IT systems · Practical guide

Validating Medidata Rave: controlling study amendments and preserving data meaning

A Medidata Rave form changes during a trial. The new display looks correct, but a previously recorded value now carries a different unit. Can reviewers identify what was collected, when, under which instruction and by whom? Trace study configuration, changes, data review and close-out. Small sponsor and CRO teams, QA and IT can use this to coordinate their own evidence with the provider and trial sites.

Table of contents

At a glance

How do you validate Medidata Rave for a clinical trial?

Validate the intended Rave use, including study configuration, roles, imports, amendments and required exports. Assess the actual coverage of supplier evidence and add tests for your protocol rules and workflows. Preserve data meaning, trace amendments and verify investigator site control. A published release or technically successful form change does not replace study-specific assessment.

  • Rave EDC, Rave Lite, and complementary Medidata products require a specifically named scope.
  • Platform validation and evidence of the specific study design complement each other.
  • A change must not misinterpret values, units and mismatches undetected.
  • User approval, technical lock, and long-term data access serve different functions.

Delimit Rave in the clinical data process

Medidata describes Rave as an EDC system for collection, management, cleaning, and reporting of clinical trial data. The key factor is your specific implementation: are source data entered directly, data transferred from other documents, or external data imported? These approaches introduce different potential sources of error. A review that merely opens blank forms does not confirm whether the full study process is adequately supported.

The regulatory framework of this guide is the clinical trials of medicinal products for human use in the EU-GCP context. The EMA guideline on computerised systems of 2023 and ICH E6(R3), Principles and Annex 1. The latter have been effective in the EU since 23 July 2025. Annex 2 only becomes effective on 15 January 2027; the consolidated publication of 2026 does not change this time limit.

EDC use is not automatically a GMP application or a complete safety reporting process. Identify data types, trial decisions, connected systems and the authoritative source. Define the actual responsibilities of the sponsor, CRO, supplier and trial site.

Sources: [1] [8] [9]

Further reading:

Maintain product scope and documented version status

Rave is currently the visible product name; Rave EDC remains in use in technical and training documentation. The latest Factsheet lists Designer as a component for study setup. Additional features such as Safety Gateway, Clinical Data Studio, and Site Cloud: End of Study are also listed. Rave Lite is a distinct variant. Do not transfer features of these products to your environment without a contractual and configuration review.

The public Medidata training digest from August 2026 references Rave 2026.2.1. This version reference is validated, not the globally latest or the build currently in use by your organization. The publicly available Amendment Manager training agenda, however, refers to Rave EDC 2024.1.0. An older course overview may clarify certain terms, but it does not replace a current functional description.

Therefore, capture the environment used, the verified software version, study versions, activated modules, interfaces, and the scope of available vendor documentation. For unresolved details such as lockout behavior or signature sequences, rely on the documentation most relevant to your specific setup. A version number in the general project folder is insufficient for this purpose.

Sources: [2] [3] [4] [1]

The study guidelines up to the used dataset adhere to

The following work matrix serves as an independent planning aid. It assigns results and responsibility questions without prescribing a universal role distribution. A small team may consolidate multiple tasks; however, the required qualifications, authorizations, and appropriate controls must remain clearly identifiable.

Start with a few key data pathways. For each pathway, the relationship between protocol rule, form, processing, and subsequent use must be clearly defined. Existing standard tests may support this evidence provided their scope and results are genuinely relevant.

Own work matrix for a limited Rave study process
ResultSpecific evidenceClarifying Responsibility
Study Rules Protocol version P2 requires body height in cm; meaning of missing entries is defined Who validates the technical specification?
eCRF Setup Formular version F2, field type, unit, and applicable check rule identified Who prepares and who evaluates the setup?
Data transfer Source, participant linkage, mapping, and error handling documented Who resolves conflicting source data?
Amendment Affected sites and participants, legacy/new installation and activation conditions documented Who is authorized to activate the change at the respective site?
Review Targeted control of values, changes, and open questions defined Who evaluates which data at what time point?
Closure Dataset, required metadata, access and receipt demonstrated Who is responsible for the available closing documentation?

Sources: [8] [10]

Integrate study configuration and vendor documentation.

