Skip to main content

Pine Park Health sees ROI from AI scheduling tool

Since artificial intelligence took over reminder calls, the mobile medical practice has given more than two FTE back to its medical assistant team. And a proactive scheduling agent has helped boost the number of visits per day.
By Bill Siwicki , Managing Editor
Michael Tadlock of Pine Park Health on AI

Michael Tadlock, COO at Pine Park Health

Photo: Pine Park Health

Scheduling needs (booking, changing, cancelling) are the most common inbound request type at Berkeley, California-based Pine Park Health, an in-home primary care provider organization – representing 19% of phone calls.

THE CHALLENGE

Older patients are "phone-first" communicators. So Pine's team of medical assistants traditionally has spent a good portion of the day answering inbound appointment request calls, calling outbound to proactively schedule overdue patients and reminding patients of upcoming visits to reduce the no-show rate.

Each scheduling call – inbound, outbound and reminder – averages about 4 minutes, including preparation, handle time, dispositioning and documentation. This was a massive time commitment, and the opportunity cost kept MAs from focusing on other complex care coordination tasks like medication refills, cardiology referrals and building out charts for new patients, said Michael Tadlock, Pine Park Health's COO.

"We run a 'top-of-license,' team-based practice where PCPs, nurses and medical assistants all play specialized roles," he explained. "To succeed as a team in caring for the complex needs of our patients, communication is essential. Properly documenting calls is a key success factor in a team-based approach to care.

"Unfortunately, thoughtful and thorough documentation of calls is one of the first things that gets sacrificed when the phone is ringing off the hook on a busy Monday morning," he added.

As a mobile practice, Pine faces significant logistical challenges.

"Our providers often drive to two senior living communities per day," Tadlock noted. "Their time is the most precious resource. No-shows are a missed opportunity for our high-need patients as in some cases it means our PCP won't be back to the same building until the following week.

"Delays in care for a poly-chronic 83-year-old patient often have significant repercussions," he continued. "Plus, missed appointments create operational waste by kicking off an effort to get the patient rescheduled, which often involves several inbound and outbound communication loops. No-shows have a significant drag on our revenue per PCP per day. No-shows are expensive for us, so we work hard to avoid them with reminders."

Pine sends a Net Promoter Score survey after each appointment. Staff know from these surveys that consistent execution when it comes to scheduling, changing and showing up on time are key drivers of satisfaction for patients. Getting great at this part of operations has an outsized impact on new patient growth through referrals, Tadlock said.

PROPOSAL

Pine Park Health knew it wanted to focus on scheduling workflows. Staff felt they were constrained and the odds of success were high.

"In fall of 2024 we attempted to build our own AI call system," Tadlock recalled. "We had successfully built our own ambient dictation tool to help ease the charting burden on our PCPs. We quickly learned real-time audio via a telephony stack was a prohibitive challenge. We then attempted to work with a tech company and started building, about the time OpenAI released advanced voice mode.

"We made good progress, but couldn't get the quality to a level I felt comfortable deploying to our patients and their families," he continued. "So, we shelved the project for six months, figuring the underlying capabilities would advance quickly. In spring of 2025, I compared Retell AI and another voice AI vendor."

Tadlock chose Retell AI as Pine's artificial intelligence voice vendor for three reasons.

"First, I loved the self-serve flow and agent design tools," he explained. "Our first attempt with the startup a year ago meant I had to request they make a code change each time I wanted to tweak something based on what I heard in the calls or read in the transcripts. This rate limitation was very frustrating.

"Second, the cost," he continued. "The vendor was cheaper and more flexible on pay-as-you-go contracting with us."

But by far the biggest reason he chose Retell AI was the attention the practice was given by the company's team, including its founders.

"The level of service as we got started together was one of the best new product onboarding experiences I've been a part of," he reported. "Their team interviewed ours. They invested the time and energy to deeply understand our patients, our use cases deeply and our goals."

MEETING THE CHALLENGE

Pine Park Health built two AI voice agents using Retell AI.

"'Ava' is our appointment reminder agent," Tadlock said. "Two days before each appointment, Ava calls our patients or their responsible party to remind them of the visit. She verifies patient identity using date of birth, can cancel or reschedule visits if the time no longer works, asks what the patient would like the visit to focus on, and writes a phone note directly into the patient's chart so the PCP is properly briefed before the visit.

"If a patient gets confused or frustrated, Ava transfers the call to our live team for escalation," he added.

"Jay" is the proactive scheduling agent. Jay looks two weeks ahead into the PCP schedules, identifies open appointment slots, and then looks for residents in those communities who haven't been seen recently. He uses a set of heuristics to build a call list that fits available slot capacity and then calls patients to see if they'd like to be seen while the PCP is on site.

Jay also validates patients using date of birth and writes notes back into the patient's chart with details captured during the call.

"We rolled this out very cautiously," Tadlock remarked. "We started with a small pilot where calls were hand-selected and manually triggered, and I reviewed every transcript and listened to each call. Over several months, we gradually increased volume, refined prompts, adjusted cadence for older patients, and trained our MA team to review transcripts and handle cancellations, changes and time-sensitive updates."

Today, Ava makes all reminder calls and Jay runs continuously as part of the scheduling workflow.

RESULTS

"Since Ava took over reminder calls, we have given approximately 2.1 FTE back to our medical assistant team," Tadlock reported. "That time has been redirected toward other care coordination workflows, allowing MAs to focus on patients with more complex or time-sensitive needs."

And Jay, the proactive scheduling agent, books visits with 55.7% of the callers he connects with.

"This has proactively increased the number of visits per day for our PCP team by filling open slots that would otherwise go unused, improving operational efficiency in a mobile care model," he concluded.

Follow Bill's health IT coverage on LinkedIn: Bill Siwicki
Email him: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.

WATCH NOW: In 2026, AI will power change in RTLS