Why this matters
Starting enrollment before you have a validated way to capture data creates holes you can’t easily fix. Sites lose track of what was recorded and when, safety signals may be delayed, and auditors will question the credibility of the study record. A clear, compliant interim plan keeps data attributable, traceable, and visible until your Electronic Data Capture (EDC) system is live.
Core concepts & definitions
- Interim capture — A temporary, validated method to collect study data when the EDC is not yet available. Examples: paper Case Report Forms (CRFs), a controlled spreadsheet (it is then an electronic record, so it needs access control and an audit trail, which a plain spreadsheet lacks), or a site Electronic Medical Record (EMR) printout packet with standardized cover sheets (patient identifiers removed before any copy leaves the site).
- Source vs. transcription — An interim page is source when data are first written on it and a transcription when they are copied from the medical record; the protocol or a source document agreement says which. Either way the site keeps its own record of every page: the investigator keeps control of the site’s data, and the sponsor never holds the only copy.
- Back‑entry controls — How interim data are entered into the EDC later: who does it, how entries are dated, how original dates are preserved, and how each interim page is matched to each EDC record.
- Reconciliation — A line‑by‑line comparison of interim records to the EDC after back‑entry, with documented resolution of any differences.
- Visibility — Operational reporting during the interim period so teams can see enrollment, visit completion, and safety events in near real time.
- Audit trail (paper & electronic) — For paper, a correction is a single‑line strike‑through that leaves the original readable, initialed and dated, with a reason when it is not obvious (no correction fluid). For EDC, the system’s audit trail records who changed what and when, the old and new values and the reason, and never hides the original entry.
- Expectedness matrix — A simple table that lists what data are expected at each visit/time point (helps you see what’s missing). It is built from the protocol’s schedule of assessments and is unrelated to expectedness in safety reporting (whether a reaction is listed in the reference safety information).
- Safety intake — The specific path for Adverse Events (AEs) and Serious Adverse Events (SAEs) to reach pharmacovigilance (PV) within required timelines. It never waits for EDC: the site reports each SAE to the sponsor’s PV immediately (usually within 24 hours of learning of it; the protocol sets the timeline) and also records it on the interim AE page.
- Key roles — Investigator of Record (IoR, the site’s principal investigator, PI); Clinical Research Associate (CRA, monitors sites); Data Manager (DM); Pharmacovigilance (PV).
- Good data principles (ALCOA+) — Data should be Attributable, Legible, Contemporaneous, Original, Accurate, and also Complete, Consistent, Enduring, Available.
The process, step‑by‑step
Make the go/no‑go call
- When to use interim capture: EDC build not validated; user access/training not complete; randomization or safety modules not ready; essential integrations (electronic patient‑reported outcomes (ePRO), labs, devices) delayed. A paper form cannot randomize: if the randomization module is not ready, subjects are randomized through a separately validated method (e.g., a stand‑alone interactive response system, IXRS) or randomization waits.
- Decision record: Document the reason, scope, start date, and the roles responsible. Store in the Trial Master File (TMF).
Design and validate the interim path
Artifacts:
- Versioned interim CRFs/templates or controlled spreadsheet (locked fields, version/date on each page).
- Instructions for Use (IFU): how to fill fields, dating/signing rules, correction rules.
- Safety intake form with PV contact info and timelines.
Validation light‑touch (fit‑for‑purpose): pilot with sample data, verify calculations or controlled fields behave as intended, confirm filing/scanning process, and record outcomes.
Train and authorize sites
- Who is allowed to use the interim forms (trained staff delegated on the delegation log); where forms are stored; how to request new versions; how to handle errors.
- Keep a training log and distribution log for form versions. Retire prior versions visibly.
Define back‑entry controls (interim → EDC)
- Ownership: designate specific back‑entry users: trained site staff on the delegation log, since sites own data entry and correction in EDC.
