Free Lesson Preview
Enjoyed this preview?
The full CRC track covers 8 courses from study start-up to close-out — the skills sponsors actually look for.
View enrollment optionsFree GCP training
Professional tracks
Clinical Research Coordinator
Full course · Data Collection and Source Documentation
Free Lesson Preview
Enjoyed this preview?
The full CRC track covers 8 courses from study start-up to close-out — the skills sponsors actually look for.
View enrollment optionsApply worksheet design principles to create SDV-ready source documentation, incorporate quality prompts, evaluate worksheets against protocol requirements, and manage version control.
You are about to walk into the exam room for Visit 4. The protocol requires vital signs at three timepoints, a focused physical examination, two participant questionnaires, a timed blood draw 90 minutes after investigational product administration, a fasting lipid panel, a 12-lead ECG, an adverse event review, and a concomitant medication reconciliation. You have approximately two and a half hours. Nine discrete procedures. Multiple data points for each one. And every value you record will, at some point, be compared against the CRF by a monitor who expects to find a complete, legible, organized source record for every entry.
If the source document worksheet sitting in that participant's binder was thoughtfully designed, it will walk you through the visit in the order the procedures actually happen. It will remind you what units to use. It will tell you the normal range for each lab value. It will have a checkbox for the timed blood draw that says "Time of IP administration: : / Time of blood draw: : / Elapsed time: ___ minutes." It will have a signature block where the investigator initials the physical exam findings. It will have space for an observation you did not anticipate -- the participant mentioning a new medication their cardiologist started last Tuesday.
If the worksheet was poorly designed -- or worse, if someone simply photocopied the CRF pages and called that "source" -- you will miss something. You may not realize it until the monitor is sitting across from you three weeks later, pointing to an empty field and asking, "Where is the systolic blood pressure for the second timepoint?"
This lesson teaches you how to design worksheets that prevent that moment.
What you will learn4 objectives
By the end of this lesson, you will be able to:
A protocol tells you what data to collect. A CRF tells you where to enter it. But neither one tells you how to collect it in real time, during a visit, with a participant sitting in front of you. That gap -- between the protocol's requirements and the practical reality of data collection -- is precisely what a source document worksheet fills.
I want to be clear about what a source document worksheet is and is not. It is not a copy of the CRF. It is not a pre-printed form that the sponsor sends along with the study materials and that you file without thinking. A well-designed source document worksheet is a working tool -- a structured document that guides the coordinator through the data collection process in the order the visit actually unfolds, capturing each observation at the point of care with the specificity and structure that data integrity demands.
The worksheet serves three functions simultaneously. First, it is a : it tells you what to measure, record, and verify at each step of the visit, reducing the risk that a data point is forgotten. Second, it is the : as we established in the previous lesson, when the coordinator records a measurement on the worksheet at the time of observation, that worksheet entry is the first recording of the data and therefore constitutes the source per ICH E6(R3). Third, it is a : during source data verification, the monitor compares the CRF against the worksheet. A well-organized worksheet makes that comparison efficient; a disorganized one makes it painful.
Over many years of reviewing source documentation across dozens of study sites, I have identified six design principles that distinguish worksheets that survive SDV without findings from those that generate monitoring visit reports full of documentation deficiencies. None of these principles is revolutionary. All of them are violated routinely.
The worksheet should follow the chronological sequence of the visit, not the organizational structure of the CRF. If the protocol requires vital signs before dosing, the ECG after dosing, and lab draws at 90 minutes post-dose, the worksheet should present those sections in that order. The coordinator should be able to move through the worksheet page by page as the visit unfolds, without flipping back and forth.
This seems obvious, yet many sponsor-provided source templates organize data by CRF module: demographics first, then vital signs, then physical exam, then labs, then adverse events. That organization makes sense for the data manager. It does not make sense for the coordinator who is standing next to the participant, measuring blood pressure, and needs to record the value immediately without searching for the right page.
Design principles create the structure. Quality prompts create the behavior. A quality prompt is any element printed on the worksheet that reminds, guides, or requires the coordinator to perform a specific verification or documentation step. Think of quality prompts as the worksheet talking to the coordinator -- nudging them toward completeness and accuracy at the moment it matters most, rather than relying on memory or habit.
