SDV stands for source data verification, a quality control process in clinical trials where a monitor compares the data entered into a study’s electronic forms against the original patient records. Its purpose is straightforward: catch transcription errors before they compromise the integrity of trial results. If a patient’s blood pressure was recorded as 130/85 in the medical chart but entered as 135/85 in the study database, SDV is the step designed to find and fix that mismatch.
How SDV Works in Practice
During a clinical trial, study coordinators at each research site collect patient data and enter it into case report forms, which are the standardized documents used to record every observation and outcome in the study. Once that data entry is marked complete, a clinical research associate (often called a monitor) visits the site, either in person or remotely, and systematically checks the entered data against the original source documents.
Source documents include anything that constitutes the original record of a patient’s clinical experience: hospital charts, lab printouts, imaging reports, signed consent forms, prescription records, even the output from medical devices. The FDA defines source data broadly as “all information in original records and certified copies of original records of clinical findings, observations, or other activities in a clinical investigation.”
When the monitor spots a discrepancy, they raise a query, which is essentially a formal question sent back to the coordinator asking them to explain or correct the entry. The coordinator responds, the monitor reviews the fix, and once everything matches, the monitor closes the query and locks the form. In some studies, the principal investigator then applies an electronic signature to confirm the data is accurate.
What SDV Can and Cannot Catch
SDV is specifically designed to detect transcription errors, the mistakes that happen when someone copies a number, date, or result from one place to another. Research into the types of errors found during SDV confirms that transcription mistakes and inexperienced data entry personnel are the primary sources of discrepancies.
But SDV has a well-recognized blind spot. Because it only compares what’s in the source record to what’s in the case report form, it cannot detect problems that occurred before the data was recorded. If a research coordinator collected a measurement incorrectly, misinterpreted the study protocol, or failed to document something altogether, the source document itself is already wrong. SDV will confirm that the wrong number was copied accurately, which is not especially helpful. It also cannot identify patterns across multiple patients or sites, like a systemic protocol deviation or a recurring safety signal.
Analysis of monitoring queries shows that SDV generates a relatively small share of the issues monitors flag. On average, SDV queries account for about 7.8% of all monitoring queries, and only 2.4% of total queries involve critical data points. Most of the important findings come from a different, broader process.
SDV vs. Source Data Review
Source data review, or SDR, is often mentioned alongside SDV but serves a different purpose. Where SDV asks “does this number match?”, SDR asks “does this data make sense in context?” The Association of Clinical Research Organizations describes SDR as a deeper, more holistic review that evaluates whether a site is following standard operating procedures, quality standards, and good clinical practice overall.
An SDR might reveal, for example, that a patient received a study drug dose that doesn’t align with the protocol’s dosing schedule, or that consent was obtained after a procedure rather than before it. These are problems SDV would never catch because the data in the case report form could perfectly match the source document while still reflecting a protocol violation. Modern monitoring programs typically use both SDV and SDR together, with SDR carrying more of the weight when it comes to protecting patient safety and ensuring the trial was conducted properly.
The Shift Away From 100% SDV
For decades, the standard approach was to verify every single data point at every site visit, a practice known as 100% SDV. This was enormously time-consuming and expensive, and growing evidence suggested it wasn’t the best use of monitoring resources. The FDA now recommends a risk-based approach to monitoring, where sponsors decide how much SDV to perform based on an assessment of what could go wrong and how much it would matter.
The factors that influence this decision include the complexity of the study design, the types of endpoints being measured, how experienced the investigators are, whether the site uses electronic data capture (which reduces manual transcription), the safety profile of the drug being tested, and the overall volume of data. For non-critical data points, the FDA guidance states that full verification may not be necessary if there are no red flags like data anomalies or inconsistencies. In those cases, checking a sample of subjects and visits can be enough to gauge overall accuracy.
This risk-based model lets sponsors concentrate their monitoring effort where it matters most. A trial testing a new cancer therapy with complex dosing and serious potential side effects will warrant more intensive SDV than a late-stage study of a well-understood medication with a strong safety record. The goal is smarter monitoring, not less monitoring.
Remote Monitoring and SDV
Traditionally, SDV required a monitor to physically travel to a research site and sit down with paper charts or hospital electronic health records. This changed significantly during the COVID-19 pandemic and has continued to evolve. The FDA recognizes remote regulatory assessments as valuable oversight tools and has issued guidance on how they can be conducted securely, including requirements for adequate internet connectivity, IT security, and privacy controls to protect patient data.
Remote SDV typically works through secure, read-only access to a site’s electronic health record system or through a platform that lets monitors view redacted source documents uploaded by site staff. This approach reduces travel costs and allows more frequent review of incoming data, though it requires careful handling of patient privacy. Federal laws governing the confidentiality of health information still apply, and sites need appropriate technical infrastructure to support remote access safely.
Why SDV Still Matters
Even as the industry shifts toward risk-based and centralized monitoring strategies, SDV remains a core component of clinical trial oversight. Regulatory agencies rely on the ability to reconstruct a trial’s findings from original records, and SDV is the process that ensures the reported data actually traces back to real patient experiences. When a pharmaceutical company submits data to support drug approval, regulators need confidence that the numbers in the submission reflect what actually happened at the clinical sites.
For anyone working in clinical research, whether as a coordinator entering data, a monitor reviewing it, or a project manager designing a monitoring plan, understanding SDV means understanding both its value and its limits. It is excellent at catching copy errors. It is not a substitute for the broader oversight activities that ensure a trial was conducted ethically and according to its protocol.