- Dating rules: EDC entry date is system‑captured; source date (original date data were recorded) is captured in a dedicated field or audit‑friendly comment.
- Source linking: add the interim page ID (barcode or unique page number) into an EDC field for traceability.
- Two‑person check: either dual data entry or single entry with independent verification against the scanned source.
Plan reconciliation
- Scope: 100% reconciliation for all data collected during the interim period. Back‑entered AE/SAE data are also reconciled with the PV safety database (the usual safety↔EDC reconciliation by DM and PV).
- Method: side‑by‑side check (scanned interim page vs. EDC form) using a reconciliation log.
- Discrepancy handling: open EDC queries so the site corrects the data, and record the final outcome in the reconciliation log.
- Acceptance threshold: “Zero unresolved differences at final sign‑off.” Back‑entry can only be completed once EDC is live, so this sign‑off comes after cutover (the switch from interim forms to EDC). Escalate anything not resolvable with documented CAPA (Corrective and Preventive Action).
Maintain visibility while interim is active
- Operational key performance indicators (KPIs): enrolled, screened, completed visits, open safety events, overdue safety follow‑ups.
- Review cadence: send a weekly summary to study leadership and PV. It only tracks safety events: each SAE still reaches PV directly from the site, within the protocol’s timeline (usually 24 hours).
- Expectedness tracking: run a simple report from your expectedness matrix to spot missing forms.
Execute go‑live & cutover
- Exit criteria: EDC User Acceptance Testing (UAT), including the edit checks, passed and signed off (part of the validation that ICH E6(R3) and 21 CFR Part 11 expect); study training complete, with each user’s access provisioned only after their training; expectedness matrix configured; critical integrations verified.
- Cutover window: announce the last interim data‑collection date/time; stop using interim forms afterward.
- Archive interim records: scan, index, and file all interim pages with page IDs; lock the archive. The site keeps its paper pages; a scan replaces a paper original only if it is a certified copy (verified as complete and accurate).
- Post‑cutover check: confirm 100% back‑entry completed and 100% reconciliation closed with sign‑off.
What good looks like
- Signed, version‑controlled interim pack (forms + IFU) in the TMF; sites trained and acknowledged.
- Documented back‑entry SOP with clear dating rules and role assignments; evidence of dual entry or verified single entry.
- Reconciliation plan with logs showing each page matched to each EDC record; all differences resolved or escalated with CAPA.
- Transparent visibility: weekly dashboards for enrollment and safety; expectedness reports to catch missing items early.
- Clean cutover: UAT passed, access ready, training complete, integrations verified; interim archive scanned and locked; “zero open reconciliation” at sign‑off.
Worked examples
Example 1 — Cardiovascular trial, EDC delayed by 2 weeks (self‑contained)
Context: Multicenter cardiovascular (CV) study. EDC randomization and AE modules are not validated by the Site Initiation Visit.
Interim design
- Forms: Paper CRF packet (Screening, Baseline, Vital Signs, Medications, AE/SAE). Each page has a page ID and version/date in the footer.
- IFU: How to record vitals, rounding rules, correction rules (single‑line strike‑through, initial/date), and where the Investigator signs.
- Safety intake: One‑page SAE form with PV email/phone and “report within 24 hours.”
- Randomization: Sites randomize through a separately validated central randomization service until the EDC module passes validation, and write the assigned number on the interim page.
Controls
- Back‑entry: Delegated site coordinators perform single data entry into EDC once their accounts are live, with independent verification by the study DM team against the scanned interim pages; differences go to the site as queries. EDC has fields for Source Date and Interim Page ID.
- Reconciliation: A log (Subject, Visit, Form, Field, Interim Value, EDC Value, Outcome). Target: 0 unresolved differences at final sign‑off.
- Visibility: A simple spreadsheet dashboard counts Screened, Enrolled, Open AEs/SAEs, and Missing Expected Forms; emailed weekly to study leads and PV.
Cutover
- Exit criteria achieved: UAT passed, user access live, expectedness matrix configured.