At the top of each visit worksheet, a pre-printed checklist can prompt the coordinator to verify preconditions before beginning data collection. A typical pre-visit checklist might include:
Designing a worksheet is not a one-time exercise followed by blind faith. Before the first participant uses a worksheet, it must be systematically evaluated against the protocol to verify that every required data point has a corresponding field, every field uses the correct units and format, and the visit windows are accurately calculated. This evaluation is the coordinator's quality gate -- the last opportunity to catch gaps before they become missing data.
The most reliable method is a structured crosswalk: take the protocol's schedule of assessments, and for each data point listed at each visit, verify that the worksheet contains a field for it. This is tedious work. It is also essential. I have reviewed worksheets where an entire assessment was missing because the person who designed the worksheet worked from an earlier version of the schedule of assessments, or because a protocol amendment added a new assessment that no one carried forward to the worksheet.
The crosswalk should verify four elements for each field:
TableProtocol-to-worksheet crosswalk verification elements
| Field presence | Every data point in the schedule of assessments has a corresponding field on the worksheet | Assessment added in a protocol amendment is missing from the worksheet entirely |
| Unit alignment | The units on the worksheet match what the CRF expects, or a conversion field is included | Worksheet captures weight in pounds; CRF requires kilograms; no conversion field exists |
| Format consistency | Date formats, time formats, and numerical precision match the CRF requirements | Worksheet uses MM/DD/YYYY; CRF uses DD-MMM-YYYY; coordinator must mentally convert during transcription |
| Visit window accuracy | The worksheet prints the correct visit window dates (calculated from the enrollment or randomization date per the protocol) | Visit window calculated from the wrong anchor date, or protocol amendment changed the window and worksheet was not updated |
A worksheet can pass the crosswalk -- every field present, every unit correct -- and still fail in practice because it is unusable. Usability failures are more insidious than missing fields because they do not show up in a desk review. They show up when the coordinator is trying to use the worksheet during a visit and discovers that the layout does not match the workflow.
Ask these questions during the evaluation:
Does the worksheet's page order match the sequence in which procedures are actually performed? If vital signs are taken before dosing but appear after the dosing section on the worksheet, the coordinator must flip backward to record them -- and might not.
Are the fields large enough for legible handwriting? A blood pressure field that provides a quarter-inch of writing space guarantees illegible entries. Legibility is a data integrity criterion per ICH E6(R3).
Is there enough space for the concomitant medication section? If the protocol population tends to be on multiple medications -- elderly participants in a cardiovascular trial, for instance -- a concomitant medication section with three lines is inadequate. Design for the realistic participant, not the ideal one.
Protocols change. Amendments add assessments, remove them, change visit windows, modify dosing schedules, or revise the schedule of events. When a protocol amendment affects the data collected at visits, the source document worksheet must be updated to reflect the new requirements. This is not optional and it is not something that can be deferred -- participants visiting on the amended protocol must use a worksheet that captures the amended requirements.
When a protocol amendment is received, the coordinator should perform a systematic review to determine whether the worksheet requires revision. The review should address three questions:
Does the amendment add, remove, or modify any data collection requirements? If the schedule of assessments changes, the worksheet must change. If new assessments are added, new fields are needed. If existing assessments are removed, the corresponding fields should be removed or clearly marked as no longer applicable (to avoid confusing the coordinator into collecting data the protocol no longer requires).
Source document worksheets come from two sources, and the version control implications differ for each.
Sponsor-mandated source documents are worksheets provided by the sponsor as part of the study materials. These are controlled documents -- the site cannot modify them without sponsor approval. When a protocol amendment affects a sponsor-mandated worksheet, the sponsor is responsible for providing the updated version. The site's responsibility is to implement it promptly and confirm that the new version captures all amended requirements. If the sponsor's update is delayed, the site should document the gap and use an interim supplementary page.
are created by the site to supplement the sponsor's materials or to serve as the primary source document when the sponsor does not provide one. These give the site more control over design and content, but they also place the full burden of version control on the site. When developing or revising a site worksheet, the coordinator should confirm with the sponsor that the worksheet captures all required data points. Some sponsors review and approve site-developed worksheets; others simply require that the site's source documentation be adequate for SDV. Either way, the site retains responsibility for keeping the worksheet current with the protocol.
Case study · Clinical Research · intermediate · 10-15 minutes
"The Missing Assessment Fields"
Scenario
Jennifer Rodriguez has been the Senior CRC on the HORIZON Study at Riverside Medical Center for 14 months. The HORIZON Study -- a Phase II dose-ranging trial of a novel anticoagulant in patients with atrial fibrillation -- has enrolled 38 participants across four dose cohorts, with most participants in the maintenance phase of the study.
