Source Data Verification (SDV) refers to the verification that data entered into a Case Report Form (CRF) or Electronic Data Capture (EDC) system match the original source data at the investigational site. The objective is to ensure data integrity, identify discrepancies early, and establish a reliable basis for statistical analysis and regulatory decisions.
What qualifies as source data and what is verified?
Source data are all original records that document the conduct of a clinical trial and the patient data collected. These include patient records, laboratory reports, physician letters, imaging findings, medication plans, device printouts, ePRO or eDiary entries, or digital raw data from medical devices.
Which source is authoritative should be defined in advance. For electronic sources, it must be clarified whether they constitute “original source” (directly captured) or transcribed data. For hybrid processes (paper plus digital), clear rules are required so that in case of discrepancies, it is unambiguous which entry is valid. This clarity is also important for subsequent audits and inspections, as it reduces discussions about retrospective additions and responsibilities.
- Concordance: Values, date/time, unit, reference range, visit association, and plausibility.
- Completeness: Are all relevant measurements, findings, and endpoints documented?
- Traceability: Are corrections dated, justified, and traceable (audit trail)?
- Protocol compliance: Were inclusion and exclusion criteria, visit windows, and examination schedules adhered to?
SDV in the context of monitoring and data management
SDV has historically been a central component of on-site monitoring. However, modern trials increasingly rely on Risk-Based Monitoring (RBM) and a risk-oriented quality strategy to reduce effort while protecting critical data and processes. SDV is deployed strategically where errors could jeopardize patient safety or the primary conclusions of the trial.
Typical SDV focus areas include informed consent documentation, inclusion criteria, primary endpoints, randomization, investigational product documentation, and safety-relevant events. In addition, query management, data validation rules, and central plausibility checks in the EDC are used to identify anomalies early. It is important that SDV, queries, and central checks are coordinated to avoid duplicate verification while preventing gaps.
Risk-based SDV and remote elements
The central question is not “100% SDV or none at all?” but rather what level of SDV is appropriate in relation to risk, trial design, and data flow. Frequently, SDV quotas are defined, e.g., 100% for critical variables and sampling for less critical data. It is essential that selection, justification, and adjustments over the course of the trial are documented, for example in the Monitoring Plan and Data Management Plan.
Remote monitoring can include SDV elements if the investigational site provides verified copies or enables secure access. Data protection, access controls, role-based permissions, and proper logging are central. In Germany, local data protection requirements, patient rights, and contractual arrangements between sponsor, CRO, and investigational site must also be observed to ensure that access remains permissible and traceable.
Common errors and misunderstandings in practice
A common misconception is to automatically equate SDV with high data quality. SDV detects errors but does not prevent them. More effective are robust processes at the investigational site, clear CRF completion instructions, training, clean interfaces (e.g., laboratory transfers), and a sound query strategy. SDV is most effective when it begins early and is used as a feedback loop.
- SDV too late: If SDV is performed only shortly before database lock, the risk of time pressure and open queries increases.
- Focus on non-critical fields: Routine values are fully verified while endpoints are insufficiently prioritized.
- Unclear responsibilities: If it is not clear who documents and who authorizes corrections, data integrity suffers.
- Inconsistent correction rules: Lack of processes jeopardizes traceability and audit trail quality.
For sponsors, SDV is also a management tool: if similar deviations occur across multiple centers (e.g., incorrect units, missing visit data), this may indicate training needs or unclear CRF guidelines. In such cases, targeted training is often more effective than a blanket increase in SDV quota.
Relevance for clinical trials
For sponsors and CROs, SDV is a lever for quality assurance but also a cost driver. A well-designed SDV strategy reduces monitoring effort without compromising the validity of the trial. In the EU/German context, it is also relevant that inspections verify the traceability of data flows, corrections, and consistency between source, CRF, and database. Full-service CROs such as mediconomics support the integration of SDV into a risk-based framework of monitoring, central data review, and quality management.
Frequently Asked Questions (FAQ)
Is 100% source data verification required by regulation?
No. A risk-based approach is expected: critical data and processes must be appropriately monitored, but a blanket 100% verification of all data is not required. What is essential are justification and documented implementation in the Monitoring Plan.
Which data should typically always be verified by SDV?
Common practice includes informed consent documentation, inclusion criteria, primary endpoints, randomization, investigational product (e.g., accountability), and safety-relevant events. The specific selection depends on the protocol and data risk.
How does SDV differ from central monitoring?
SDV compares CRF/EDC entries with source data at the investigational site, often on-site. Central monitoring uses aggregated data analyses, plausibility checks, and quality indicators across all centers to identify risks and patterns early.
Regulatory References
- ICH E6(R3) Good Clinical Practice: Expectation of risk-based quality management and appropriate oversight of critical data and processes.
- Regulation (EU) No 536/2014 (Clinical Trials Regulation): Requirements for the conduct of clinical trials and the reliability of the data basis.
- EU GMP Guideline, Annex 11 (Computerised Systems): Relevance for data integrity and audit trail in computerized systems (e.g., EDC) in the quality context.