How to Fight Insurance Denials for CGMs and Insulin Pumps: A Patient Guide

Continuous glucose monitors (CGMs) like Dexcom G7, FreeStyle Libre 3 Plus, and Eversense 365, and automated insulin delivery (AID) systems like Omnipod 5, Tandem t:slim X2 with Control-IQ, Tandem Mobi, Medtronic MiniMed 780G, and the iLet Bionic Pancreas are now standard-of-care tools for managing diabetes—especially Type 1 and insulin-treated Type 2. Yet insurance companies routinely deny coverage with vague or outdated rationales: "not medically necessary," "experimental," "must try older device first," or arbitrary frequency/insulin-dose thresholds that contradict 2025 clinical guidelines. These denials force patients to choose between rationing supplies, paying thousands out-of-pocket annually, or living without technology that prevents hypoglycemia, reduces A1c, and improves quality of life. This guide explains why insurers deny CGMs and pumps, which citations win appeals, and exactly how to argue each common denial template.

---

Why Insurers Deny CGMs and Insulin Pumps

1. "Not medically necessary" or "doesn't meet criteria"

The insurer claims you don't meet their coverage policy—often because you're on "too little" insulin (e.g., basal-only, or fewer than 3 injections/day), your A1c isn't high enough or is too well-controlled, or you haven't had documented severe hypoglycemia. Some plans still cite outdated 4-fingerstick-per-day requirements that Medicare dropped in 2021.

2. "Experimental / investigational" (especially for AID systems)

Omnipod 5, t:slim X2 with Control-IQ, Tandem Mobi, MiniMed 780G, and iLet are FDA-cleared closed-loop systems with years of real-world use, yet some insurers label them experimental—particularly for Type 2 diabetes, pregnancy, or pediatric patients—or demand you "fail" an older non-AID pump first.

3. Step-therapy or "try older model first"

Insurers require you to use a previous-generation device (e.g., MiniMed 770G before 780G, Dexcom G6 before G7, or a non-automated pump before an AID system) even when your physician prescribes the current model and the older device lacks features critical to your safety (e.g., no smartphone app, no predictive low-glucose suspend).

4. Frequency limits or replacement denials

You're told sensors/pods can only be replaced every X days (ignoring adhesive failures, sensor errors, or manufacturer-recommended wear time), or that a pump under warranty can't be replaced despite mechanical failure, or that you must wait 4–5 years for an upgrade even when technology has advanced significantly.

5. Missing documentation or visit requirements

The insurer says your chart lacks an in-person or telehealth visit within the past 6 months, hypoglycemia unawareness isn't "documented," or your endocrinologist didn't fill out their proprietary form—even when your medical record contains all the clinical justification.

---

The Citations Insurers Respect

When you appeal, reference specific, named guidelines and policies by year. Generic phrases like "evidence-based care" are ignored; precise citations force the medical director to acknowledge binding standards. Below are the authoritative sources that win CGM and pump appeals in 2025–2026.

Medicare Local Coverage Determination (LCD) L33822 – Glucose Monitors

• Original 2017 LCD: Established therapeutic CGM coverage for insulin-treated diabetes.
• July 2021 revision: Removed the 4-fingerstick-per-day requirement and explicitly allowed smartphone-integrated CGMs (e.g., Dexcom G7, Libre 3 Plus).
• April 2023 revision (effective 2023-04-16): Expanded coverage to all insulin regimens (including basal-only) and to non-insulin-treated diabetes when the beneficiary has a history of "problematic hypoglycemia"—defined as recurrent Level 2 hypoglycemia (<54 mg/dL) or at least one Level 3 severe hypoglycemia event requiring assistance.
• 2025 requirements: Diabetes diagnosis + insulin treatment (any regimen) or problematic hypoglycemia history + in-person or telehealth evaluation within 6 months prior to CGM order + follow-up every 6 months.
• Companion Article A52464 clarifies documentation and HCPCS codes (E2102/E2103 for CGM integrated with pumps; A4238/A4239 for CGM supply allowance).
• 42 CFR §422.101(b): Medicare Advantage plans cannot impose stricter criteria than the applicable LCD. If you have MA and your plan cites thresholds not in L33822, cite this regulation.

American Diabetes Association (ADA) Standards of Care in Diabetes—2025

Published in Diabetes Care Vol. 48, Supplement 1, January 2025.

• Section 7 (Diabetes Technology):

- Real-time CGM (rt-CGM) or intermittently scanned CGM (isCGM) "should be offered" for diabetes management in adults with diabetes on MDI or CSII (continuous subcutaneous insulin infusion / pump therapy)—Level A evidence for Type 1 diabetes, Level A/B for Type 2 diabetes on insulin.

