AI Software for Diabetes Care
Powered with Artificial Intelligence. Built for Human Intelligence.
AI Software for Diabetes Care from Unlimited Systems predicts which supply and device claims are likely to deny. It suggests codes for CGM, training, and monitoring services from the record. Diabetes care bills many small services, each with its own limits, next to visits, devices, and medications that need approval. Unlimited AI checks each of those services against payer rules automatically. Educators and billers can then spend their time where judgment is needed.
The Practical Side of AI for Specialty Practices
Avoid the pitfalls, cut through the noise, and put AI to work on everyday challenges. With Johnathon Trout (JT), Director of Product Management.
AI Built Around the Realities of Diabetes Care
A diabetes program sees patients for visits, group classes, nutrition sessions, device starts, and data reviews. Each service has its own code, time unit, and yearly limit. CGMs and pumps may fall under the medical, DME, or pharmacy benefit, and each path wants different paperwork. Newer medications often need prior approval backed by lab values. AI reads the record against each payer's rules and raises only the services where documentation, coverage, or limits fall short.
- Code 95250 covers CGM sensor placement, hook-up, and training on clinic-owned equipment. Code 95249 covers the same startup on patient-owned equipment.
- Code 95251 covers analysis and report of CGM data, and payers limit how often it can be billed.
- For Medicare, diabetes self-management training bills under G0108 for individual sessions and G0109 for groups, each per 30 minutes.
- Medicare's training benefit needs a referral from the treating provider and has yearly hour limits.
- Medical nutrition therapy bills under 97802, 97803, and 97804 in time units, with a separate code for group sessions.
- Under Medicare, CGMs and supplies are billed by a DME supplier under the glucose monitor coverage policy, which relies on the practice's order and notes.
Where AI Can Improve Diabetes Care Operations
- Problem
- Device and supply claims deny when the order or notes do not show what the payer requires, such as insulin use or a history of problematic low blood sugar. The supplier then asks the practice for records.
- Automation + Intelligence
- Each device order is checked against the payer's coverage criteria before it is sent. The platform looks in the chart for the insulin regimen, glucose history, and recent visit the payer wants. Orders with a gap are held with the missing item named, and claims the practice bills are scored for denial risk.
- Human Benefit
- Staff answer fewer supplier requests for missing notes.
- Business Benefit
- Fewer device and supply denials trace back to the practice's documentation.
- Problem
- CGM startup and data review codes depend on who owns the device and how often the data was reviewed. Billing 95251 again too soon is a common denial.
- Automation + Intelligence
- Unlimited AI reads the device record and visit note and drafts 95249, 95250, or 95251 to match. Each 95251 draft is checked against the last billed review for that patient. Every code cites the note or report behind it.
- Human Benefit
- Billers stop checking CGM billing dates by hand.
- Business Benefit
- CGM services are billed when earned, and repeat reviews in the same period stay off the claim.
- Problem
- Training sessions deny when no referral is on file or when the patient has used up the year's hours. Group and individual sessions also bill under different codes.
- Automation + Intelligence
- Each training session is checked for a referral on file and the hours the patient has left. Sessions draft as G0108 or G0109 from the attendance record. Patients close to their limit are flagged before the next class is booked.
- Human Benefit
- Educators know who can attend before the class fills.
- Business Benefit
- Training sessions are billed within the benefit, and fewer deny for a missing referral.
- Problem
- Newer diabetes medications often need prior approval backed by A1c results and prior drug trials. Gathering that proof slows the start of treatment.
- Automation + Intelligence
- AI prior authorization checks each medication order against the plan's criteria. It pulls A1c results, prior medications, and diagnosis details from the chart for the request. Approval and renewal dates are tracked as treatment goes on.
- Human Benefit
- Staff send complete requests rather than chasing lab results.
- Business Benefit
- Fewer treatment starts stall on a missing approval.
- Problem
- Glucose data sent from home supports monitoring codes only when readings arrive often enough and staff log their review time.
- Automation + Intelligence
- Reading days and review minutes are counted for each patient in the period. Patients whose uploads have stopped are flagged for a reminder. The draft claim lists only the monitoring codes those counts back up.
- Human Benefit
- Educators reach patients whose meters have gone quiet.
- Business Benefit
- Monitoring revenue follows the data, and short periods are not billed.
- Problem
- Diabetes quality measures, such as A1c control, eye exams, and kidney health checks, close only if a gap is caught before or during a visit.
- Automation + Intelligence
- Open gaps are listed for each patient from the same records billing uses. Patients due for an eye exam or kidney health evaluation are flagged when a visit is booked. Results are tracked for payer programs and MIPS reporting.
