Testing new treatments on patients with life threatening diseases requires immediate understanding of whether the treatment helps or harms. Data Safety Monitoring Boards (DSMB) provide this critical oversight. Proper auditing of these boards ensures trial integrity and patient safety.

The Stakes

• Early Signal Detection: If early data shows a promising new drug outperforms standard treatment, continuing inferior treatment is unethical. If it causes harm, the trial must stop immediately.
• Herceptin Example: This cancer treatment received approval based on an interim analysis enabled by continuous DSMB benefit risk monitoring.
• Adverse Events: DSMB members review unblinded data to spot patterns and rare complications across treatment groups that demand immediate action.

The Challenge

A DSMB requires clear rules documented in a DSMB Charter. This defines roles, decision making processes, and strict procedures to prevent the trial team, investigators, and sponsor from seeing unblinded data. Evaluating benefit risk data requires systems that withstand GCP inspection scrutiny. This necessitates proper quality assurance oversight.

What Gets Audited

Auditors review several key areas:

• The Charter: Must be thorough, signed before the trial starts, and align meeting frequency with safety data flow.
• Governance: Defines the chair, minuting, and sponsor representation. Sponsors must lack voting rights and exit during deliberations.
• Data Protection: The Charter must detail compliance with GDPR or equivalent regulations when reviewing patient records, images, or ECG tracings.
• Competency: Verification of CVs, training records, and protocol understanding for all stakeholders. Changes in composition require proper transition documentation.
• Independence: Verified through Financial Disclosure Forms and conflict of interest checks.
• Meeting Records: Confirm adherence to the Charter and verify sponsor compliance with DSMB decisions.
• Blinding Safeguards: Evidence that blinded stakeholders remained blinded during safety data unblinding, supported by comprehensive process documentation.
• Computerized Systems: Validation documentation for systems like eDSMB, including the full system lifecycle and post change revalidation.

When to Audit

• Routine: Audit after 30 to 40 percent of patients are randomized and two or three DSMB meetings have occurred. This provides sufficient data while leaving time for corrective actions.
• Milestone Based: Adjust timing if the DSMB only meets at major milestones.
• Reaudits: Conducted to verify corrective actions or address operational issues.
• For Cause: Triggered anytime by suspected scientific or medical misconduct.

Planning and Preparation

• Timeline: The sponsor announces the audit 4 to 12 weeks in advance to coordinate logistics. The auditor sends an Audit Plan 4 to 6 weeks prior to execution.
• Scope Assessment: The auditor evaluates data volume, process complexity, and involved systems to determine duration, site visits, and required subject matter experts.
• Pre Audit Review: The auditor examines governance documents (Charter, SOPs, templates) and requests a clinical database line listing showing data reviewed by the DSMB. Minutes, images, and ECG tracings are reviewed on site for data protection purposes.

Who Can Audit

Any trained GCP or Pharmacovigilance auditor can perform a DSMB systems audit. Medical expertise is required for medical data review. Knowledge of CSV requirements is essential for assessing computerized systems.

The Bottom Line

DSMB audits ensure the system functions correctly. Decision makers need accurate data, clear authority, independence, and documented reasoning to protect trial participants. The goal is to establish processes that enable rapid responses to unexpected benefits or potential harms.