CPT code 97158

Editorial approval scope: The team checked current source fidelity, scope boundaries, dates, arithmetic, reader usefulness, practical workflow, and general-information limitations.

97158 joins group treatment with protocol modification

The American Medical Association CPT overview identifies CPT as the code set for reporting medical services and procedures. The licensed code set and its guidelines provide exact wording and reporting instructions. The AMA's behavioral health coding guide places 97158 in the adaptive behavior services family.

Licensed 2026 CPT materials place 97158 in the adaptive behavior family as a 15-minute, face-to-face group service administered in the physician or other QHP role; the code label includes protocol modification. “QHP” is a coding role, not an automatic synonym for BCBA. A credential alone does not establish scope, payer recognition, enrollment, authorization, or billing eligibility.

The ABA Coding Coalition FAQ offers stakeholder guidance, not AMA or payer authority. It says 97158 is intended for QHP-led sessions with two to eight patients, 97154 and 97158 are not reported concurrently, and the applicable group code is reported for each patient. It also says an actual protocol change is not required in every QHP-led 97158 session. Apply licensed CPT instructions and the member’s current payer rule before release.

Each participant needs attributable evidence

Each claim and unit calculation belongs to one patient. Support the member’s attendance, eligible time, QHP-led service, clinical work, and response. When a payer requires protocol development or modification, record what was evaluated, retained, tested, or changed. Never multiply one member’s time by group size on a single claim.

Connect the group event to a member-level record containing:

• member, payer, product, authorization, date, setting, modality, and service
• participant arrival, departure, eligible intervals, and group census
• qualified professional's role, license, enrollment, roster, and contract evidence
• treatment-plan link and protocol version before the session
• data, clinical rationale, QHP-led group work, protocol evaluation or modification when required, participant response, decision, and follow-up
• protocol version after the session and implementation instructions when changed
• time calculation, exclusions, modifiers, concurrent services, signatures, and corrections
• claim, acknowledgment, adjudication, remittance, payment, and appeal state

Administrative tools may compare times and sources. A qualified clinician authors the clinical reasoning, and a coding or billing reviewer applies the current code and payer rules. Preserve each person's contribution and every protocol version.

Payer group rules can differ

The current Texas Medicaid Children’s Services Handbook is a narrower payer example. It defines groups as two to eight and uses 97158 for LBA or delegated-LaBA direct group time to develop a new or modified protocol. Claims require HO or HN. Texas derives timed units from each client’s billable minutes for the calendar day. It permits synchronous audiovisual 97158 with modifier 95, but LaBAs and RBTs may not deliver remotely. It generally excludes concurrent ABA providers during one child’s session except its separate-family-service circumstance.

The June 24, 2026 TRICARE Autism Care Demonstration manual, Change 57 says 97158 may be delivered by an authorized ABA supervisor or delegated assistant behavior analyst. Groups may not exceed eight; telehealth is prohibited; units are 15 minutes; contractors authorize at least four and no more than six units per day; and reimbursement is calculated for each participant. The service is limited to treatment-plan goals for generalizing mastered skills. The cited ACD manual does not list 97154 among its Category I codes or concurrency table, so it cannot support a TRICARE-specific 97154/97158 pair rule.

These named program rules do not create national group-size, delegation, modifier, unit, telehealth, or coverage rules. Store payer, product, version, effective date, and source for every release decision.

Routine group treatment requires separate analysis

Routine group time depends on the governing source. The Coalition says a QHP-led 97158 session need not produce an actual protocol change; Texas Medicaid uses 97158 to develop a new or modified group protocol. Do not convert time from clinician presence or title alone. Identify the service, role, participant time, and payer rule; hold conflicts for qualified review.

A fictional group separates phases of work

Assume a fictional payer requires member-level evidence that the QHP evaluated or modified the group protocol. Three patients each attend six 15-minute intervals, producing 3 × 6 = 18 attended participant-intervals. The QHP supports the first four intervals for all three members. The final two intervals per member use the established protocol outside the assumed payer rule.

The internal evidence-support rate is (3 × 4) ÷ (3 × 6) = 12 ÷ 18, or 66.7%. This ratio is not a universal CPT test. Each member record supports four units, or 60 minutes; the aggregate 12 participant-units must never be placed on one member’s claim. The final six participant-intervals require separate service and payer analysis.

If all supported time matches the three authorizations, authorization alignment is 12 authorized supported participant-intervals divided by 12 supported participant-intervals due for release, or 100%. These ratios cannot establish clinical benefit, payer acceptance, adjudication, or payment.

Coding edits do not decide clinical need

The CMS Medicaid NCCI FAQ says NCCI edits address correct coding, not medical necessity; the program does not perform medical review or prior authorization. Some private insurers voluntarily adopt Medicare NCCI methodologies, but CMS does not control their implementation. Medically unlikely edit (MUE) values are not utilization guidelines, and claims at or below an MUE can still undergo medical review.

Record the payer, product, code pair or unit rule, edit version, date, provider configuration, modifier logic, result, and appeal path. A passed edit cannot supply missing service evidence, authorization, participant-level time, or a covered-provider relationship.

Prepare for the January 2027 transition

The AMA CPT coding resources identify the 2026 code set as effective January 1, 2026. The AMA’s September 2025 panel-action summary records panel acceptance, for January 2027, of six added adaptive behavior codes, revised guidelines and codes 97151 through 97158, and deletion of Category III codes 0362T and 0373T.

The panel summary is transition notice, not final licensed code text. Its placeholder codes are not for claims and should not be configured. Before 2027 services, verify the licensed 2027 code set, payer implementation, authorizations, contracts, templates, claims configuration, training, and effective dates. Preserve 2026 rules for older service dates.

Measure member-level support and outcomes

Use explicit numerators and denominators:

• Evidence-support rate: supported participant-intervals divided by attended participant-intervals reviewed
• Authorization-alignment rate: authorized supported participant-intervals divided by supported participant-intervals due for release
• Rule-review hold rate: participant-intervals held for provider, service, or payer-rule review divided by participant-intervals reviewed
• Release-gate pass rate: member claim records passing every gate divided by member claim records due
• Claim-acceptance rate: mature submitted claims accepted for adjudication divided by all mature submitted claims

Report pre-adjudication rejects, adjudicated denials, remittance, and payment separately.

Define the group, member, interval, protocol change, census, exclusion, payer, product, code-set version, claim maturity, and final state. Pair billing measures with client choice, access, clinical fit, safety, protocol integrity, treatment fidelity, and outcomes.