Three weeks ago, the sponsor issued Protocol Amendment 3, which added two new assessments to Visits 6 through 10: a standardized bleeding assessment questionnaire (BAQ) and a structured medication adherence interview. The amendment was reviewed at the site, the IRB approved the modification, and the study team began implementing the changes -- updating the consent form, notifying active participants, and adjusting the visit schedule.
What Jennifer did not do was update the source document worksheet. The worksheet in use -- Version 2.0, effective since Protocol Amendment 2 -- did not include fields for either the BAQ or the medication adherence interview.
Over the following three weeks, three participants completed Visit 7 using the outdated worksheet. The coordinator conducting those visits -- a newer team member -- administered the BAQ and the adherence interview verbally, as the protocol required, but had no structured place to document the responses. The BAQ responses were jotted on a sticky note attached to the worksheet. The adherence interview responses were recorded in the margins of an unrelated page. For one participant, the adherence interview was not documented at all -- the coordinator recalled conducting it but could not locate any written record.
Jennifer discovers the problem when she begins preparing for the upcoming monitoring visit and realizes the worksheets for those three participants are missing entire assessment sections. She contacts Dr. David Park, the sponsor's medical monitor, to discuss the situation.
Dr. Park's response is measured but direct: "We need three things. First, reconstruct whatever data you can from the existing records -- the sticky notes, the margin entries, anything in the medical record. Document the reconstruction effort, including what was recoverable and what was not, with an explanation for each data point. Second, create the updated worksheet immediately -- Version 3.0 -- with fields for both new assessments. Third, write a protocol deviation report for each affected participant documenting that the assessments were performed but not captured on an adequate source document."
Question 1 of 3. Not yet answered.
Progress saves itself — finish reading and this lesson completes
The distinction between a worksheet and a CRF bears emphasis because I have seen the confusion cause real problems. The CRF is designed for data management -- it is structured by domain (demographics, vitals, labs, adverse events) to support data analysis and regulatory review. The worksheet is designed for data collection -- it is structured by visit flow to support the coordinator's workflow. These are different organizing principles, and conflating them produces a document that serves neither purpose well.
Every page of the worksheet must include the participant number (or study identifier), the visit name or number, and the date. This is not merely good practice -- it is a data integrity requirement. If a page separates from the worksheet, it must be identifiable as belonging to a specific participant at a specific visit. Pages without identification are orphaned data: attributable to no one.
I have seen sites where the participant identifier appeared only on the first page of a multi-page worksheet. When the monitor reviewed the binder and found a loose page with vital sign recordings but no participant number, the data on that page were essentially unverifiable. The coordinator remembered which participant it belonged to. The monitor, correctly, pointed out that memory is not documentation.
Every field that captures a quantitative measurement must pre-print the expected unit. Blood pressure in mmHg. Temperature in degrees Celsius (or Fahrenheit, depending on the protocol's convention). Weight in kilograms. Creatinine in mg/dL or micromol/L. If the protocol and the CRF expect kilograms but the clinic scale reads in pounds, the worksheet should include both fields -- the measured value in pounds and the converted value in kilograms -- with the conversion formula printed on the page.
Unit errors are among the most common causes of data queries. A coordinator records "72" for heart rate. The CRF asks for heart rate in beats per minute. Is 72 in beats per minute? Almost certainly. But "almost certainly" is not what monitors want. They want the unit on the source document, matching the unit on the CRF, leaving no room for inference.
For clinical assessments that have established reference ranges -- vital signs, laboratory values, physical examination parameters -- the worksheet should print the protocol-specified normal ranges adjacent to the data entry field. This serves two purposes: it prompts the coordinator to flag abnormal values at the point of collection, and it provides the context that the investigator needs when reviewing the data and making clinical significance determinations.
A blood pressure field that reads "/ mmHg" is adequate. A blood pressure field that reads "/ mmHg (Protocol-defined criteria for intervention: SBP >180 or <90; DBP >110 or <50)" is substantially better. The second version prompts the coordinator to act if the reading meets intervention criteria, rather than discovering the issue hours later during data entry.
For every procedure the protocol requires at a visit, the worksheet should include a completion check -- a checkbox or Y/N field indicating that the procedure was performed. This is separate from the data captured during the procedure. The check field answers the question "Was it done?" The data fields answer the question "What was found?"
