Data Monitoring Committees: Alira Health’s Methodology and Best Practices

Alira Health’s Approach to Setting up an Independent Statistical Center

Drafting and Reviewing the DMC Charter

The ISC could draft or review the DMC charter. The DMC charter will include the planned frequency of meetings, the minimum quorum for a closed session to be held, the data expected to be discussed, and the preferred platform to be used to share confidential and unblinded materials. It will also detail who will attend the open and closed sessions of the DMC meetings. A communication flow is a required component of the DMC charter, specifying to whom the DMC reports are sent and who receives the DMC recommendations. It is important that the charter covers how the blinding of the trial will be maintained for the sponsors, investigators, and participants, as well as the procedures for maintaining blinding and confidentiality when preparing reports for the DMC open sessions.

Establishing a DMC-Specific Statistical Analysis Plan (SAP)

There also will be a DMC-specific SAP and associated Table, Listing, Figure (TLF) shells which need to be reviewed and agreed to by the sponsor and the DMC to ensure that only the necessary and required information is provided. Once the DMC has begun reviewing unblinded data, the DMC will work with the ISC to provide ad-hoc information not included in the DMC SAP; in most cases, to protect the blind, the sponsor will not be informed of the nature of these requests until after the study has unblinded.

Defining Data Transfer Agreements (DTA) and Data Flow

Data Transfer Agreements will need to be set up to ensure the timely delivery of the data, especially if external vendor data is required. Ideally, the ISC should have access to all relevant study data, in order to respond to DMC questions in a timely manner and without informing the sponsor of specific data requests and transfers. The DTA should include timelines as to when the data is required to allow adequate time for the DMC report to be programmed and delivered. Considerations will also be given to cleanliness of the data and any coding requirements.

Ensuring Data Quality and Timely Access

Data that is recommended to be reviewed during the DMC meetings includes information on recruitment, data quality, and completeness. Any protocol violations will be summarized. Demographics, disposition, exposure, safety, and relevant efficacy on the trial should also be presented. Other items discussed could be any assumptions on the study design as well as any newly published research that may impact the trial.

Managing DMC Meetings and Communication

The ISC will organize and schedule the DMC kick-off meeting. For regular or ad-hoc DMC meetings, the ISC will receive the relevant data, produce, and distribute the DMC reports for both open and closed sessions. They will also attend the DMC meetings and ensure the signed DMC recommendation is sent to the sponsor in a timely manner. The independent statistician will present the data at the DMC meeting in a manner that is understandable to all the attendees and be able to answer questions. The ISC is also responsible for drafting and finalizing the open and closed minutes.

Only when the study has officially finished, the database locked, and the relevant unblinding has occurred will the ISC transfer all the unblinded material back to the sponsor upon a written request. The unblinded material generally includes at least the closed reports, unblinded datasets, and closed minutes.

Preparing and Delivering DMC Reports

The ISC generates a DMC report and there are various ways of doing this. They can either receive the raw or SDTM data and independently program the analysis datasets and the TLFs for the DMC report after merging on unblinded treatment data, or the ISC receives blinded analysis data sets and independently programs the TLFs for the DMC report after merging on unblinded treatment data. Some sponsors provide the blinded data sets and TLF programs for which the ISC will check that the programs run correctly on their system prior to receiving the unblinding information. There are advantages and limitations of each of these models, each of which should be discussed with the sponsor in order to develop the most appropriate programming approach prior to the first DMC meeting.

A DMC report is not to be seen as a subset of a clinical study report (CSR). The rules for deriving clinical endpoints could be different for a DMC report versus a CSR, as well as the layout of TLFs should permit the DMC to quickly review the relevant data before making a recommendation, which may require different or additional TLFs from the final analysis TLFs. Additionally, interim data are not fully validated, which may require some derivations or conventions be different between the DMC reports and the clinical study report. A specific DMC statistical analysis plan would specify any differences for programming or TLFs.