A data safety monitoring board or data monitoring committee works from a charter written before the first participant is enrolled. This template gives you 14 sections with guidance on what each must settle, fill-in wording to adapt, and a meeting agenda to go with it.
Free sandbox · No credit card · 21 CFR Part 11 aligned
Purpose and scope
Membership and independence
Three members, one statistician
Data flow and reports
Report shells to agree
Stopping guidance
Statistician to confirm
Recommendations and sponsor response
Confidentiality and firewall
Key points
Background
A charter is the committee's operating manual. It makes the committee's job explicit: protect participants and the integrity of the trial by reviewing the data at planned intervals, and give the sponsor a recommendation. Without one, committees improvise on membership, what they may see, how they vote and how the sponsor responds, and those are exactly the places where bias and disputes appear.
FDA's guidance for sponsors on data monitoring committees (final in 2006) is the established reference in the US, and FDA published a draft update, "Use of Data Monitoring Committees in Clinical Trials", in February 2024. Check whether it has since been finalised. EMA's guideline on data monitoring committees addresses independent committees in the EU and states, in summary, that the committee may review unblinded data and advises the sponsor on continuing, modifying or ending the trial. Across these sources, a complete charter typically covers roles and responsibilities, composition, handling of conflicts of interest, meeting timing and procedures, the plan for analyses, any prespecified stopping criteria, how recommendations are made and communicated, and confidentiality. The sections below follow that spine and add practical items teams need.
The committee is separate from other bodies. It is not the ethics committee or IRB, not the sponsor's safety review team and not an adjudication committee. See the SAE reporting timelines guide for how individual safety events reach the sponsor and regulators, and the SAE report form template for the reporting form.
Typical reasons: the trial has mortality or serious morbidity endpoints, the intervention carries notable risk, the population is vulnerable, the trial is large or long, or there is a planned interim analysis with a stopping rule. Short, low-risk, single-site trials often rely on the investigator and sponsor safety review instead. Record the reasoning in the protocol.
"DSMB" is common in US academic and NIH-funded work, "DMC" or "IDMC" in industry and Europe. A charter title should match the protocol, with the other term noted once so readers searching either find it.
Template
Adapt every row. Text in square brackets is a placeholder. Fill-in wording is deliberately plain so you can replace it with your protocol's own terms.
| Section | What it must settle | Fill-in wording (adapt) |
|---|---|---|
| 1. Purpose and scope | Why the committee exists, which trial(s) it covers, what it advises on. | This charter sets out the responsibilities, membership and procedures of the [DSMB/DMC] for protocol [ID], "[title]". The committee advises the sponsor on participant safety and trial integrity. |
| 2. Roles and responsibilities | What the committee reviews and decides; what it does not (no day-to-day management). | The committee will review accumulating safety data [and efficacy data at planned interim analyses] and recommend whether the trial should continue unchanged, be modified, or be stopped. The committee advises; the sponsor decides. |
| 3. Membership and independence | Number, expertise (clinical, statistical, ethics), chair, who is excluded. | The committee has [3-5] voting members: [clinical specialty], [clinical specialty], and one biostatistician. Members have no financial or other interest in the sponsor or the product beyond reasonable fees. Sponsor employees do not vote. |
| 4. Conflicts of interest | Disclosure form, update rules, how conflicts are judged. | Members disclose relevant interests before appointment and update them [annually / at each meeting]. The chair and sponsor review disclosures and decide whether a member should step back. |
| 5. Meetings | Schedule, triggers for ad hoc meetings, quorum, format, attendance, minutes. | The committee meets [before enrolment starts], then [every N months or after N participants], and ad hoc if requested by the chair. A quorum is [number] voting members. Minutes are recorded for open sessions; closed-session records are held separately. |
| 6. Data flow and reports | Which data, who prepares reports, timing, format, blinding. | An independent statistician prepares the closed report from the locked-down dataset cut dated [date]. The open report is prepared by [role]. Reports are sent [N] days before the meeting. |
| 7. Safety review | Which safety measures are reviewed, how events are summarised, how SAEs and suspected unexpected reactions are fed in. | The committee will review [adverse events by severity and relatedness, serious adverse events, discontinuations, laboratory and vital sign trends] by treatment group in the closed report. |
| 8. Stopping guidance | Prespecified rules or criteria (if any), who defines them, how they are applied. | Guidance for [safety / futility / efficacy] is described in the statistical analysis plan, section [x]. Guidance is not a rigid rule; the committee weighs the totality of the evidence. |
| 9. Voting and decisions | How decisions are made, consensus or vote, how dissent is recorded. | The committee aims for consensus. If a vote is needed, a majority of voting members decides. Dissenting views are recorded in the closed minutes. |
| 10. Recommendations and sponsor response | Who receives the recommendation, in what form, and how fast the sponsor responds. | After each meeting, the chair sends a written recommendation to [sponsor contact] within [N] working days. It states only the recommendation and, where needed, the reasoning that does not reveal unblinded results. |
| 11. Communication and escalation | Who may talk to the committee, urgent safety signals, contact routes. | Urgent concerns are raised to the chair by [route]. Members do not discuss unblinded results outside closed sessions. |
| 12. Confidentiality and firewall | Who may see unblinded data; storage; what happens to papers; copies. | Unblinded outputs are accessible only to the committee and the independent statistician. They are held in [secure location]. Sponsor staff are not given access to treatment-level results. |
| 13. Documents and records | Minutes, reports, disclosure forms, file location, retention. | Open-session minutes and recommendations are filed in the trial master file. Closed-session materials are held by [independent party] until [event] and then handled per the sponsor's retention plan. |
| 14. Charter review and changes | Amendment process, version control, sign-off. | The charter may be amended by agreement of the chair and sponsor. Each version is dated and signed before the first participant is enrolled and filed in the trial master file. |
Adapted from the topics in FDA and EMA guidance plus common practice. Items such as numbering, notice periods and quorum are examples to settle with your biostatistician and sponsor.
