Skip to content

CMS & Regulatory Compliance

Texas Health and Safety Code 183.005

Texas Health & Safety Code §183.005: Multi-Specialty

AI-assisted Texas notes require licensed scope, TMB-standard review, and patient disclosure. Book your audit at https://cal.com/merryai/demo.

Executive Key Takeaways

  • Texas Health and Safety Code §183.005 permits AI use in the EHR only when the practitioner acts within licensed scope, the use is lawful, and every AI-created record is reviewed under Texas Medical Board standards
  • Section 183.005(b) requires disclosure of AI use to patients, which should be logged per encounter across every specialty
  • An affirmative physician attestation checkpoint before EHR note lock creates a discrete review artifact stronger than a default signature
  • Zero offshore data transmission and RAM-only processing narrow exposure under Chapter 183 attorney general enforcement
  • Validation via Access Clinical Prompts at templates.scribing.io
Regulatory Verification Framework
2026 Audit Ready
Tex. Health & Safety Code §183.005(a)(1)Verified Compliant

Practitioner Scope of License

Practitioner must act within the scope of license, certification, or authorization regardless of AI use

Tex. Health & Safety Code §183.005(a)(2)Verified Compliant

Lawful AI Technology Authorization

Particular AI use must not be otherwise restricted or prohibited by state or federal law

Tex. Health & Safety Code §183.005(a)(3)Verified Compliant

TMB Medical Record Standards (22 TAC §165.1)

Practitioner must review all records created with AI consistent with Texas Medical Board medical record standards (22 TAC §165.1)

Tex. Health & Safety Code §183.005(b)Verified Compliant

Patient AI Technology Disclosure

Practitioner must disclose the use of AI technology to patients

Statutory Compliance Architecture for Texas Health and Safety Code 183.005

Documenting clinical encounters under Texas Health and Safety Code §183.005 requires a specific sequence: the practitioner acts within licensed scope, confirms the AI use is not otherwise restricted, reviews every AI-created record against Texas Medical Board standards, and discloses AI use to the patient. In a multi-specialty group, that sequence must hold identically across behavioral health, cardiology, orthopedics, and pediatrics, even though each specialty documents different diagnostic evidence.

Chapter 183 carries real enforcement. The Texas attorney general may pursue injunctive relief and civil penalties for violations of the chapter, which converts documentation workflow design from an operational preference into a liability control. Merry AI treats each statutory element as an engineering requirement with a corresponding checkpoint, log entry, or architectural constraint.

Scope-of-Practice Boundaries Under §183.005(a)(1)

The platform drafts, physicians decide. Merry AI structures dictated history, examination findings, and patient statements into the note, but it does not issue diagnoses, select codes, append modifiers, or enter orders. Diagnostic conclusions remain the act of the licensed practitioner, which keeps the clinician within scope regardless of AI involvement, as §183.005(a)(1) contemplates.

Physician Review and Attestation Under §183.005(a)(3)

Subsection (a)(3) requires review of all records created with artificial intelligence, consistent with Texas Medical Board medical record standards at 22 TAC §165.1. Relying on a default EHR signature is a weak evidentiary position, because a signature alone does not demonstrate that AI-generated content was examined.

Merry AI inserts an affirmative physician attestation checkpoint before EHR note lock. The dialog summarizes which sections were AI-drafted, requires explicit confirmation, and records the timestamp, user, and note identifier. This produces a discrete artifact showing review occurred, rather than an inference drawn from routine signing.

Specialty Diagnostic Scaffolds Across Multi-Specialty Practice

Diagnostic criteria function as documentation scaffolds, not decision engines. In behavioral health, Merry AI organizes symptom clusters, duration, functional impairment, and PHQ-9, GAD-7, or C-SSRS results under DSM-5-TR-aligned headings, leaving diagnosis selection and rule-outs to the clinician. In cardiology, it surfaces recorded LVEF values and exertional symptom descriptions, then prompts the physician to assign NYHA class explicitly rather than inferring it.

Orthopedic and pediatric notes follow the same rule. Quantified range of motion, 0–5 strength grading, and special tests such as Lachman or McMurray are transcribed only when dictated; measurements are never fabricated. Pediatric notes reflect Bright Futures periodicity, growth percentiles, milestones, and screening results drawn from the encounter itself.

Patient Disclosure, Data Handling, and EHR Write-Back

Section 183.005(b) mandates disclosure of AI use to patients. Merry AI provides configurable disclosure language for visit summaries, intake consent, and portal notices, and logs that disclosure occurred for each encounter, so the practice can demonstrate compliance per visit rather than by policy statement alone.

Data handling follows a conservative posture. Although §183.005 does not itself address data residency, Merry AI enforces zero offshore data transmission and RAM-only processing, with no vendor-side persistence of audio or note content. This narrows breach exposure under Chapter 183 and parallels HIPAA Security Rule technical safeguards at 45 CFR §164.312.

DOM, iframe, and Multi-Tab Audit Controls

Write-back is restricted by design. Merry AI maintains a versioned whitelist of DOM selectors per EHR and per note type, distinguishing progress, psychotherapy, and operative notes. It respects iframe boundaries, re-binds on tab and route changes, and never injects into orders, problem lists, or billing screens. Each injection event is logged with timestamp, user, note ID, and field selector, addressing any allegation of unsupervised AI modification of the record.