- Human Benefit
- Care teams see each patient's open gaps before the visit.
- Business Benefit
- Gaps close during visits the practice already has, which supports quality scores.
The Evolution of Intelligence in Diabetes Care
Eligibility and benefit checks run on their own for visits, classes, nutrition sessions, devices, and medications.
Training without a referral, CGM reviews inside a billed period, and device orders missing required notes are flagged before they go out.
Device-related claims and visits near a filing deadline move ahead of small training balances.
Each flagged item names the fix, such as the referral to request, the glucose history to add, or the next date a CGM review can be billed.
Emerging capability: routine renewals for stable medication approvals could be sent on their own when the chart has the needed labs. The platform's design allows for this step, though it is not live today.
Work by Exception for Diabetes Care
A patient's CGM data review is drafted under 95251. The same review code was already billed for this patient in the current period.
Unlimited AI holds that one claim and shows the date of the earlier review. A biller moves it to the next eligible date or drops it. Every other visit, class, and CGM start from the day moves on with no manual check.
AI Tracks the Limits. Your Diabetes Team Coaches Patients.
Living with diabetes takes daily effort, and patients lean on educators, dietitians, and clinicians to keep going. That time should go to coaching, adjusting plans, and teaching device use. It should not go to counting training hours or digging up glucose logs for a supplier. Automation picks up that work. Every unclear code, limit, or coverage question still goes to a person.
AI Across the Diabetes Care Revenue Cycle
Why Unlimited Systems for Diabetes Care
- CGM, training, nutrition, and monitoring codes are drafted from the record with citations.
- Training hours and referrals are tracked against each patient's yearly benefit.
- Device orders are checked against payer coverage criteria before they go out.
- Medication approvals are built from A1c results and prior drug history in the chart.
- Diabetes quality gaps are tracked from the same records billing uses.
Frequently Asked Questions About AI for Diabetes Care
AI software for diabetes care automates the billing and coverage checks behind diabetes programs. Unlimited AI drafts CGM, training, nutrition, and monitoring codes from the record. It checks training hours and referrals, reviews device orders for required notes, and gathers labs for medication approvals. It also flags claims with a high chance of denial before they go out. Staff review only the services with a real gap.
Supply and device denials often trace back to missing notes, such as insulin use, a history of problematic low blood sugar, or a recent visit. The platform checks each device order against the payer's coverage criteria before it is sent. For claims the practice bills, it also uses payer policy and past remits to score risk. Gaps are flagged with the missing item named.
Unlimited AI reads the device record and visit note to draft the right CGM code. It uses 95249 for startup on patient-owned equipment, 95250 for startup on clinic-owned equipment, and 95251 for analysis and report of the data. Each 95251 draft is checked against the patient's last billed review. Every code cites the record behind it.
Diabetes self-management training hours are tracked for each patient against the yearly benefit. Each session is checked for a referral from the treating provider before it is billed. Sessions draft as G0108 for individual training or G0109 for groups. Patients close to their limit are flagged before the next session is booked, so no one is surprised by an uncovered class.
Prior authorization for diabetes medications, such as GLP-1 drugs, often depends on A1c results and earlier drug trials. AI prior authorization checks each order against the plan's criteria and gathers those details from the chart. Each request is followed until the plan decides. Renewal dates are watched so a patient's treatment does not stop for a lapsed approval.
Remote glucose monitoring billing depends on reading days and staff review time. The platform counts both for each patient in the period. Patients whose uploads have stopped are flagged for outreach, so the care team can help them reconnect. The claim includes only the monitoring codes the counts support, so short periods are never billed on a guess.
Diabetes quality measures, such as A1c control, retinal exams, kidney health evaluation, and blood pressure control, are tracked from the same records billing uses. Open gaps are listed for each patient and flagged when a visit is booked. Care teams can then close a gap during a visit the patient already has. Results feed payer programs and MIPS reporting.
AI does not replace diabetes educators or billers. It counts hours, checks referrals, drafts cited codes, and gathers labs for approvals. Educators keep their time for teaching and coaching. Billers review flagged claims, handle appeals, and work with suppliers on records. Low-confidence items are routed to staff, and each automated step is logged.
GET STARTED
See What AI Can Do for Your Diabetes Care Practice
AI Software for Diabetes Care from Unlimited Systems predicts which supply and device claims are likely to deny. It suggests codes for CGM, training, and monitoring services from the record. Diabetes care bills many small services, each with its own limits, next to visits, devices, and medications that need approval. Unlimited AI checks each of those services against payer rules automatically. Educators and billers can then spend their time where judgment is needed.
6,500+ specialty providers