A suitable supplier qualification can cover a significant amount of work. Verify that the specific version, standard features, environment, and test scenario actually match the requirements. Document any remaining gaps relevant to your study. The EMA guidelines do not require indiscriminate repetition of all supplier assessments; what is essential is suitability and traceable coverage.

For the own configuration, relevant parameters include visit structure, field types, units, mandatory fields, edit checks, and derived values. Specify expected results prior to execution. A test using fully completed typical data is insufficient, particularly when missing information, inconsistent units, or delayed data transfers could potentially influence a critical decision.

Use controlled synthetic data and explain relevant differences between test and production. At the actual interface boundary, check what is accepted, rejected or routed for resolution. Define who may correct trial site data. Technical administrator access does not provide that authority.

Sources: [8] [10] [2]

Separate decisions for formulation change and site activation

Medidata describes QuickPublish for specific changes, such as label updates or validation rules. The Amendment Manager Agenda includes, among other items, migration planning as well as Dry Run and Live Run. These public records illustrate different tools available for change management. They demonstrate that there is no general rule permitting every change to be implemented without data impact or additional validation evidence.

First, classify the impact on the process: pure text correction, altered input limit, new derivation, or changed meaning of an existing field. Then determine which previous and future datasets are affected. Even a visually minor change may be significant, for example, if a unit changes in the label.

Plan the activation of each testing site with the required relevant study authorizations. Technical readiness and organizationally permissible usage are separate milestones. Document how, during the transition phase, the applicable form version remains clearly identifiable for a participant and how already entered data is handled. A failed transition must follow a controlled decision-making process rather than silent post-hoc remediation.

Sources: [5] [4] [10] [8]

Example: Test case – a unit is modified

The following example serves as a hypothetical scenario, not a completed Rave test. For a study, a licensed configuration change is intended to display body weight in meters instead of centimeters. A specifically defined data conversion will involve multiplying existing numeric values by 100. Whether and with which tool this conversion can be implemented in your Rave environment must be independently assessed.

Background involves three participants: S01 at 180 cm, S02 at 170 cm, and S03 with no numerical value. In the target, there remain three participants with two numerical values and one missing value. The following errors are intentionally introduced; they are not alleged product defects.

Example: Soll-matrix – Data significance before and after an amendment
Test case and resultExpected resultControl and Decision-making
S01: 180 cm, valid prior state 180 ÷ 100 = 1.80 m Verify value and new unit together; ensure old value is contextually traceable
S02: 170 cm, valid previous baseline 170 ÷ 100 = 1.70 m Participant assignment and independent billing align
S03: no numerical value available Further missing, not 0 m Out-of-specification status and, if applicable, the need for clarification shall be addressed.
Error A: label change only for S01 180 m is recognized as a false acceptance. 180 ÷ 1.80 = 100. Reject the change.
Error B: missing values are replaced with 0 0 m is not accepted as a substitute value Address meaning changes; identify affected datasets
Export upon takeover 3 participants, 2 numerical values, 1 outlier; units identifiable S01 and S02 together amount to 3.50 m as a pure control sum; no clinical parameter value.

Sources: [8] [10] [4]

Data review, audit trail and investigator sign-off

A data review assesses whether information is clear and complete for its intended purpose and adequately justified. An audit trail review examines relevant creation and modification events along with their context. The EMA guidelines outline a risk-based, study-specific approach. Not every unchanged entry requires the same level of review; an unexpected unit change, for example, may necessitate targeted traceability.

A technical lock prevents or restricts certain actions in accordance with system rules. It does not automatically validate the professional appropriateness of all data. Similarly, electronic confirmation is only as reliable as its scope, timing, and the actual prior review conducted. The reviewer's oversight must not be reduced to an undifferentiated end-of-study summary step.

Define which questions permit a decision and which must be resolved first. Additionally, assess blinding: Changes, roles, exports, or error reports could reveal information. The specific Rave configuration must implement these requirements; general product descriptions are insufficient to infer its behavior.

Sources: [8] [10] [3]

Further reading:

Review and verify transfer and closing documentation with context

Medidata presents Rave Web Services as an integration option. The detailed developer documentation is access-restricted. Without the appropriate reference, you should neither assume a specific reset behavior nor automatic retry, or complete transfer of all metadata. Request the required interface documentation for your data pathway.

