At Phoebe Physician Group, an affiliate of Albany, Georgia-based Phoebe Putney Health System, which serves a mostly rural, 41-county service area, caregivers say missing doctor's appointments is becoming increasingly acceptable among patients.
problem
That led to a 12% no-show rate across physician groups, more than double the national average of about 5% reported by the Medical Group Management Association. For example, practices in urban areas can offer taxi vouchers, but not in southern Georgia. Automated reminder texts and calls didn't help, either.
Compounding the problem were Phoebe Physician’s size, the clinic’s vast market, and the challenges of staffing in rural areas. High staff turnover and inexperience led to inconsistent scheduling, double bookings, and inconsistent appointment confirmations.
“Even when staff think they're sending these reminders, they often aren't, or aren't doing so effectively,” says Matthew Robertson, chief administrative officer at Phoebe Physician Group, “which is why we decided to explore how artificial intelligence technology could help increase patient volumes while minimizing disruption for healthcare workers and improving the patient experience.”
suggestion
Things are, Berkeley Research Group. Phoebe Physician staff notified BRG about the issue, telling them that reminder text messages and phone calls weren't working, and that the organization needed to eliminate the human element to free up staff time and ensure work was completed.
“BRG proposed an AI tool they developed in collaboration with Trajum ML, Tool Development and Piloting MelodyMD,” Robertson explains. “The tool leverages machine learning to analyze years of patient visit data and predict the likelihood that a given patient will not show up for their appointment.”
“When a new patient appointment comes in, MelodyMD communicates with Phoebe Physician's booking system to analyze patient no-shows and automatically creates an adjacent appointment slot if the likelihood of no-shows exceeds a set threshold,” he added.
Rising to the challenge
The tool's developers examined data points to identify those most highly correlated with a patient's likelihood of not showing up, including patient demographics, provider specialty, appointment scheduling lead time, past appointment history, insurance, etc. As new patient visit data was added, the developers continued to refine the model.
“One of the key elements we worked out over time was to set a daily double-booking cap,” Robertson points out, “so that only patients who were likely not going to show up were subject to double-booking. We then applied exclusions to certain clinics and appointment types.”
“As the model was rolled out and tested, we adjusted our reminder process to improve communication with patients and ensure our teams had enough time to fill newly open appointment slots,” he continued.
AI tools have also enabled organisations to measure their performance and implement improvements at the next level, he added.
-
Patient access. Regularly monitor utilization, number of no-shows, completed visits, cancellations, cancellations within 24 hours, and rescheduled visits.
-
Referral management. You can regularly monitor your referral volumes, patient leakage and retention rates, and the volume of splitters and competitors.
-
Provider Scorecard. Routinely monitors work relative value units, visit type, evaluation and administrative coding, average visits per session, average days to schedule new and existing patients, no-show rates, and payer mix by provider.
-
Physician/advanced practice physician productivity. Regularly monitors work relative value units, visit types, evaluation and administrative coding, and current procedural terminology details by provider.
-
Non-provider staffing. Regularly monitor full-time equivalent paid pay, productivity and overtime hours to ensure staffing meets demand.
result
From January 2023 to February 2024, Phoebe Physician saw an average increase in practice volume of 168 cases per week, equating to approximately 7,800 additional practice cases and $1.4 million in net new patient revenue.
Robertson said while people still not showing up and local perceptions of getting a doctor's appointment remain obstacles, the double-booking factor has helped to greatly mitigate the impact of these occurrences.
Advice for others
“It's definitely important to talk to providers early in the process and be transparent about what's going on,” Robertson advises. “You need to clearly articulate the problem, the goals, and the potential impact you expect the AI technology to have. When we told 20 primary care providers that they were getting 84 no-shows a day, they were shocked. It was this story that got them on board with trying a new solution.”
“We also need to consider health equity,” he continued. “For example, how do we ensure that AI tools don't introduce implicit scheduling bias? How do we refine algorithms so that patients who are underinsured or have certain types of insurance aren't disproportionately affected by the possibility of their appointment slots always being double-booked and having longer wait times to be seen?”
And of course, AI is only as good as the underlying data, he added.
“We worked with BRG to collate three years of patient data and use it to establish an effective model,” he concluded.
Follow Bill's HIT articles on LinkedIn: Bill Siwicki
Email: bsiwicki@himss.org
Healthcare IT News is a publication of HIMSS Media.
