Purpose

This policy establishes standards for the use of artificial intelligence tools by speech-language pathologists and support staff within the hospital system. It ensures that AI-assisted workflows comply with the Health Insurance Portability and Accountability Act (HIPAA), institutional quality assurance requirements, and professional standards of care while enabling clinicians to use technology responsibly to improve documentation efficiency and clinical productivity.

Scope

This policy applies to all speech-language pathologists, speech-language pathology assistants, clinical fellows, and administrative support staff within the Department of Speech-Language Pathology who use AI tools in connection with patient care, clinical documentation, instrumental assessment reporting, discharge planning, or any other professional function performed within or on behalf of the hospital system.

Definitions

• AI Tool: Any software application using artificial intelligence, machine learning, or large language models to generate, summarize, analyze, or process text, clinical data, or other content.
• Protected Health Information (PHI): Any individually identifiable health information transmitted or maintained in any form, as defined under HIPAA (45 CFR 160.103).
• Business Associate Agreement (BAA): A written contract between the hospital and a third-party vendor that ensures the vendor will appropriately safeguard PHI in accordance with HIPAA requirements.
• AI-Assisted Documentation: Any clinical note, report, or communication for which AI tools were used in drafting, summarizing, formatting, or generating any portion of the content.
• Clinical Recommendation: A professional judgment regarding diagnosis, prognosis, treatment approach, diet modification, or discharge disposition.

Policy Statements

1. No AI tool shall process, store, or transmit PHI unless the hospital has executed a Business Associate Agreement with the tool’s vendor and the tool has been approved by the Information Security Office.
2. AI tools shall not generate clinical recommendations, diagnostic impressions, or diet-level determinations independently. All AI-generated clinical content must be reviewed, verified, and approved by a licensed SLP before entry into the medical record.
3. AI use shall be documented in departmental quality assurance records and shall be subject to periodic audit.
4. Clinicians retain full professional and legal responsibility for all documentation entered into the electronic health record, regardless of whether AI tools were used in its preparation.
5. The department shall maintain a current list of AI tools approved for use within the hospital system. Use of unapproved tools with any patient-related information is prohibited.

Approved Uses

• Drafting clinical documentation templates for evaluations, daily treatment notes, and discharge summaries using de-identified clinical scenarios.
• Generating standardized language for instrumental assessment reports (e.g., MBSS, FEES) that the clinician then customizes with patient-specific findings.
• Summarizing clinical research articles to support evidence-based practice decisions.
• Assisting with drafting patient education materials on swallowing safety, communication strategies, or cognitive-linguistic exercises.
• Using approved, BAA-covered AI features embedded within the hospital’s electronic health record system.

Prohibited Uses

• Entering any PHI, including patient names, medical record numbers, dates of service, diagnoses, or any identifying details, into any AI tool that lacks an executed BAA with the hospital.
• Using AI to generate clinical recommendations, modified diet levels, or aspiration risk assessments without independent clinician verification against objective clinical data.
• Submitting instrumental assessment recordings (video, audio, or imaging) to any AI platform not approved by the Information Security Office.
• Using AI-generated text verbatim in the medical record without clinician review and individualization.
• Using personal AI accounts or consumer-grade AI tools for any patient-related work.

Data Protection Requirements

1. All AI tools used with patient information must comply with HIPAA Privacy and Security Rules (45 CFR Parts 160 and 164) and must be covered by an executed BAA.
2. PHI shall not be stored on, transmitted through, or processed by any AI system outside the hospital’s approved technology infrastructure.
3. Any AI tool integrated into clinical workflows must undergo a security risk assessment conducted by the Information Security Office prior to deployment.
4. In the event of a suspected data breach involving AI tools, staff must immediately notify their supervisor and the Privacy Officer in accordance with the hospital’s Breach Notification Policy.

Disclosure Requirements

1. When AI tools are used in the preparation of clinical documentation, the clinician shall include a notation in the medical record (e.g., “Documentation prepared with AI-assisted drafting tools; reviewed and finalized by [Clinician Name], [Credentials].”).
2. Patients or their authorized representatives who inquire about AI use in their care shall receive a clear and accurate explanation of how AI was used and the clinician’s role in reviewing all content.
3. AI use disclosure practices shall align with the hospital’s existing informed consent and patient rights frameworks.

Documentation Standards

1. All AI-assisted documentation entered into the medical record must carry the reviewing clinician’s signature, credentials, and date, along with notation of AI assistance.
2. The department shall maintain a quarterly log of AI tools in use, their approved applications, and any incidents or concerns reported by staff.
3. AI-assisted documentation is subject to the same quality review standards as all other clinical documentation, including peer review and audit processes.
4. Original AI-generated drafts need not be retained, but the final clinician-reviewed version in the medical record constitutes the official record.

Compliance & Accountability

1. The SLP Department Director, in coordination with the Information Security Office and the Compliance Department, is responsible for enforcing this policy.
2. Violations involving unauthorized disclosure of PHI through AI tools will be managed in accordance with the hospital’s HIPAA Sanctions Policy and may result in disciplinary action up to and including termination and regulatory reporting.
3. All department staff shall complete HIPAA-compliant AI use training upon hire and annually thereafter.
4. AI use practices shall be included in the department’s annual quality assurance plan and subject to internal audit.

Review Schedule

This policy shall be reviewed at minimum annually by the SLP Department Director, a representative from the Information Security Office, and the Compliance Department. Updates shall be issued as needed in response to changes in regulation, institutional policy, or available technology.

Acknowledgment

I, ______________________________ (printed name), acknowledge that I have read, understand, and agree to comply with this policy governing AI use in speech-language pathology services within the hospital system.

Signature: ______________________________ Date: ______________

Title/Credentials: ______________________________ Department: ______________________________

Supervisor Signature: ______________________________ Date: ______________