Browser-native DOM injection writes endocrine notes into Cerbo EMR without server APIs
Endocrinology documentation inside Cerbo EMR rarely breaks down at the clinical reasoning step. It breaks down at the transcription step, when a clinician who has just interpreted two weeks of continuous glucose data, adjusted basal insulin, and reviewed a thyroid panel must retype all of it into the chart. Merry AI removes that retyping step.
Merry AI runs as a Chrome Extension content script inside the clinician's already authenticated Cerbo session. The extension locates Cerbo's editable note regions, including Assessment, Plan, Supplements, and Lab Review, by querying the page DOM for their visible section headers. After the clinician reviews the draft in the overlay, Merry writes each section into its matching textarea or contentEditable field. No Cerbo API credentials, marketplace listing, or server-side connector is involved at any point.
A 14-day CGM summary belongs in the Assessment, not in an unread attachment
Every diabetes follow-up note should state the same core glycemic metrics in the same order, so trends are readable across visits and defensible during audit. Merry's Cerbo endocrinology template records the following fields by default:
- Time in Range (TIR) percentage: the share of readings between 70 and 180 mg/dL over the most recent 14 days, with the consensus target above 70% for most adults.
- Time below 70 mg/dL: reported as a percentage with a target under 4%, plus time below 54 mg/dL with a target under 1%.
- Glucose Management Indicator (GMI) value: calculated as 3.31 + 0.02392 × mean glucose in mg/dL, documented alongside the most recent laboratory A1c for comparison.
- Sensor wear and data sufficiency: CGM active time, with at least 70% wear over 14 days required before the summary is treated as representative.
- Insulin dose adjustments by component: basal units before and after, insulin-to-carbohydrate ratio changes, and correction factor changes, each tied to the glucose pattern that justified it.
Structured insulin titration language matters because a line such as basal glargine reduced from 24 to 21 units for 6% time below 70 mg/dL with nocturnal clustering documents the clinical decision, the data, and the risk in one sentence. That sentence is what a reviewer looks for when confirming moderate or high medical decision-making.
Endocrine diagnostic thresholds stay explicit so coders can verify medical necessity
Merry templates cite the criteria that support each diagnosis rather than listing the diagnosis alone. Common thresholds written into the injected Assessment include:
- Type 2 diabetes diagnostic values: A1c of 6.5% or higher, fasting plasma glucose of 126 mg/dL or higher, or 2-hour OGTT glucose of 200 mg/dL or higher, with confirmatory testing noted.
- Prediabetes range documentation: A1c of 5.7% to 6.4%, fasting glucose of 100 to 125 mg/dL, or 2-hour OGTT of 140 to 199 mg/dL.
- Thyroid function interpretation: TSH paired with free T4, and antibody or imaging results when Graves disease or toxic nodular goiter is under evaluation.
- Bone density risk stratification: DEXA T-score of -2.5 or lower for osteoporosis, T-score between -1.0 and -2.5 for osteopenia, and FRAX 10-year fracture probability.
- PCOS Rotterdam criteria count: at least two of ovulatory dysfunction, hyperandrogenism, and polycystic ovarian morphology, with mimicking disorders excluded.
Cerbo chart parts already hold many of these values as discrete vitals and lab variables. Merry's narrative references those values in the Lab Review and Assessment text, so the narrative and the discrete data agree.
Chrome DOM injection places field-level notes into Cerbo with no IT project
The three available integration paths differ sharply in setup burden, field placement, and data retention. The comparison below reflects a typical independent endocrinology practice on Cerbo.
| Criterion | Traditional API Marketplace Integration | Manual Copy-Paste | Merry AI Chrome DOM Injection |
|---|---|---|---|
| Setup time | Weeks to months of vendor contracting and IT review | None | Under 10 minutes: install extension, sign in |
| Cerbo API credentials | Required | Not required | Not required |
| Marketplace or interface fees | Often recurring per provider | None | None |
| Field-level placement | Usually a single note blob | Manual, section by section | Assessment, Plan, Supplements, and Lab Review mapped individually |
| Audio retention | Often stored in vendor cloud | Depends on external scribe tool | RAM only, destroyed after note generation |
| Clinician attestation | Preserved | Preserved | Preserved inside native Cerbo signing |
| Context switching | Low | High, two windows per visit | None, overlay sits inside the Cerbo chart |
Installing Merry in Cerbo takes three steps and no administrator ticket
The clinician workflow is short and stays inside the browser window the clinician already uses for charting:
- Install the Merry Chrome Extension from the Chrome Web Store and sign in with your Merry account.
- Open the Cerbo chart note for the patient encounter, so the Assessment, Plan, and Lab Review editors are visible on screen.
- Insert the reviewed Merry draft after editing it in the overlay. Each section writes into its mapped Cerbo field, ready for your signature.
Stable section headers keep mapping reliable. Practices that rename Cerbo note sections should update the matching Merry template headers at the same time.
G2211 capture depends on documenting the longitudinal relationship, not visit length
HCPCS add-on code G2211 applies when the practitioner serves as the continuing focal point for a patient's care or manages a single serious or complex condition over time. It is reported with office or outpatient E/M codes 99202 to 99205 and 99211 to 99215, and with eligible home or residence E/M codes. Endocrinology fits this definition closely: insulin-requiring diabetes with CKD, thyroid cancer surveillance, pituitary and adrenal disorders, and high-risk osteoporosis are all longitudinal by nature.
Merry's Plan section states continuity explicitly, naming the chronic condition, prior titration history, and the next scheduled reassessment. That language gives the billing team a documented basis for G2211 rather than an assumption.
Modifier 25 changes the G2211 analysis. G2211 is generally not payable when the base E/M carries modifier 25, except when the same-day service is an annual wellness visit, vaccine administration, or a covered Part B preventive service. When a separately identifiable E/M occurs on the same day as a procedure such as a thyroid fine-needle aspiration, Merry separates the E/M Assessment and Plan from the Procedure Note under distinct headings.
Modifier 59 never belongs on an E/M code. When two distinct procedures occur, such as biopsies of separate thyroid nodules, Merry documents each indication, anatomic site, and session so coders can choose modifier 59 or a more specific XS or XU modifier.
RAM-only audio processing leaves no recording outside the Cerbo chart
Encounter audio exists only in volatile memory while Merry generates the draft. Once the note is produced, the audio is destroyed, and no recording or transcript is retained on Merry servers. The only durable artifact is the clinician-attested note inside Cerbo, operating under an executed BAA.
State AI statutes reinforce clinician authorship as the record of truth
California and Texas AI laws point in the same direction: a licensed clinician must remain accountable for clinical content. California SB 1120 requires that medical necessity determinations by health plans be made by licensed clinicians rather than by algorithms, which raises the value of notes that state criteria clearly. California AB 3030 governs disclosure of generative AI in patient communications and exempts content reviewed by a licensed provider. Texas HB 1709 addresses AI governance and should be reviewed by compliance counsel against your specific deployment.
Clinician attestation remains the final control on every injected note. Merry drafts, the clinician edits and signs inside Cerbo, and the chart records the clinician as author of record.




