Einstein Healthcare Network has spent the last year on an ambitious—and successful—effort to reduce their high number of patient no-shows. To understand the new approach they took, The Daily Briefing spoke with Barry Freedman, CEO of Einstein, as well as members of Advisory Board's Patient Access Consulting team who worked with Einstein on the effort.
Q: Can you tell us about the no-show problem at Einstein?
Ayal Bitton, Advisory Board:When we started working with Einstein, their no-shows were off the charts—in the 20% to 30% range at several physician practices. We're talking almost 160,000 missed visits annually.
Einstein is an academic, urban health system that predominantly serves a pretty socioeconomically challenged population, so those high rates weren't particularly surprising. But we felt an urgent need to address the issue.
Barry Freedman, President and CEO, Einstein Healthcare Network
Barry Freedman, Einstein CEO: Not only was our no-show rate a problem in itself, but it was creating a very long wait times for patients. Whenever schedulers looked at the schedule, it was full of visits—but many people never showed up to these visits. This meant there were long delays to get new patients in or to get repeat patients back. We knew this was something we wanted to address in a more systematic way, as a medical group rather than leaving it to any individual practice.
Q: How did you begin to approach the problem?
Bitton: When we think about helping a health system with no-shows, we typically start by employing tools like patient reminders and strong policies and workflows. But in Einstein's case, we found they had high no-show rates across the board, even at practices that were employing these tools.
That didn't necessarily mean that those mitigation strategies could not be effective. What it meant is we needed to understand more about the root causes driving specific no-shows behaviors for specific patient subsets.
Jordan Holland, Advisory Board: We began working closely with Einstein's IT department to pull lots of historical scheduling data. We ended up looking at 20 to 25 different attributes to understand what was truly meaningful in predicting no-shows—including patients' demographics, geographic segmentations, clinical attributes, and others. We didn't go in asking, "How can we predict no-shows across the nation?" We asked, "How can we predict them at Einstein?"
“Appointment slots aren't interchangeable like airplane seats, so "percentage-based" approaches tend to backfire when overbooked patients show up at the same time”
David Sweeney, Advisory Board: We also knew we wanted to avoid traditional no-show mitigation strategies that follow the airline industry's playbook, overbooking a given percentage of appointments. Appointment slots aren't interchangeable like airplane seats, so "percentage-based" approaches tend to backfire when overbooked patients show up at the same time.
So our tool worked a little differently. First and foremost, we wanted to reduce the chance of a particular patient becoming a no-show. We wanted to understand the specific factors or social determinants that make them a no-show risk—whether that's transportation issues, financial concerns, or something else—so they could address those particular issues head-on, whether that's with an Uber, an personalized reminder call, a home visit, etc.
Then, having tried to prevent no-shows, we wanted the tool to help practices strategically overbook based on no-show probability for each patient.
Freedman: Once we identified how big a problem no-shows were, I felt a great sense of urgency to move quickly. So when we saw this tool, we thought it would complement our existing efforts and help us make big gains on reducing our wasted capacity.
Q: What sort of results have you seen so far?
Bitton: When we launched the tool, it accurately predicted about 75% of patients who didn't show up at Einstein as being highly likely to no-show—and of patients who did show up, only 5% had been flagged as likely no-shows. And what's interesting is that the tool itself is self-learning; it'll keep evolving with more data to ensure it is consistently representing accurate correlations, even as we impact patient behavior.
Sweeney: By implementing the tool's recommendations, we've reduced the no-show rate from 19.4% to 16.5%, which is particularly noteworthy given the higher-risk patient population. We found in the pilot phase that if you utilize the tool and perform the recommended overbooking strategies on a day-to-day basis, you have the potential to reduce total overall patient no-shows by up to 30%.
“If you utilize the tool and the strategies, you have the potential to reduce total overall patient no-shows by up to 30%”
Freedman: And that's improved our bottom line, plain and simple. If you calculate 4% more visits on the base of almost 1 million visits, that's 40,000 additional visits a year. That's a lot of revenue just going to the physician companies. But those visits also drive downstream revenue, whether it be for radiology procedures, infusion procedures or future hospital admissions. Even more importantly, in these days and times, you can't have idle capacity and you can't have patients frustrated by a lack of access.
We've also noticed that our physicians are more satisfied; RVUs per physician have been increasing, which are tied to our compensation model. I think that we've also seen some improvements in patient satisfaction because they are experiencing better access. Certainly, the administrative team has seen the difference in terms of efficiency ratios, which makes them see that the efforts that they have put into this are paying off. So staff morale is better too.
Q: Could you speak more to how you see the tool being used in the future?
Freedman: Certainly, we are going to use these tools in all of our physician companies and our clinics. As we acquire or open new physician practices or satellites, we are planning on using this as a baked-in tool at all of them.
Reynolds: I'd also add that we've continued to expand this approach and specifically the tool's capabilities and sustainability by implementing it at other organizations—with drastically different geographic, socioeconomic and fundamental patient populations, but with similarly positive results.
Addressing no-shows should just be one part of an organization's overall strategy to improving patient access to care. But it can be a key part: The root-cause analysis of why patients are likely to no-show at your organization can lend a lot of insight to your broader patient access strategy as you begin to understand patient behavior.
“I think everybody that is experiencing a problem with no-show rates needs to consider applying tools like this to achieve better efficiency.”
Q: What advice would you give to other health system leaders facing a similar problem?
Freedman: While I think the no-show rates are always going to vary substantially based on the setting, I think that tools that allow you to predict the risk of patient no shows and the ability to create patient risk profiles are incredibly valuable. I think everybody that is experiencing a problem with no-show rates needs to consider applying tools like this to achieve better efficiency.
Struggling with no-shows? Speak with our patient access consulting team.
Patient access and navigation is impacted by a host variables across the health system—from scheduling and appointment capacity, to physician productivity and facility design. Our experts work with health system leaders across the country to identify roadblocks, and develop a coordinated, cross-continuum strategy to promote a seamless patient experience, and improve access across the system.
Contact us to learn how we can help you improve patient access to care.