2026 CMS Documentation Alignment and Statutory Comparison

CMS data underscore the documentation stakes. The 2024 Medicare improper payment rate for E/M services was 10.3%, with incorrect coding and insufficient documentation driving most errors. For 2026, HCPCS G2211 extends to home or residence E/M visits (99341–99350) and may be paid alongside an annual wellness visit, vaccine administration, or Part B preventive service by the same practitioner. Merry AI surfaces a longitudinal complexity synopsis for the clinician but never appends G2211 automatically.

NCCI modifier 25 and 59 require demonstrable distinctness. The note must separate a significant, separately identifiable E/M service from same-day procedure evaluation, and must capture lesion-specific location and rationale where modifier 59 or XS applies. Merry AI structures that separation; the clinician and billing team assign modifiers.

Compliance DimensionManual ChartingStandard Generic AI ScribesMerry AI Compliance Architecture
§183.005(a)(1) scopePhysician authors all contentMay suggest diagnoses or codesDrafts documentation only; no diagnoses, codes, or orders
§183.005(a)(3) reviewImplicit through authorshipDefault EHR signatureAffirmative attestation before note lock
§183.005(b) disclosureNot applicableOften undocumentedConfigurable language with per-visit logging
Data residencyLocal EHR onlyFrequently global cloud routingZero offshore transmission; RAM-only processing
Write-back auditabilityManual entryCopy-paste or brittle injectionVersioned selector whitelist with injection logs
G2211 and modifier 25/59Depends on clinician timeGeneric narrativeComplexity synopsis and E/M–procedure separation

Clinicians reviewing the underlying evidence on ambient documentation accuracy can consult peer-reviewed literature at https://www.ncbi.nlm.nih.gov/pmc/. Practices preparing for a Texas review should map each subsection of §183.005 to a specific control before the next audit cycle.

To evaluate your current workflow, Schedule a 15-Minute Workflow Audit and we will trace one encounter per specialty from capture through attestation and note lock.

Regulatory & Compliance FAQ

What does Texas Health and Safety Code §183.005 require when AI assists with multi-specialty documentation?

Section 183.005 permits a practitioner to use artificial intelligence for diagnostic purposes, including diagnosis or treatment recommendations drawn from the medical record, only when three conditions hold: the practitioner acts within licensed scope, the use is not otherwise restricted by state or federal law, and the practitioner reviews all AI-created records consistent with Texas Medical Board standards. Subsection (b) separately requires disclosure of AI use to patients, regardless of specialty.

Does a standard EHR signature satisfy the review requirement in §183.005(a)(3)?

A signature shows the note was finalized, but it does not independently demonstrate that AI-generated content was examined against Texas Medical Board record standards under 22 TAC §165.1. Merry AI adds an affirmative attestation checkpoint before note lock that identifies AI-drafted sections, requires explicit physician confirmation, and logs timestamp, user, and note identifier. That discrete artifact gives the practice clearer evidence of review if the attorney general or a payer examines the record.

How should patient disclosure under §183.005(b) be documented in a multi-specialty practice?

The statute requires disclosure of AI use but does not prescribe a single format, so practices should adopt a consistent, documented method. Merry AI supports configurable disclosure language in intake consent, after-visit summaries, and portal notices, and records that disclosure occurred for each encounter. Per-visit logging allows a multi-specialty group to demonstrate compliance for behavioral health, cardiology, orthopedic, and pediatric visits alike, rather than relying on a general policy statement. Counsel should approve final wording.

Can the AI assign HCPCS G2211 or modifiers 25 and 59 on behalf of the physician?

No. Coding and modifier assignment remain clinician and billing-team decisions, consistent with the scope-of-practice condition in §183.005(a)(1). Merry AI surfaces a longitudinal complexity synopsis supporting G2211, which CMS extends to home or residence E/M visits in 2026, and structures the narrative to separate a significant, separately identifiable E/M service from same-day procedures. It captures lesion-specific detail for modifier 59 or XS, but never appends codes automatically.

Where can clinics verify compliant templates?

Clinics can review specialty-specific prompt structures at https://templates.scribing.io, which organize documentation around DSM-5-TR criteria, NYHA class and LVEF, orthopedic range of motion, and Bright Futures periodicity. Each template keeps diagnostic conclusions with the clinician, consistent with §183.005(a)(1), and assumes the attestation checkpoint required for (a)(3) review. Practices should validate templates against their own EHR note types and Texas Medical Board record standards before deployment.

Explore More Knowledge

Related Clinical Rules & Compliance Guides

View all clinical rules & compliance
Illinois Eavesdropping Act: Family Medicine Scribing
CMS & Regulatory Compliance

Illinois Eavesdropping Act: Family Medicine Scribing

Capture all-party verbal consent under 720 ILCS 5/14-2 in every family medicine visit. Book your audit at https://cal.com/merryai/demo.

Dr. Sarah Chen, MD•7 min read
Read guide
Washington MHMDA Compliance for Psychiatry Scribing
CMS & Regulatory Compliance

Washington MHMDA Compliance for Psychiatry Scribing

Separate RCW 19.373.030 consent before psychiatric recording; zero audio retention. Book your audit at https://cal.com/merryai/demo.

Dr. Sarah Chen, MD•7 min read
Read guide
Illinois Eavesdropping Act: Multi-Specialty Compliance
CMS & Regulatory Compliance

Illinois Eavesdropping Act: Multi-Specialty Compliance

Timestamp all-party verbal consent in chart before capture; audio shreds from RAM at 0 days. Book your audit at https://cal.com/merryai/demo.

Dr. Sarah Chen, MD•7 min read
Read guide