Addiction documentation in Sigmund Software breaks down at the group session
Clinical documentation inside Sigmund Software rarely fails because clinicians lack dedication. It fails because a single IOP process group can produce 8 to 12 separate progress notes. Each one must show individualized participation, response to intervention, and current risk status rather than cloned group text.
Sigmund's AURA platform already provides logic-based assessments, customizable templates, auto-fill, electronic signatures, and its own integrated AI note tools. Merry AI is an independent, clinician-side tool that works alongside that environment. It is not a Sigmund feature, and Sigmund does not publicly document group-audio diarization within its own product.
Merry AI writes into Sigmund fields through the browser, not through a server API
The Chrome Extension reads the rendered page in the active Sigmund tab, confirms the client and encounter shown on screen, and maps each generated note section to the visible narrative fields, such as Interval History, Assessment, Interventions, and Plan. The text is placed into those fields client-side, in the same location a clinician would type.
Because the write happens inside the clinician's own authenticated session, Sigmund's existing save behavior, required-field validation, and audit logging apply exactly as they do for manually typed text. Merry AI does not alter structured fields that Sigmund's logic-based assessments control, so branching logic and signability rules stay intact.
No marketplace listing, interface engine, or vendor API key is involved. There is no HL7 or FHIR connection to provision, no per-transaction API fee, and no firewall change. The integration surface is the browser window the clinician already has open.
Browser-native DOM mapping bypasses EHR marketplace paywalls entirely
| Factor | Traditional API Marketplace Integration | Manual Copy-Paste | Merry AI Chrome DOM Injection |
|---|---|---|---|
| Setup time | Weeks to months of vendor review | None | Minutes per workstation |
| Recurring cost | Marketplace and per-call API fees | None | No API fees |
| IT involvement | Interface engineering and credentials | None | Chrome install or policy allowlist |
| Field accuracy | Depends on exposed endpoints | Error-prone, manual placement | Section-to-field mapping on the open chart |
| Wrong-patient risk | Low, ID-bound | High in multi-tab charting | Blocked by on-screen client ID check before each write |
| Group session support | Rarely exposed for group notes | One note retyped per client | One verified draft written per participant |
| Audit trail | Vendor-dependent | Native, but no draft history | Native Sigmund audit plus clinician attestation |
The comparison matters most for small and mid-size addiction programs that cannot fund marketplace integration fees yet carry heavy group documentation loads. Copy-paste costs nothing up front but creates wrong-chart risk every time a counselor has several client tabs open.
Group session diarization produces one draft per participant, never one note cloned eight times
The diarization engine segments group audio by speaker, then attributes segments to participants using the session roster the clinician confirms at the start of group. Each attendee receives a separate draft containing shared elements (topic, modality, facilitator interventions) and individualized elements (participation level, statements tied to treatment goals, disclosed cravings or triggers).
Speaker attribution is probabilistic, so every draft is flagged for clinician review. Clinicians confirm which statements belong to which client and add mental status and safety observations that speech alone cannot capture. Only then does the extension write each approved draft into that client's own Sigmund note, one chart at a time.
Before every individual chart write, the extension compares the client identifier displayed in the open Sigmund chart against the draft's assigned participant. If they do not match, the write is blocked. This directly addresses the multi-tab error of placing one client's group note into another client's record.
Addiction notes must anchor to DSM-5-TR criteria and ASAM dimensions, not generic SOAP structure
Substance use disorder documentation needs substance type, route, frequency, quantity, severity specifier, co-occurring conditions, and functional impact on work, relationships, legal status, and overdose risk. Merry AI addiction templates prompt for each element and align narrative content with scores already recorded in Sigmund assessments, such as withdrawal scales or craving ratings, rather than inventing values.
Level of care decisions depend on multidimensional assessment across withdrawal potential, biomedical and psychiatric conditions, readiness to change, relapse risk, and recovery environment. Templates keep narrative tied to the clinician's recorded findings so that placement and continued-stay decisions remain defensible on review, consistent with CMS behavioral health documentation guidance.
Federal confidentiality under 42 CFR Part 2 adds a layer beyond HIPAA for SUD treatment records. Merry AI templates keep psychotherapy process content and sensitive disclosures in distinct sections, so programs can route them according to their Part 2 consent and psychotherapy note policies instead of merging them into the general progress note.
G2211 applies to addiction medicine E/M visits, not to group psychotherapy codes
HCPCS add-on code G2211 describes the complexity of longitudinal care for a single serious or complex condition. It is reported only with office or outpatient E/M codes 99202 through 99205 and 99211 through 99215, and, from January 1, 2026, with home or residence codes 99341, 99342, 99344, 99345, and 99347 through 99350, per the CMS G2211 FAQ. An addiction physician managing buprenorphine maintenance over many months often fits this pattern.
G2211 does not attach to group psychotherapy code 90853. CMS also states that G2211 is not payable when the base E/M carries modifier 25, except when the other same-day service is a preventive service, immunization administration, or annual wellness visit, as described in CMS MLN006764.
Merry AI surfaces existing clinical evidence of the ongoing relationship, such as dose adjustments, toxicology review, and relapse prevention planning, but it never inserts billing-justification boilerplate or selects codes. CMS cautions against cloned and auto-populated text in its EHR features decision table, and code selection remains the clinician's decision.
Clinician attestation remains the final control in every Sigmund chart
No draft reaches the signed record without clinician review, edits, and attestation inside Sigmund. California programs should also maintain AI governance policies and apply AB 3030 disclosures wherever AI-generated communications reach patients. Programs in Texas and other states should confirm current AI and telehealth requirements with counsel before rollout.
Programs evaluating this workflow can start with addiction and behavioral health templates at templates.scribing.io, then map them to their Sigmund note configuration during a 15-minute workflow audit.




