Search
Book a Demo
Book a Demo

Decision Support Interventions

170.315(b)(11) DECISION SUPPORT INTERVENTIONS DISCLOSURE

Purpose

This document satisfies the transparency and disclosure requirements of 45 CFR 170.315(b)(11) (Decision Support Interventions). It describes the Decision Support Interventions (DSIs) implemented within EHRYourWay, including their purpose, source attributes, configuration capabilities, user interaction, feedback mechanisms, and Intervention Risk Management (IRM) practices.

Decision Support Interventions Overview

EHRYourWay implements two types of Decision Support Interventions:

Evidence-Based DSIs – Rule-based interventions derived from clinical guidelines and established medical knowledge:

  • Drug-drug interaction alerts
  • Allergy warnings
  • Preventive care reminders based on clinical guidelines

Predictive DSIs – AI-enabled interventions that produce recommendations based on patient data:

  • Diagnosis suggestions via Documentation AI Assistant (optional, user-configured)

Evidence-Based Decision Support Interventions

Drug-Drug Interaction Alerts

Alerts users when prescribed medications have known interactions. Triggered during medication entry and e-prescribing workflows.

Source Attributes:

  • Developer: EHRYourWay / Third-party drug database provider
  • Evidence Source: Medi-Span
  • Last Updated: Updated monthly with regular drug database releases
  • Known Limitations: Interactions are based on published data; not all potential interactions may be included

Drug-Allergy Interaction Warnings

Alerts users when a prescribed medication conflicts with documented patient allergies. Triggered during medication entry and e-prescribing workflows.

Source Attributes:

  • Developer: EHRYourWay / Third-party drug database provider
  • Evidence Source: Medi-Span
  • Last Updated: Updated monthly with regular drug database releases
  • Known Limitations: Dependent on accurate allergy documentation in patient chart

Preventive Care Reminders

Prompts users when patients are due for preventive services based on clinical guidelines. Triggered when accessing patient charts or during encounters.

Source Attributes:

  • Developer: EHRYourWay
  • Evidence Source: USPSTF recommendations, CMS quality measures, clinical practice guidelines
  • Last Updated: Annually
  • Known Limitations: Reminders are based on documented patient data; missing or incomplete data may affect accuracy

Predictive Decision Support Intervention

Diagnosis Suggestions via Documentation AI Assistant

When enabled by a practice and configured with a Suggested Diagnosis field, the Documentation AI Assistant may suggest potential diagnoses based on clinical content captured during the encounter. This feature is optional and must be explicitly configured by the practice.

Source Attributes:

  • Intervention Name: Diagnosis Suggestions
  • Developer: EHRYourWay (integration); AI model provided by OpenAI
  • Model Type: Large language model (AI/ML-based)
  • Evidence Source: AI model trained on broad medical and general knowledge; EHRYourWay does not control or modify training data
  • Intended Use: To assist clinicians by suggesting potential diagnoses for consideration; not intended to replace clinical judgment
  • Intended Users: Licensed clinicians reviewing AI-assisted documentation
  • Decision-Making Role: Informational only—suggestions must be reviewed, validated, and accepted or rejected by the clinician before use
  • Known Limitations: AI-generated suggestions may be inaccurate, incomplete, or inappropriate for a specific patient context; model may reflect biases present in underlying training data; suggestions are not based on the full patient chart history unless included in the encounter content
  • Known Risks: Over-reliance on AI suggestions without clinical validation; potential for automation bias
  • Cautioned Uses: Should not be used as the sole basis for diagnosis; not validated for pediatric, rare disease, or complex multi-system conditions; clinician review required before finalizing any diagnosis
  • Date of Last Update: Ongoing updates via AI vendor

User Interaction

EHRYourWay enables users to:

  • View decision support interventions within clinical workflows
  • Interact with alerts, recommendations, and notifications
  • Accept, defer, or override interventions
  • Access source attribute information for each intervention

Configuration Capabilities

Authorized users may:

  • Enable or disable specific DSIs
  • Configure user roles for intervention delivery
  • Enable or disable the Suggested Diagnosis field in Documentation AI Assistant
  • Submit feedback on DSI performance

Feedback Mechanisms

EHRYourWay supports user feedback on Evidence-Based DSIs. Users can:

  • Provide electronic feedback on interventions
  • Feedback data is available for export in computable format, including intervention type, action taken, user feedback (if provided), user, date, and location

Intervention Risk Management (IRM) for Predictive DSIs

EHRYourWay maintains IRM practices for Predictive DSIs, informed by the National Institute of Standards and Technology (NIST) AI Risk Management Framework.

Risk Identification

Risks evaluated include:

  • Inaccurate or inappropriate recommendations
  • Model bias from underlying training data
  • Automation bias (over-reliance on AI suggestions)
  • Security and privacy vulnerabilities

Risk Mitigation

Mitigation controls include:

  • Clinician review and validation required before any AI-suggested diagnosis is finalized
  • Clear labeling of AI-generated content
  • Suggested Diagnosis feature is optional and must be explicitly enabled by the practice
  • Business Associate Agreements (BAAs) with AI vendors handling protected health information
  • Selection of established AI vendors with published responsible AI practices

Risk Monitoring

Ongoing monitoring includes:

  • User feedback collection and review
  • Periodic evaluation of AI vendor updates and model changes

Governance

EHRYourWay maintains oversight of Decision Support Interventions. Integration of new DSIs or modifications to existing interventions requires review of clinical appropriateness, security, privacy, and operational considerations.

Transparency Statement

All Decision Support Interventions are implemented with transparency regarding purpose, evidence base, data inputs, limitations, risks, and the requirement for human oversight. Predictive DSIs are clearly identified as AI-generated and require clinician validation before use in patient care.