A case report form (CRF) is a standardized document used in clinical trials to collect data about each study participant. Every measurement, observation, and outcome specified by the trial’s protocol gets recorded on this form, creating the structured dataset that ultimately determines whether a treatment works. Whether printed on paper or built into an electronic system, the CRF is the primary tool that turns raw clinical observations into analyzable research data.
What a CRF Actually Does
A CRF captures the specific information needed to answer a study’s research question. It’s not a medical chart or a patient record. It’s a research-specific document designed to collect only the data points the study protocol requires. Participants are identified by subject numbers rather than names, so the forms can be shared with sponsors and statistical teams without compromising privacy.
Beyond data collection, CRFs serve several other functions in a clinical trial. They ensure the study protocol is being followed at every site. They document who performed each research task and when. And they create a traceable record that regulators can audit years later to verify the trial’s integrity. The International Council for Harmonisation (ICH), which sets global standards for clinical research, formally defines a CRF as “a printed, optical, or electronic document designed to record all of the protocol required information to be reported to the sponsor on each trial subject.”
What Goes Into a CRF
The contents of a CRF vary by study, but they typically revolve around safety and efficacy endpoints. A CRF for a cancer drug trial, for instance, would include sections for tumor measurements, lab results, side effects, and dosing records. A CRF for a depression study might include validated questionnaires like the Beck Depression Inventory alongside vital signs and medication logs.
Common data domains across most CRFs include:
- Demographics: age, sex, race, and other baseline characteristics
- Medical history: relevant prior conditions and treatments
- Vital signs and physical exams: collected at scheduled study visits
- Adverse events: any side effects or health problems during the trial
- Concomitant medications: other drugs the participant is taking
- Efficacy outcomes: whatever the study is measuring to determine if the treatment works
Industry standards from organizations like CDISC help harmonize how these domains are structured across trials, so data collected at different sites and by different sponsors can be compared and submitted to regulators in a consistent format.
Paper CRFs vs. Electronic CRFs
The pharmaceutical industry has largely migrated from paper CRFs to electronic versions (eCRFs) due to improvements in data quality and cost efficiency. Paper forms haven’t disappeared entirely, particularly in smaller academic studies or resource-limited settings, but electronic systems now dominate industry-sponsored trials.
Paper CRFs carry a greater risk of missing data, especially if forms are poorly designed or not administered carefully. Best practices for paper include using a common format for date fields, headers, and response options, numbering pages clearly (e.g., “1 of 5”), using large fonts, and avoiding complex skip patterns that require the person filling out the form to jump between sections.
Electronic CRFs solve many of these problems by building in automated checks. An eCRF system can flag a blood pressure reading that falls outside a plausible range, pop up a calendar for date fields so formats stay consistent, and enforce skip logic automatically so the right questions appear based on previous answers. These built-in safeguards catch errors at the point of entry rather than weeks later during data review.
How CRFs Are Designed
Good CRF design starts before a single question is written. A multidisciplinary team, typically including investigators, a biostatistician, and a data manager, develops the forms alongside the study protocol and statistical analysis plan. This concurrent development ensures the CRF collects exactly what’s needed to answer the research question, nothing more and nothing less.
The guiding principles are clarity and parsimony. Questions should be written at a 5th to 8th grade reading level with unambiguous response options. Each data point should be collected only once unless repeated collection serves a specific validation purpose. Free-text responses are minimized because they’re harder to analyze and more prone to inconsistency. For multi-site international trials, forms go through language translation and back-translation to ensure reliability across sites.
Before a CRF is finalized, the full team reviews it, and new forms are pilot-tested. Electronic versions undergo formal user acceptance testing to confirm that edit checks, skip logic, and data exports all work correctly. Form completion guidelines are developed at the same time as the CRF itself, so site staff know exactly how to fill out each field.
Rules for Filling Out and Correcting a CRF
International guidelines place clear responsibilities on investigators when it comes to CRF data. The investigator must ensure that data reported on the CRF is accurate, complete, legible, and submitted on time. When CRF data is derived from source documents like medical records or lab reports, it must be consistent with those source documents, or any discrepancies must be explained.
Corrections follow strict rules designed to maintain an audit trail. Any change to a CRF entry must be dated, initialed, and explained if necessary. The original entry can never be obscured. On paper, this means drawing a single line through the error so the original value remains readable. In electronic systems, the software automatically logs every change with a timestamp and the identity of the person who made it. The investigator is required to retain records of all changes and corrections.
These rules exist because regulators need to trust the data. If an entry was changed, they want to know who changed it, when, and why. A CRF without a clean audit trail raises red flags during an inspection.
How Errors Get Caught and Fixed
Even with careful design, discrepancies show up in CRF data. The process for resolving them is called query management. After a monitoring visit, the sponsor or clinical monitor reviews the data and flags any inconsistencies on a data clarification form. These might be values that don’t match the source records, missing fields, out-of-range numbers, or logic conflicts between related data points.
Site staff are typically expected to resolve these queries within five working days, though sponsor-specific timelines outlined in the study contract take priority. The resolution process involves reviewing the original source data, correcting or explaining the discrepancy, and returning the resolved query to the sponsor or monitor. For paper-based studies, resolved queries go into the corresponding CRF notebook.
Edit checks in electronic systems fall into several categories: endpoint checks for missing or out-of-range values, safety checks to ensure serious adverse events are reported promptly, protocol compliance checks to verify the visit schedule was followed, and external checks for data transferred from outside sources like central laboratories.
Source Data Verification
CRF data doesn’t exist in a vacuum. It’s drawn from source documents: the original medical records, lab printouts, imaging reports, and other primary records generated at the clinical site. Source data verification (SDV) is the process of comparing what’s on the CRF against those original records to catch transcription errors.
During monitoring visits, a clinical research associate reviews the CRF entries side by side with the source documents. If a patient’s lab report shows a hemoglobin level of 12.3 but the CRF says 13.2, that’s a transcription error that needs correction. SDV is one of the most time-consuming parts of clinical trial monitoring, and it’s a major reason electronic systems that pull data directly from source records are increasingly valued for reducing manual transcription altogether.
Who Keeps the CRF and for How Long
Both the sponsor and the investigator retain CRF records after a trial ends. Regulatory guidelines require that the sponsor not have exclusive control of CRF data, and the investigator must have continuous access to the data they reported. Completed, signed, and dated CRFs are classified as essential documents that must be archived. So are records of any corrections made after the initial data entry and a signature sheet documenting every person authorized to make entries or corrections on the forms.
Retention periods vary by jurisdiction and sponsor requirements but commonly extend 15 years or more after the trial concludes. These archived CRFs allow regulators to reconstruct the trial’s findings and verify data integrity long after the study has ended.