CharmHealth pediatric encounters already expose every field Merry AI needs inside the browser
Clinical documentation inside CharmHealth rarely fails because pediatricians lack dedication. It fails because a single well-child visit asks for growth interpretation, Bright Futures anticipatory guidance, immunization counseling, and caregiver concerns inside a short appointment slot. CharmHealth renders its template-driven SOAP note, its left-pane vitals, and its Growth Chart as standard web page elements. That design means a Chrome Extension can read and write them inside the clinician's own logged-in session.
Merry AI's Chrome Extension runs as a content script scoped only to the CharmHealth domain. It locates the Subjective, Objective, Assessment, and Plan regions by their section labels, then writes the clinician-approved draft into each field through the DOM and fires standard input events so CharmHealth registers the change. Nothing reaches CharmHealth servers except the note the clinician saves, exactly as if it had been typed by hand.
Browser-native DOM mapping bypasses EHR marketplace paywalls entirely
Traditional EHR integrations depend on API credentials, vendor partnership agreements, and interface engine work. For an independent pediatric practice, that path often means recurring marketplace fees and an IT project before a single note is written. Manual copy-paste avoids those costs but moves the burden back onto the clinician, one field at a time.
| Dimension | Traditional API Marketplace Integration | Manual Copy-Paste | Merry AI Chrome DOM Injection |
|---|---|---|---|
| Setup time | Weeks to months of vendor approval and interface work | None | Minutes: install extension and sign in |
| IT involvement | Required for credentials, interfaces, and testing | None | None beyond optional Chrome allowlisting |
| Recurring integration fees | Marketplace or per-provider API fees are common | None | No EHR API fees |
| Field-level placement | Depends on which endpoints the vendor exposes | Clinician pastes into each section | Each SOAP section written to its matching field |
| Growth z-score handling | Rarely supported for pediatric references | Clinician reads the chart and types values | Computed, source-cited, and compared to prior visit |
| Error surface | Interface mapping failures | Wrong-field and wrong-patient pastes | Insertion halts if a field label is not found |
| Clinician attestation | Varies by vendor workflow | Clinician signs as usual | Draft stays unsigned until clinician attests |
The practical difference shows up at the end of clinic. A pediatrician who sees 25 children a day avoids dozens of section-by-section pastes, and the growth paragraph lands in Objective while counseling lands in Plan, where an auditor expects to find them.
Every growth statement names its reference standard and compares against the prior visit
A growth number without context cannot support clinical judgment or an audit. Merry's pediatric template reads age, sex, weight, length or height, and BMI from the CharmHealth vitals pane, computes z-scores against the correct reference, and pulls the most recent prior encounter value for comparison.
Reference selection follows CDC guidance. Children under 24 months are measured against the WHO 2006 Child Growth Standards using weight-for-age and weight-for-length. Children from 2 to 20 years are measured against the CDC 2000 growth charts using weight-for-age and BMI-for-age. Above the 95th BMI percentile, Merry applies the CDC 2022 extended BMI-for-age charts and also reports BMI as a percentage of the 95th percentile, because standard z-scores compress at the extremes.
A typical injected Objective statement reads: Weight-for-age z = 0.42 (66th percentile, CDC 2000), up from z = 0.31 at the prior well visit. BMI-for-age z = 0.85 (80th percentile, CDC 2000), previously z = 0.62. Trajectory remains within one major percentile line. The source, the current value, and the prior value appear in one sentence, so the reviewer never has to open the Growth Chart to verify the claim.
When a trajectory crosses two major percentile lines between visits, Merry writes a flagged statement into the Assessment field rather than a routine sentence, and it proposes the matching Bright Futures counseling language for the Plan field. The clinician decides whether the finding is clinically meaningful.
Merry cross-checks every computed value against the percentiles CharmHealth displays on its own Growth Chart. If the two disagree, insertion pauses and the discrepancy is shown to the clinician. This prevents two conflicting growth numbers from ever coexisting in the same chart.
Clinicians move from install to first inserted note in three steps
- Add the Merry extension to Chrome from the Chrome Web Store and sign in with your Merry AI account.
- Open the CharmHealth encounter for the patient. Merry detects the SOAP template and vitals pane automatically and prepares the draft after the visit.
- Click Insert to write the reviewed draft into each SOAP field, then review, edit, and sign inside CharmHealth as usual.
No firewall changes, admin consoles, or CharmHealth configuration are required. Practices that manage Chrome centrally can allowlist the extension through Google Admin policy, and Merry provides its security documentation and an executed BAA for compliance review.
Preventive and problem-oriented documentation must stay visibly separate to support modifier 25
Pediatric visits frequently combine a well-child preventive service, vaccine administration, and evaluation of a chronic problem such as obesity or asthma. NCCI policy permits modifier 25 only when the E/M service is significant and separately identifiable from the other service on the same date. Merry places problem-oriented content in its own labeled Assessment and Plan block so the separate work is visible rather than buried in the preventive narrative.
G2211 is a Medicare add-on code, so in most pediatric panels it applies only to Medicare-covered children, such as those with end-stage renal disease. Medicaid and commercial plans set their own add-on policies. CMS allows G2211 with a modifier 25 office E/M on the same day as vaccine administration or a Part B preventive service, and from January 1, 2026 that allowance extends to home or residence E/M codes. Longitudinal growth trends strengthen the complexity narrative, but they do not establish eligibility on their own. Merry never generates modifier 59 language, and coding decisions remain with the clinician and billing team.
Adolescent confidentiality and final attestation remain under clinician control
Adolescent confidential content stays separate. Psychosocial screening disclosures, such as HEADSS findings, route to a clinician-controlled confidential block, and psychotherapy notes as defined under HIPAA at 45 CFR 164.501 are never merged into the general SOAP note. The clinician decides what enters the shared record and what caregiver portal views can display.
State AI laws add disclosure obligations. California AB 3030 requires disclosure on generative AI patient communications unless a licensed provider reviews them, and every Merry draft is clinician-reviewed before use. California SB 1120 governs payer utilization review rather than clinical documentation, and Texas HB 1709 was a proposed AI governance framework. Multi-site pediatric networks in either state should obtain counsel review before scaling.
Every injected draft remains unsigned until the clinician reviews, edits, and attests. Merry writes text into fields. It does not sign, submit claims, or alter prior encounters.
DOM injection works best when its maintenance cost is stated openly
Selector maintenance is the real cost of browser-native integration. When CharmHealth updates its interface, field labels or layouts can shift. Merry monitors CharmHealth releases, and if a section label cannot be found, the extension stops and copies the draft to the clipboard rather than guessing a destination. A missed insertion is recoverable. A note in the wrong field is not.