- CGM for non-insulin-treated Type 2 diabetes: Recommendation 7.14, Level B/E—CGM "may be offered" or "may be useful" in selected patients (driven by the MOBILE extension and IMMEDIATE trials).

- Automated insulin delivery (AID) systems (closed-loop): "should be offered to youth and adults with Type 1 diabetes"—Level A recommendation. The Standards explicitly state AID should be considered over non-automated pumps or MDI when appropriate.

• Section 15 (Management of Diabetes in Pregnancy): Real-time CGM is recommended in Type 1 diabetes pregnancy—Level B (based on the CONCEPTT trial).

ATTD International Consensus on Automated Insulin Delivery

Endocrine Reviews 2023;44(2):254–280 (Phillip M, Nimri R, et al.)

Recommends AID for all individuals with Type 1 diabetes regardless of baseline A1c, hypoglycemia history, or prior pump experience. Supports brand switching when clinically warranted (e.g., form factor, integration with preferred CGM, user preference). This consensus rebuts "must fail older pump first" step-therapy.

AACE 2024 Algorithm + Technology Statement

American Association of Clinical Endocrinology endorses AID as the preferred insulin delivery mode for Type 1 diabetes and selected insulin-using Type 2 diabetes patients, citing superior glycemic control and hypoglycemia reduction.

Pivotal Trials

• ADAPT (Automated Insulin Delivery Amongst Pregnant Women with Type 1 Diabetes): Demonstrated 1.4% A1c advantage of AID vs. MDI + CGM in pregnancy.
• MOBILE and IMMEDIATE: Showed benefit of CGM in non-insulin-treated Type 2 diabetes with hypoglycemia risk.
• CONCEPTT: Real-time CGM in pregnant women with Type 1 diabetes improved time-in-range and neonatal outcomes.
• Individual pivotal trials for each AID system (e.g., Omnipod 5 pivotal, t:slim X2 Control-IQ pivotal, MiniMed 780G pivotal, iLet Bionic Pancreas pivotal) are published and support FDA clearance; cite these when the insurer calls the device "experimental."

---

How to Argue Against Each Denial Reason

1. "Not medically necessary" / "Doesn't meet criteria"

If you're on basal-only insulin or fewer than 3–4 injections/day:

• Counter: Cite LCD L33822 April 2023 revision, which expanded CGM coverage to "all beneficiaries on any insulin including basal-only." If you have Medicare Advantage, cite 42 CFR §422.101(b)—MA plans cannot require more injections than the LCD.
• Counter: Cite ADA Standards 2025 Section 7—CGM "should be offered" for adults on MDI or CSII (Level A/B). Basal-bolus and basal-only both fall under this recommendation.
• Action: Ask your physician to document in the appeal letter: "Patient is on insulin therapy [specify regimen and total daily dose]. Per ADA 2025 Level A/B and Medicare LCD L33822 (2023), CGM is standard of care for all insulin-treated diabetes."

If the insurer says your A1c is "too good" or "not high enough":

• Counter: Neither Medicare LCD L33822 nor ADA 2025 impose A1c thresholds for CGM. In fact, ATTD 2023 consensus states AID (and by extension CGM) is appropriate "regardless of baseline A1c."
• Counter: A well-controlled A1c often reflects intensive self-management (frequent monitoring, insulin adjustments) that CGM will sustain or improve while reducing hypoglycemia risk and burden. Denying CGM because control is good punishes success.
• Action: Letter should state: "Patient's A1c of [X]% reflects diligent self-care with [current method]. CGM will maintain glycemic control, reduce hypoglycemia risk, and decrease daily burden. No guidelines require elevated A1c for CGM coverage; ADA 2025 and ATTD 2023 recommend CGM for all insulin-treated patients."

If you're on non-insulin Type 2 diabetes medications:

• Counter: LCD L33822 April 2023 allows CGM for non-insulin-treated diabetes if you have a history of problematic hypoglycemia (recurrent Level 2 <54 mg/dL or at least one Level 3 severe event requiring assistance). Sulfonylureas and meglitinides can cause severe hypoglycemia.
• Counter: ADA 2025 Recommendation 7.14 (Level B/E): CGM "may be offered" to non-insulin T2DM patients, especially those with hypoglycemia risk.
• Action: Document every hypoglycemic event <54 mg/dL with dates and glucose values, and any severe event requiring glucagon/EMS/ER. Letter: "Patient has documented problematic hypoglycemia [list events] on [medication]. Per LCD L33822 (2023) and ADA 2025, CGM is covered and medically appropriate."

If the insurer still cites a 4-fingerstick-per-day requirement:

• Counter: Medicare removed this requirement in July 2021 (LCD L33822 revision). Any policy citing it is outdated.
• Action: "The 4-fingerstick/day criterion was eliminated from LCD L33822 in July 2021. Patient meets current 2023–2025 LCD criteria [list them]."