Proposed test scenarios include incomplete delivery, duplicate request, and correction of a previously transferred value. For each scenario, define how source, transfer instruction, target inventory, and error message must align. Successful connection or acceptance of a file does not constitute valid participant assignment.

Site Cloud: End of Study is a dedicated offering for the distribution and confirmation of closing documentation. Specifically evaluate which data and metadata are received by the review authority, which functions remain necessary for understanding, and how future access is secured. A receipt confirmation certifies the transfer; completeness and suitability of the materials must be assessed separately.

Sources: [6] [7] [8]

Limit effort with clear prerequisites

The effort increases significantly due to inconsistent protocol guidelines, retrospective form changes, numerous data sources, and unclear responsibilities. Therefore, a small sponsor must early establish a reliable list of critical data pathways and point of contact. Pure test case counts are a poor measure of effort when expected outcomes remain disputed.

Prepare the protocol version, data management requirements, role matrix, interface overview and known change needs. Request appropriate supplier evidence and the versions it covers. Identify someone authorised to answer process questions and someone who follows technical findings to resolution. Include the actual workflows of the trial site.

In operation, new releases, relevant defect notifications, and study modifications are evaluated for their impact. A unchanged product name does not imply an unchanged process. Conversely, every release does not necessitate a complete reexecution of all previous tests. Document the justified selection, remaining limitations, and required actions.

Sources: [8] [10] [3]

Checklist prior to activation and after changes

Use this internal checklist as a basis for discussion. An open point requires a specific decision with defined responsibility and a scheduled date; a tick without accessible evidence will provide little support later. The list is supplemented with the actual risks identified in your study.

For this example change, establish correct assignment of all three participants, conversion of both numerical values and preservation of the missing entry. Also verify the traceable historical version, the export and authorised activation at the affected trial sites.

  • Are Rave variant, environment, modules, and study status clearly identified?
  • Is the original source defined for each essential data pathway?
  • Do vendors provide actual usage functions and versions?
  • Are own rules, units, out-of-range values, and interfaces verified against expected results?
  • Are affected participants and sites determined for each amendment?
  • Are necessary authorizations and technical activation separately traceable?
  • Are permissions, blinding and investigator oversight implemented appropriately?
  • Are defect handling, review, and, if necessary, reconfirmation addressed?
  • Are finalization data, relevant metadata, access, and retention agreed upon?
  • Are open findings assessed and the usage decision authorized documented?

Sources: [8] [10]

Project experience and collaboration

Reconcile study changes with values already collected

For a study change, we begin with the scientific meaning of the affected data. Your study team explains which unit or form rule will change and how to treat data already collected. With the technical teams, we plan checks for configuration, activation and later analysis. A limited change may need only targeted checks. If several study versions or data transfers are affected, we agree the mapping and decision rule first. We explicitly account for the controls that the investigator site must perform.

Frequent Questions on Rave Study Modifications

Has Medidata Rave already been validated if the supplier provides evidence?

Supplier documentation may cover standard functions and operational capabilities. Verify version, scope, test conditions, and remaining limitations. For your study, the specific configuration, protocol rules, data pathways, and relevant workflows must be adequately demonstrated. This does not imply complete retesting of the product nor the unverified adoption of a general certificate. Clearly document which evidence you are accepting and which gaps you will address independently.

Sources: [8] [10]

Which Rave version should we reference in the validation plan?

State the actually deployed and supplier-verified version including relevant modules and study configurations. The public training digest indicates Rave 2026.2.1, but does not establish universal currency across all environments. The 2024.1.0 Amendment Manager agenda serves as an older functional validation. Additionally, document how subsequent service changes are identified and aligned to the existing validation scope.

Sources: [3] [4] [8]

Can QuickPublish be used without additional review?

The technical pathway alone does not determine the testing requirement. Medidata outlines QuickPublish for certain types of changes; the impact on the process of such changes must still be assessed. A spelling correction can be reviewed with a focused approach. However, if the unit, a derived value, or a valid input range is altered, affected data and downstream processes must be evaluated. The scope should be defined based on the specific change.

