A year ago, Mount Sinai Health System’s pharmacy could not work through prior authorizations fast enough to keep up with demand. Now its biggest constraint is refrigerator space.
The New York City-based system’s GLP-1 volume has roughly doubled over the past year as the pharmacy took on new clinics and the Medicare GLP-1 Bridge program launched. AI-assisted prior authorizations let the back office absorb the surge, according to Mark Makhinson, PharmD, senior director of outpatient and specialty pharmacy. The bottleneck moved to the shelves.
In December, Dr. Makhinson said the system’s AI-enabled prior authorization tool had cut completion time by at least 50%, with a goal of 75% to 80%. Since then, the Medicare GLP-1 Bridge program has added a new layer of work.
The program went live July 1, and Mount Sinai had a GLP-1 module running in its prior authorization platform within about a week, Dr. Makhinson said. Bridge claims are billed to a separate third party rather than a patient’s Medicare Part D plan, so staff must first bill the Part D plan, receive a rejection and then submit to the Bridge program. The tool automates that sequence. It also scans the medical record for preexisting conditions and lab values, such as obstructive sleep apnea or A1C levels, that determine whether a patient is eligible.
The workload per patient is also heavier. Many Bridge patients seek three or four approvals as they weigh oral and injectable options, he said.
“Having this tool has been indispensable for us,” Dr. Makhinson said.
Between 200 and 250 members of Mount Sinai’s internal pharmacy team use the platform daily. Prior authorization efficiency has improved 60% to 75%, up from about 50% a year ago, Dr. Makhinson said. The gains came from two directions: Staff grew more comfortable with the tool, and the tool got better at searching the medical record and pulling out the right information.
The team would likely be past its 80% target if volume had held steady, he said. Instead, GLP-1 volume roughly doubled as the pharmacy took on clinics it previously lacked the capacity to serve.
That volume brings a particular strain. GLP-1s and SGLT2 inhibitors carry a heavy prior authorization burden but do not have the same value proposition as specialty drugs, Dr. Makhinson said. To extend the same high-touch service to those patients, the team had to find efficiency elsewhere.
The effect shows up in turnaround time. A year ago, the pharmacy averaged about two and a half days from prescribing to the release of a clean prescription to the pharmacy. It now averages under two days and has dropped to 1.7 days in some months, he said.
“If you can automate that PA process then you can focus on outreaching to the patient,” Dr. Makhinson said. “You can focus on the financial assistance and navigating that patient to the pharmacy.”
Appeals have sped up too. A denied prescription once took days to appeal because a pharmacist had to write it from scratch. Now the tool gathers the data and drafts the appeal in about five minutes, and a pharmacist reviews and submits it. Both first-pass approvals and appeal approvals are rising, he said.
The new constraint is physical. GLP-1s require refrigerator and freezer capacity for storage, packing and shipping, along with physical space for the shipping process.
“Our problem has flipped over the last year,” Dr. Makhinson said. “Before, we couldn’t keep up with volume from a back-end perspective. Now our choke point is on the dispensing side, just because we are limited in space.”
That fix has a longer runway, he said, since it requires acquiring new space, building it out and securing licensing.
AI oversight has matured over the same period. A year ago, the pharmacy was among the first teams to present to Mount Sinai’s newly formed AI governance committee. The system now has several committees that oversee AI by use case, with different guardrails for patient-facing, provider-facing and administrative tools.
For health systems with smaller pharmacy operations, Dr. Makhinson’s advice is to think beyond specialty drugs. Retail medications such as GLP-1s and SGLT2 inhibitors are getting harder to obtain as payers add access hurdles, and health systems are positioned to manage a patient’s whole medication list, which can improve adherence and reduce readmissions, he said.
A system that runs only a specialty pharmacy will need automation once it expands into GLP-1s, he said, because volume can jump “2x, 3x, 5x overnight.” Many patients on biologics are also taking a GLP-1, and the drug class’s indications keep expanding.
“The only way to keep up with the demand is to automate that process,” Dr. Makhinson said.
Prior authorization is only one piece, he added. The next step is extending the same AI-assisted, human-in-the-loop model to other back-end work in the medication delivery process.
Becker’s 3rd Annual Fall Chief Pharmacy Officer Summit is the top tier event that the people leading the future of pharmacy convene. This 3-day event is built specifically for Chief Pharmacy Officers and senior pharmacy leaders. The sessions examine the workforce, financial and clinical priorities shaping pharmacy today. Becker’s is where candid discussions flourish and new ideas thrive. Becker's 3rd Annual Fall Chief Pharmacy Officer Summit