Running it
Most committees split each meeting. An open session includes sponsor and investigator representatives and covers blinded matters: recruitment, protocol adherence, data quality and any operational issue. A closed session includes only the committee, and the independent statistician where needed, and covers treatment-level data. Some groups add an executive session for committee members alone. The charter should say which sessions are held, who attends each, and what is recorded. EMA's Q&A discusses how disagreement between the committee and the sponsor should be handled; the charter should say what happens before such a case arises.
Enrolment by site against plan, eligibility and protocol deviation counts, data completeness and query status, follow-up and discontinuations (pooled), and overall event counts that do not reveal arm. A data capture system with by-site filtering and exports makes these easy to produce. See risk-based monitoring and SDV software.
Safety and, where relevant, efficacy outputs by treatment group, prepared by someone outside the sponsor's team. If your platform restricts blinded roles from treatment values, as described in blinded and unblinded role management, the independent statistician should receive the dataset through a separate, controlled route and not through a blinded user's export.
minutes
Recruitment, conduct, data quality
Safety by arm, planned analyses
Recommendation drafted
Chair briefs the sponsor
Build the study, visits and safety forms in the free sandbox, so the open-report outputs exist from the first participant. Free sandbox, every feature, no credit card.
How to use it
Write the reasoning into the protocol: risk, population, size, duration and endpoints. If you decide not to have one, say who monitors safety instead.
Use the table above. The statistician must agree sections 6 to 9, since the data cuts, the interim looks and any stopping guidance come from the statistical analysis plan.
Invite clinicians with the relevant experience and a statistician. Collect conflict disclosures before the first meeting and file them.
Mock tables for the open and closed reports let the committee ask for what it needs before data exist. Build them from the analysis plan.
Review the protocol, sign the charter, confirm membership, and agree dates. Record the decisions.
Follow the agenda, document votes and send a short recommendation letter to the sponsor. Keep closed material away from the main file.
File open materials in the master file, handle closed materials as the charter states, and revisit the charter if the protocol changes. See how to set up a trial master file.
In Capture
Capture is an EDC and clinical trial platform, not a DSMB service: it does not run your committee, write your charter or produce the independent closed report. It gives the study team the data the committee depends on: adverse event and safety forms, query status, by-site enrolment views, role-based access with blinded roles that never receive treatment-arm values, and CSV or Excel exports with a data dictionary, and SDTM datasets as SAS XPT with Define-XML for the independent statistician.
Enrolled
86 / 120
Open queries
14
SAEs reported
5
Discontinued
7
Behind plan
Before the first meeting
In the protocol, with the reasoning.
By the chair and the sponsor, dated, filed.
Disclosure forms collected; sponsor staff are not voting members.
An independent statistician prepares closed outputs.
Open and closed shells agreed against the statistical analysis plan.
Dates, secure transfer, who holds unblinded data.
Or a statement that none is prespecified.
Who receives it, in what form, and the sponsor's response time.
Where closed material is held and for how long.
Common mistakes
The recurring problems are vague data-flow sections, a statistician who joins after the charter is signed, a sponsor employee who attends closed sessions "just to listen", and recommendations that include unblinded numbers. Another is copying a charter from a different trial without changing the stopping guidance, which then conflicts with the statistical analysis plan. Read the plan first: see the statistical analysis plan and how to write a statistical analysis plan.
Keep the charter consistent with the protocol, the monitoring plan and the risk management plan. A charter that promises quarterly reviews the protocol does not budget for is a deviation waiting to happen.
In practice they are the same kind of independent committee. "DSMB" is common in US academic and NIH-funded work, and "DMC" or "IDMC" is common in industry and Europe. Use the term in your protocol.
Guidance summaries list roles and responsibilities, composition, conflict-of-interest handling, meeting timing and procedures, planned analyses, prespecified stopping criteria where applicable, the recommendation process and confidentiality. This template adds data flow, voting, communication and amendment sections.
No. The need depends on risk, population, size, duration and endpoints. Trials with serious outcomes, high-risk interventions or planned interim analyses are the usual candidates. Record the decision and the alternative safety oversight in the protocol.
Typically clinicians experienced in the condition, a biostatistician, and sometimes an ethicist or patient representative. Regulators expect independence from the sponsor. Check the current FDA and EMA guidance for your case.
Usually an independent statistician who is not part of the sponsor's study team. The charter should name the role and the secure route for the dataset.
The committee advises and the sponsor decides, though sponsors usually have a process for responding promptly and recording the reasons if they do not follow a recommendation. State this in the charter.
No. Capture is an EDC and trial platform. It provides the safety forms, queries, role-based access with blinded roles and dataset exports a committee depends on. It does not author charters, run meetings or produce the independent closed report.
No. Use it as a structure, then check it against current FDA and EMA guidance, your protocol and your biostatistician. It is not legal advice.
Keep exploring
SAE reporting timelines guide
How individual events reach regulators.
SAE report form template
The event report that feeds safety review.
Clinical trial protocol template
Where the decision to have a committee is recorded.
Blinded and unblinded role management
Keeping treatment data away from blinded users.
Risk management plan template
The wider oversight plan.
Statistical analysis plan
Source of interim analyses and stopping guidance.
Free sandbox with every feature. No credit card, and you pay only when you go live with real participants.