Sources: [5] [8] [10]

Can missing values during a migration be defaulted to zero?

Only if zero has that defined meaning. A missing measurement does not automatically mean zero. In the height example, S03 has no numerical value. Test missing, not measured and, where relevant, not applicable according to your defined data states. A populated column must not give a false impression of completeness or alter the interpretation.

Sources: [8] [10]

Must every previous participant be migrated after an amendment?

Which prior participant data are affected is determined by the approved study amendment together with the relevant data collection time points. Explicitly specify which data are to be transferred and which are to remain unchanged. The Amendment Manager, according to the public agenda, supports migration planning; appropriate current documentation and assessments are required to define the exact scope and behavior of locked data.

Sources: [4] [8] [10]

Is a technically locked study automatically considered professionally concluded?

No. Technical locking, data cleansing, reviewer control, and organizational final decision address different questions. Review open queries, relevant changes, and required confirmations according to the study protocol. The public Rave training overview lists site-lock extensions but does not describe all customer-specific prerequisites. Therefore, document the specific meaning of the applied lock and the approach to handling a subsequent required correction.

Sources: [3] [8]

Must every audit trail event be manually verified?

The EMA guideline provides for a risk-based, study-specific review. Define relevant data, events, time points, and responsible parties. A unit conversion, unexpected changes following confirmation, or suspicious editing patterns may warrant special attention. An unfiltered overall summary is not sufficient for an effective review. Selection and evaluation must address the actual risks associated with the data process.

Sources: [8] [10]

Does a Rave closure package replace the planning of archiving?

The package can serve as a key component. However, it is essential to understand which datasets, metadata, and change records are included and how they remain interpretable after contract termination. Site Cloud: End of Study supports distribution and receipt confirmation; this does not imply automatic suitability of any delivered format. Specifically evaluate required dynamic functions, access rights, and the applicable retention strategy.

Sources: [7] [8]

Is Rave EDC also a Safety System?

Rave captures clinical trial data; this alone does not include complete safety incident handling or regulatory reporting. Medidata offers Safety Gateway as a complementary product. When safety-related information is shared, a clearly defined handover with defined responsibilities, data scope, error management, and reconciliation is required. Neither shared participant identifier nor common platform ensures a complete or timely reporting process.

Sources: [2] [1] [8]

What should a small sponsor or CRO team start with?

First, select a critical data pathway and describe its source, processing, users, and future usage. Then, clarify the scope of the product, supplier documentation, and study-specific rules. A well-structured contingency plan covering normal, error, and change states provides greater clarity than an unverified standard test list. From this, the full scope of additional evidence and required involvement can be realistically derived.

Sources: [8] [10]

Sources and review date

Source date: 6 October 2026. Rave 2026.2.1 is documented against a public training overview; Amendment Manager details originate from a 2024.1.0 agenda. Protected vendor packages, current customer configurations, and RWS references were not reviewed. Regulatory framework is EU-GCP for clinical human medicinal product studies. Tables and calculation examples are standalone work aids.

  1. Medidata: Rave – Electronic Data Capture
  2. Medidata Rave Fact Sheet
  3. Medidata Global Education: Rave EDC Training Digests
  4. Medidata: Rave EDC Amendment Manager – Training Agenda
  5. Medidata: Rave EDC for Frictionless Mid-Study Changes
  6. Medidata: Developer Central
  7. Medidata: Site Cloud – End of Study
  8. EMA: Guideline on computerised systems and electronic data in clinical trials
  9. EMA: ICH E6 Good Clinical Practice – Status
  10. ICH E6(R3): Principles and Annex1 – EU Step5

Sources checked: .

Daniel Herrmann, Daniel Herrmann Consulting

Daniel Herrmann Consulting

Study documentation aligned with Rave evidence

We support you in translating supplier invoices, study configurations, and operational responsibilities into a verifiable documentation scope.

  • Cloud and SaaS validation →

    Maintain the validated state of cloud and SaaS systems as suppliers issue new versions. We define the boundary between your responsibilities and the supplier’s and agree how each release will be assessed.

Send a project enquiry Customer testimonials →