This distinction matters because some procedures generate no data when the finding is normal. A physical examination that reveals nothing abnormal may produce only "WNL" (within normal limits) in the data fields. But the check field confirms that the examination was performed -- that the investigator actually examined the participant, rather than simply assuming nothing changed since the last visit. During SDV, the monitor can scan the check fields to confirm all required procedures were completed, then drill into the data fields for the specific values.
No worksheet can anticipate every observation. A participant mentions a new symptom in passing. A blood draw requires four attempts before a viable sample is obtained. The investigator notices a finding during the physical examination that does not fit any predefined category on the worksheet. The worksheet must provide unstructured space -- a comments section on each page, or a dedicated observations area -- where the coordinator can document what actually happened.
This is, in my view, the most underappreciated design principle. Worksheets that consist entirely of structured fields and checkboxes create an illusion of completeness while leaving no room for the messy, unpredictable reality of clinical encounters. The comment space is where data integrity lives in its most honest form: the coordinator recording what they observed, in their own words, at the time it happened.
Each item gets a checkbox. The coordinator checks each one before collecting the first data point. This takes 30 seconds and prevents errors that take hours to remediate -- the participant who was not actually fasting for their fasting lipid panel, the consent that expired two weeks ago, the outstanding adverse event from the prior visit that was never followed up.
Not every field on a worksheet carries equal weight. Some fields are critical to the primary endpoint. Others capture secondary or exploratory data. Mandatory field indicators -- an asterisk, a bold border, a shaded background -- visually distinguish the fields that must be completed from those that are important but not strictly required. This does not excuse incomplete documentation. But it directs the coordinator's attention to the data points where an omission would be most consequential.
The alignment here should be to the protocol's Critical to Quality (CTQ) factors, a concept formalized in ICH E6(R3) Section 3.10. If the sponsor has identified specific data points as critical to the trial's objectives, the worksheet should visually emphasize those fields. The coordinator should know, at a glance, which measurements are primary endpoint data.
Any procedure that is time-dependent -- a post-dose blood draw, an ECG at a specific interval, a pharmacokinetic sample -- requires explicit time documentation. The worksheet should include separate fields for:
Pre-printing the acceptable window on the worksheet means the coordinator can verify in real time whether the procedure was performed within the protocol-specified window, rather than discovering a timing deviation during data entry or, worse, during monitoring.
For assessments that require the investigator's clinical judgment -- physical examination findings, eligibility confirmation, adverse event causality, clinical significance determinations -- the worksheet must include a designated signature or initial block with a date field. This is not optional decoration. Per ICH E6(R3), certain assessments must be attributable to the investigator. A worksheet that captures the investigator's findings without a mechanism for the investigator to attest to those findings creates an attributability gap.
The signature block should be positioned immediately adjacent to the assessment it covers -- not at the end of the worksheet in a generic "investigator approval" section. The investigator should initial the physical exam findings after completing the exam, not after reviewing 10 pages of data at the end of the visit. Contemporaneous attestation is the standard.
Does the worksheet accommodate protocol deviations? If a timed procedure is performed outside its window, is there a field to document the actual time and a space to note the deviation? Or does the worksheet assume everything will go according to plan?
These are the questions that separate a worksheet designed by someone who has coordinated visits from one designed by someone who has only read the protocol.
Does the amendment change any visit windows, timing requirements, or procedure sequences? If a blood draw window changes from 90 +/- 10 minutes to 60 +/- 15 minutes, the pre-printed window on the worksheet must be updated. If procedures are resequenced, the worksheet page order may need revision.
Does the amendment change units, scales, or reference criteria? If the protocol changes the threshold for a clinically significant vital sign finding, and that threshold is printed on the worksheet as a reference range, it must be updated.
Every version of the worksheet must carry a version number, an effective date, and -- ideally -- a brief change description. When a new version is deployed, the prior version must be retired from active use. "Retired" means physically removed from participant binders (for sites using paper worksheets) and replaced with the current version. The retired version must be retained in the site's files -- it is part of the audit trail showing what document was in use during earlier visits -- but it must not be available for active data collection.
The IRB does not typically review source document worksheets (they are operational documents, not participant-facing materials), but if a worksheet includes any participant-facing content -- instructions, questionnaire text, or information that participants see during the visit -- the IRB may need to review it. When in doubt, consult your IRB.
The challenge: Consider what went wrong and what a strong remediation plan looks like.
Analysis