---

2. "Experimental / investigational" (AID systems)

If Omnipod 5, t:slim X2 Control-IQ, Tandem Mobi, MiniMed 780G, or iLet is denied as experimental:

• Counter: All these devices are FDA-cleared Class II/III medical devices with published pivotal trials and real-world data. They are not experimental.
• Counter: ADA 2025 Section 7 gives AID a Level A recommendation for youth and adults with Type 1 diabetes. ATTD 2023 consensus (published in Endocrine Reviews) and AACE 2024 both endorse AID as preferred therapy.
• Action: List the device's FDA clearance date, pivotal trial publication, and the above guidelines. Example: "Tandem t:slim X2 with Control-IQ is FDA-cleared (2019), supported by published pivotal data (Brown SA et al., NEJM 2019), and recommended as Level A standard-of-care by ADA 2025 Section 7 and ATTD 2023. Labeling it 'experimental' contradicts ADA, ATTD, and AACE guidance."

If the insurer says AID is experimental for Type 2 diabetes:

• Counter: FDA clearances for Omnipod 5, t:slim X2, and MiniMed 780G include Type 2 diabetes (age 6+ or 7+ depending on device). AACE 2024 explicitly endorses AID for insulin-using Type 2 diabetes. ATTD 2023 notes AID benefits apply to Type 2 as well.
• Action: "Patient has Type 2 diabetes on [total daily insulin dose]. [Device] is FDA-cleared for Type 2 diabetes and recommended by AACE 2024 and ADA 2025. Denial contradicts FDA labeling and clinical guidelines."

If the insurer says AID is experimental in pregnancy:

• Counter: ADAPT trial (NEJM 2024) demonstrated AID safety and efficacy in pregnant women with Type 1 diabetes (1.4% A1c improvement vs. MDI+CGM). ADA 2025 Section 15 recommends rt-CGM in pregnancy (Level B); AID, which includes rt-CGM, is thus consistent with ADA guidance.
• Action: "Patient is pregnant with Type 1 diabetes. AID in pregnancy is supported by the ADAPT trial (2024) and ADA 2025 pregnancy guidelines. It is not experimental."

---

3. Step-therapy / "Try older model first"

If required to try MiniMed 770G before 780G, or Dexcom G6 before G7, or non-AID pump before AID:

• Counter: ATTD 2023 consensus and ADA 2025 do not mandate sequential device trials. ATTD explicitly supports brand switching and initial AID prescription when clinically appropriate. Step-therapy that delays access to superior technology (e.g., automated basal adjustments, smartphone connectivity, smaller form factor) exposes patients to preventable hypoglycemia and hyperglycemia.
• Counter: If the older device lacks a feature critical to your safety or adherence (e.g., no predictive low-glucose suspend, no smartphone alerts, requires separate receiver you can't carry), document this. Example: "Patient is a teacher who cannot carry a separate receiver; Dexcom G7's smartphone-only display is essential for discreet monitoring and adherence. G6 with receiver is not a reasonable alternative."
• Action: Letter: "Step-therapy is not supported by ADA 2025 or ATTD 2023. Patient's physician has determined [newer device] is medically appropriate and superior to [older device] for [specific reasons: automated insulin adjustments, alert customization, form factor, integration]. Requiring an older device trial delays evidence-based care and increases hypoglycemia risk."

---

4. Frequency limits or replacement denials

If sensors/pods denied for "too frequent" replacement:

• Counter: Manufacturer instructions specify wear time (e.g., 10 days for Dexcom G7 sensors, 3 days for Omnipod 5 pods). LCD L33822 Companion Article A52464 establishes supply allowances consistent with labeled use. Adhesive failures, sensor errors (e.g., "Signal Loss," calibration failures), and skin reactions are documented reasons for early replacement.
• Action: Keep logs of sensor/pod failures with dates and error messages. Letter: "Patient experienced [number] sensor failures in [timeframe] due to [adhesive failure/signal loss/skin reaction]. Replacement frequency aligns with manufacturer labeled use and LCD A52464 allowances. Arbitrary limits below labeled wear time violate coverage policy."

If pump replacement denied because warranty hasn't expired or "must wait 4–5 years":

• Counter: Pump warranty (typically 4 years) is a replacement guarantee, not a maximum coverage term. If your pump has failed (mechanical error, battery issues, software glitches), is out-of-warranty and malfunctioning, or lacks features now standard-of-care (e.g., no AID), replacement is medically necessary. ATTD 2023 supports transitioning to AID when clinically beneficial, regardless of prior pump age.
• Action: Document pump malfunctions (error codes, service calls, loaner pumps). If seeking upgrade to AID: "Patient's current pump [model, year] lacks automated insulin delivery. ADA 2025 Level A and ATTD 2023 recommend AID for Type 1 diabetes. Delaying upgrade to [new AID pump] denies access to standard-of-care technology that reduces A1c and hypoglycemia."

