically around people, processes and tools. All three combined is what makes a difference,” Fajardo continued.

He went on to note, “AI is really getting much more acceptance now in health care. So, we think the timing is better now as well because a couple of years ago … it wasn’t really embraced by health care organizations.”

Fajardo and Gregorio recalled an interesting insight from the pilot. While trialing the Patient No-Show Predictor, Children’s Specialized Hospital learned that many patients from the Jersey Shore area were either cancelling appointments or showing up late during the summer when traffic was heavy on Fridays.

As a result, the facility became more mindful of scheduling during that timeframe — especially if it’s a newer patient who may not be familiar with coming to the location.

Some providers may not necessarily be aware of how patient no-shows affect profitability, according to PHS. “It varies. Some places have a really good sense of it. Some have absolutely no clue what their pain point is,” Gregorio said.

Depending on the specialty, the U.S. has a no-show rate of between 14% to 39%, according to various studies.

Assuming a 20% no-show rate, PHS estimates missed appointments and cancellations can cause a large health care system to lose over $110 million every year.

A medium-sized organization that sees roughly 250,000 visits a year yields about a $13.7 million cost and a 10-person practice with 48,000 appointments a year can lose $2.64 million annually, PHS says.

However, if organizations use the Patient No-Show Predictor, PHS believes it can save $5 million for a large health care system, $682,000 for a medium-sized health care system and $132,000 for a 10-person practice

Based on discussions with facilities, Fajardo believes they are not tracking no-show rates and are unaware of the impact on finances or patient outcomes.

So, it’s a big goal to “get organizations to at least start using data to get a better feel of how dire their situation may be and what the low-hanging fruit is,” he said.

“We initially do a landscape review to find out where exactly an organization’s pain points are around patient no-shows and late cancellations,” he said.

While that can be a “very, very complicated” investigation, Fajardo said PHS can have a customized solution up and running “within weeks.”

“Then, we gradually evolve the tool over time,” he said.

Editor’s note: This story was updated at 9:30 a.m. ET Jan. 28, 2025, to correct the name spelling of Predictive Health Solutions’ Bert Fajardo.