Skip to content

CMS & Regulatory Compliance

California SB 1120

California SB 1120 Compliance for Pain Management

Capture dated failed therapies and ROM metrics UR physicians must weigh under SB 1120. Book your audit at https://cal.com/merryai/demo.

Executive Key Takeaways

  • Under Cal. Health & Safety Code § 1367.01(k) and Insurance Code § 10123.135(j), plan AI tools must base utilization review on the documented record and cannot deny, delay, or modify pain care on medical necessity grounds
  • Objective severity metrics such as ROM degrees, NRS scores, and Oswestry percentages, together with dated failed therapies, form the evidentiary core of a successful UR appeal
  • Merry AI writes structured exam values into Tebra through Chrome DOM injection only after clinician confirmation, which keeps the physician as author of record
  • Encounter audio is held in RAM only and shredded after transcription, so the signed chart is the sole durable record
  • Validation via Access Clinical Prompts at templates.scribing.io
Regulatory Verification Framework
2026 Audit Ready
Cal. Health & Safety Code § 1367.01(k)(1)Verified Compliant

health plan AI

health plan AI, algorithm, or software tools must base UR/UM determinations on enrollee medical history and individual clinical circumstances

Cal. Insurance Code § 10123.135(j)(1)Verified Compliant

parallel AI utilization review requirements for disability insurers

parallel AI utilization review requirements for disability insurers, including fair, equitable, and non-discriminatory application

Cal. Health & Safety Code § 1367.01(e) and (k)(2)Verified Compliant

denials

denials, delays, or modifications based on medical necessity reserved to licensed physicians or competent licensed health care professionals

Cal. Health & Safety Code § 1367.01(h)Verified Compliant

UR decision timeframes and written denial communication identifying criteria and the responsible health care professional

UR decision timeframes and written denial communication identifying criteria and the responsible health care professional

Statutory Compliance Architecture for California SB 1120

Documenting pain management encounters under California SB 1120 means writing for two readers at once: the treating physician and the insurer's utilization review apparatus. Since January 1, 2025, Health and Safety Code § 1367.01(k) and Insurance Code § 10123.135(j) have required that any artificial intelligence, algorithm, or software tool used by a health plan or disability insurer base its determination on the enrollee's medical history, the individual clinical circumstances presented by the requesting provider, and other relevant clinical information in the record.

The statute draws a firm line around authority. A tool may support review, but it cannot deny, delay, or modify services based on medical necessity. Only a licensed physician or a licensed health care professional competent in the clinical issues may make that determination. In interventional pain, epidural steroid injections, radiofrequency ablation, and long-term opioid therapy routinely trigger review. In those cases, the quality of the chart largely determines whether the human reviewer has enough objective evidence to approve.

Objective Severity Metrics as Statutory Evidence

Population-level denial rules fail when the record contains specific, dated, quantified findings. Plan tools are obligated to consider facts such as lumbar flexion of 45 degrees, a Numeric Rating Scale score of 8/10, an Oswestry Disability Index of 52 percent, and a straight leg raise positive at 40 degrees. Narrative phrases such as 'limited motion' or 'severe pain' give a reviewer nothing to weigh against an algorithmic threshold.

Dated Failed Therapies and Step Therapy Records

Conservative care documentation must be chronological and verifiable. Each failed therapy should carry a start date, end date, dose or frequency, and reason for discontinuation. Examples include six weeks of physical therapy ending 2026-09-01, naproxen 500 mg twice daily discontinued for gastrointestinal intolerance, and duloxetine 60 mg discontinued for inadequate response. When these entries live in discrete fields rather than prose, an appeal letter can cite them line by line.

Comparative Liability Across Documentation Models

Documentation method determines appeal strength. The table below compares three documentation models across the dimensions SB 1120 makes material.

Compliance DimensionManual ChartingStandard Generic AI ScribesMerry AI Compliance Architecture
Objective severity captureVariable; depends on clinician recallMetrics embedded in narrative proseROM degrees, NRS, ODI written to discrete fields
Failed therapy chronologyOften undated or summarizedParaphrased without reliable datesTime-stamped entries with dose and stop reason
UR tool legibility under § 1367.01(k)Low; requires manual abstractionModerate; free-text parsing requiredHigh; structured, queryable evidence
Physician authorshipClearBlurred by generated assessmentsInjection only after clinician confirmation
Audio retentionNoneOften retained for model improvement0-day RAM-only buffer, shredded post-transcription
Audit trail integrityNative EHRSingle pasted note bodyNative Tebra save and sign workflow preserved

Tebra DOM Injection for Structured Exam Data

Field-Level Writing Through the Chrome Extension

Merry AI writes structured exam data directly into Tebra's existing web charting inputs through a Chrome content script. The extension first locates the active note container. It then resolves the discrete fields for range of motion, pain scale, functional indices, and failed therapies. It populates each field only after the clinician confirms the value in a side panel. Because it never bypasses Tebra's native save and signature actions, the EHR's own audit trail records the physician as author.

Quantitative and categorical items go to discrete fields, while clinical reasoning stays in the narrative. The assessment then references the structured values directly. For example: lumbar flexion limited to 45 degrees with NRS 8/10 despite six weeks of physical therapy and trials of an NSAID and duloxetine. A reviewer reading the note sees narrative and data that agree.