---

5. Missing documentation or visit requirements

If denial cites "no in-person visit within 6 months":

• Counter: LCD L33822 (2025) allows in-person or telehealth evaluation within 6 months prior to CGM order. If you had a telehealth visit, it counts.
• Action: Pull visit notes (in-person or telehealth) from within 6 months. Letter: "Patient was evaluated [in-person/via telehealth] on [date] by [physician], within the 6-month LCD L33822 requirement. Visit note documents diabetes management, insulin regimen, and CGM/pump indication."

If insurer claims hypoglycemia unawareness isn't "documented":

• Counter: Hypoglycemia unawareness is a clinical diagnosis. Gold or Clarke scores ≥4 are validated tools; alternatively, patient-reported inability to detect lows <70 mg/dL is sufficient. ADA 2025 and LCD L33822 recognize this as an indication for CGM.
• Action: Ask your physician to complete a Gold or Clarke questionnaire and document the score in your chart, or write: "Patient reports inability to perceive hypoglycemic symptoms until glucose <54 mg/dL, consistent with hypoglycemia unawareness (Gold score X). This is a Level A indication for CGM per ADA 2025."

If the insurer wants their proprietary form filled out:

• Counter: Insurers cannot require proprietary forms if medical records contain all necessary documentation (diagnosis, insulin regimen, visit date, clinical justification). However, pragmatically, having your physician complete the form and submit a detailed letter citing guidelines often speeds approval.
• Action: Submit both the form and a letter of medical necessity with the citations above. State: "All required clinical information is documented in the medical record and summarized in this letter per ADA 2025, LCD L33822, and ATTD 2023 guidelines."

---

What We Do

We help patients and physicians draft appeal letters and gather the documentation insurers actually respect—specific guideline citations, trial references, and policy excerpts that rebut each denial template. We don't write vague "this is medically necessary" letters; we write letters that cite ADA Standards of Care 2025 Section 7 Level A, Medicare LCD L33822 April 2023 revision, ATTD 2023 consensus (Endocrine Reviews 44(2):254–280), and device-specific pivotal trials. Our approach turns a frustrating, time-consuming appeal process into a clear, evidence-backed argument that medical directors find difficult to dismiss.

---

Sources

1. Centers for Medicare & Medicaid Services. Local Coverage Determination (LCD) L33822 – Glucose Monitors. Original 2017; revised July 2021, April 2023 (effective 2023-04-16). Companion Article A52464.

https://www.cms.gov/medicare-coverage-database/view/lcd.aspx?lcdid=33822

2. Code of Federal Regulations Title 42 §422.101(b) – Medicare Advantage plan requirements (parity with original Medicare LCDs).

3. American Diabetes Association. Standards of Care in Diabetes—2025. Diabetes Care 2025 Jan;48(Suppl 1). Section 7 (Diabetes Technology) and Section 15 (Management of Diabetes in Pregnancy).

https://diabetesjournals.org/care/issue/48/Supplement_1

4. Phillip M, Nimri R, et al. ATTD International Consensus on Automated Insulin Delivery. Endocrine Reviews 2023;44(2):254–280.

https://doi.org/10.1210/endrev/bnac022

5. American Association of Clinical Endocrinology. AACE 2024 Algorithm for Glycemic Control and Diabetes Technology Statement.

6. Lee TTM, Collett C, Bergford S, et al. Automated Insulin Delivery in Women with Pregnancy Complicated by Type 1 Diabetes (ADAPT). New England Journal of Medicine 2024 (ADAPT trial).

7. Feig DS, Donovan LE, Corcoy R, et al. Continuous Glucose Monitoring in Pregnant Women with Type 1 Diabetes (CONCEPTT). Lancet 2017;390(10110):2347–2359.

8. Aleppo G, et al. MOBILE study: CGM in non-insulin-treated Type 2 diabetes. JAMA 2017;317(4):379–387; extension data in Diabetes Technology & Therapeutics 2021.

9. Yoo HJ, et al. IMMEDIATE trial: CGM in Type 2 diabetes. Diabetes Care 2023.

10. Brown SA, Kovatchev BP, Raghinaru D, et al. Six-Month Randomized, Multicenter Trial of Closed-Loop Control in Type 1 Diabetes (t:slim X2 Control-IQ pivotal). New England Journal of Medicine 2019;381:1707–1717.

11. FDA device clearance letters and pivotal trial summaries for Dexcom G6/G7, FreeStyle Libre 2/3 Plus, Eversense 365, Omnipod 5, Tandem t:slim X2/Mobi, Medtronic MiniMed 780G, and Beta Bionics iLet. Available at https://www.fda.gov/medical-devices.