Audio is held only in volatile memory until transcription completes, and the buffer is then cleared. Nothing is written to disk, cloud storage, or training pipelines. The signed chart becomes the only durable record. That is the record SB 1120 directs plans to rely on. This design also aligns with the statute's requirement that patient data not be used beyond its intended and stated purpose under HIPAA and the Confidentiality of Medical Information Act.

Reimbursement Integrity and Appeal Readiness

The same structured fields support coding defensibility. HCPCS G2211 recognizes the longitudinal complexity of ongoing care for a single serious or complex condition. Chronic radicular pain with comorbid depression and multiple failed medication classes fits that description when the record shows it. CMS does not pay G2211 when the base E/M is reported with modifier 25. The exception is when the separate service is a preventive service, vaccine administration, or annual wellness visit.

Same-day trigger point injections require a distinct E/M history, exam, and medical decision-making to justify modifier 25, with the procedure documented separately. Multi-site work, such as a left L4-L5 facet injection and a right sacroiliac joint injection, needs separate anatomic site fields and indications. That separation supports modifier 59, or the appropriate X-modifier, against NCCI edits. Separation in the chart makes separation in the claim credible.

Clinicians preparing evidence-based appeals should pair discrete chart data with current literature on interventional pain documentation, available through https://www.ncbi.nlm.nih.gov/pmc/. When a denial arrives, confirm that the written notice identifies the criteria applied and the responsible health care professional, as § 1367.01(h) requires. Verify that the decision timeframe was met. Then cite the specific dated metrics the plan's tool was required to consider.

Practices uncertain about current workflows can Schedule a 15-Minute Workflow Audit to review field mapping, attestation steps, and appeal packet structure against SB 1120 requirements.

Regulatory & Compliance FAQ

What does Health and Safety Code § 1367.01(k) require of plan AI tools reviewing pain management services?

Section 1367.01(k), added by SB 1120, requires that any algorithm or AI tool used in utilization review base its determinations on the enrollee's medical history, individual clinical circumstances, and other clinical information in the record. The tool cannot supplant provider decision-making, must be applied fairly and equitably, and must be open to inspection by the Department of Managed Health Care. In pain management, this makes documented ROM, pain scores, and failed therapies legally material evidence.

Can an algorithm deny an epidural steroid injection or radiofrequency ablation under SB 1120?

No. Under Health and Safety Code § 1367.01(k) and Insurance Code § 10123.135(j), an AI tool cannot deny, delay, or modify health care services based in whole or in part on medical necessity. Only a licensed physician or licensed health care professional competent in the relevant clinical issues may make that determination. Confirm that the denial notice identifies the responsible reviewer and the criteria applied. Then appeal using dated objective metrics from discrete chart fields.

Which objective metrics most effectively support a pain management utilization review appeal?

Reviewers respond to quantified, dated findings. These include joint-specific range of motion in degrees, Numeric Rating Scale scores, Oswestry or Neck Disability Index percentages, straight leg raise angles, motor strength grades, and PHQ-9 or opioid risk screening scores. Pair them with a chronological failed-therapy record listing start date, end date, dose or frequency, and reason for discontinuation. Discrete fields let an appeal letter cite each element precisely rather than paraphrasing narrative text.

How does 0-day RAM-only audio handling relate to SB 1120 data use obligations?

SB 1120 directs plan tools to rely on documented clinical information. It also requires that patient data not be used beyond its intended and stated purpose, consistent with HIPAA and the Confidentiality of Medical Information Act. Merry AI holds encounter audio only in volatile memory and clears the buffer after transcription. It never writes audio to disk, cloud storage, or training pipelines. The physician-signed Tebra note remains the sole durable record subject to review.

Where can clinics verify compliant templates?

Clinics can review specialty-specific pain management prompts and structured exam templates at https://templates.scribing.io. Each template maps range of motion, pain scale, functional index, and failed-therapy fields to discrete inputs suitable for Tebra DOM injection. Compare template output against your current notes and confirm that every UR-critical metric is dated and quantified. Also verify that clinician attestation occurs before any value is written into the chart.

Explore More Knowledge

Related Clinical Rules & Compliance Guides

View all clinical rules & compliance
Florida Statute 934.03 Compliance for Psychiatry
CMS & Regulatory Compliance

Florida Statute 934.03 Compliance for Psychiatry

Enforce all-party verbal consent before psychiatric audio capture under Fla. Stat. 934.03(2)(d). Book your audit at https://cal.com/merryai/demo.

Dr. Sarah Chen, MD•7 min read
Read guide
Illinois 720 ILCS 5/14-2: Internal Medicine Consent
CMS & Regulatory Compliance

Illinois 720 ILCS 5/14-2: Internal Medicine Consent

Document all-party consent before ambient capture in Illinois internal medicine visits. Book your audit at https://cal.com/merryai/demo.

Dr. Sarah Chen, MD•7 min read
Read guide
Texas HB 149 AI Disclosure for Family Medicine
CMS & Regulatory Compliance

Texas HB 149 AI Disclosure for Family Medicine

Timestamp HB 149 disclosures before capture and protect G2211 integrity. Book your audit at https://cal.com/merryai/demo.

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