- Back‑entry & reconciliation finished in three business days after go‑live.
- Archive: All interim pages scanned to PDF/A, indexed by Page ID, filed in TMF; the site keeps the signed paper pages in its own files.
Outcome: No missed SAEs; auditors can follow each EDC entry back to a signed, dated interim page. The team closes interim with a signed memo summarizing results.
Example 2 — Investigator turnover and EDC refusal (self‑contained)
Context: Mid‑study change of Investigator at one site. The new IoR is unfamiliar with the EDC and asks to “use clinic notes instead.”
Risk: Unstructured EMR notes cause missing fields and poor traceability.
Governance path
- Escalation: DM raises to Study Lead and Quality; decision captured in a governance memo.
- Options weighed: delegating EDC entry to trained site staff on the delegation log (the IoR still reviews and signs the casebooks in EDC), targeted training, or a time‑boxed limited interim packet.
- Interim compromise: Permit a limited interim packet (AE/SAE, ConMeds (concomitant medications), Visit Summary) with a 7‑day deadline to complete EDC training.
Controls & support
- Onboarding: Same‑day EDC access request; targeted training (1‑hour) on the forms used most at that site, with the account activated only after the training. The former investigator’s account is disabled and the delegation log updated.
- Back‑entry: Site coordinator enters; CRA checks 100% of entries against interim pages for the first two subjects.
- Reconciliation: DM reconciles 100% of the site’s interim pages to EDC, with a weekly log review until every page shows zero unresolved differences.
Outcome: Site transitions to EDC within a week. The governance memo, training logs, and reconciliation logs show a controlled, compliant response without shutting down enrollment.
Pitfalls & misconceptions
- “We’ll back‑enter later” without a plan. If interim forms aren’t validated and version‑controlled, you cannot prove data integrity later.
- Assuming safety can wait for EDC. Define and test a safety intake path (phone/email/fax form) with PV timelines from day one.
- Mixing versions. Sites sometimes print an old interim form. Use a distribution log and clearly mark superseded versions.
- No linkage between interim and EDC. Always capture a Source Date and Interim Page ID inside the EDC for traceability.
- Unilateral site closure by DM. Closure or suspension is a cross‑functional governance decision; DM provides evidence and options.
- Skipping the archive. Scanning and indexing interim pages is not optional—make the interim record enduring and findable.
Practice in the wild (checklist)
- Write a short decision memo: why interim is needed, scope, start date, responsible roles.
- Build interim forms (with page IDs and version footers) and a clear IFU; pilot them with sample data.
- Define safety intake (contact, form, 24‑hour clock) and brief PV.
- Train sites; record training and distribution logs; retire prior versions.
- Configure back‑entry rules (roles, Source Date capture, Interim Page ID field, dual entry or verified single entry).
- Set up reconciliation (log template, 100% scope, acceptance = zero unresolved differences).
- Publish interim KPIs (enrollment, expected vs. received forms, open AEs/SAEs).
- Confirm go‑live exit criteria (UAT passed, access provisioned, expectedness matrix live, integrations verified).
- Scan & archive all interim records; lock and index by Page ID.
- Obtain sign‑off that back‑entry and reconciliation are complete.
Key points to remember
- Do not start enrollment without a validated capture pathway.
- Use interim capture only when EDC is not yet ready (e.g., build not validated, access/training incomplete, or critical modules/integrations delayed); document the decision (scope, start date, roles) in the TMF.
- Interim materials must be version‑controlled and guided by an IFU; keep training and distribution logs and visibly retire superseded versions.
- Safety intake has a standalone, validated path from day one to meet PV timelines (e.g., 24‑hour SAE reporting).
- Back‑entry controls: assign ownership; preserve the original Source Date; capture an Interim Page ID in the EDC to ensure audit‑ready traceability; system captures the EDC entry date.
- Apply a two‑person check—either dual data entry or single entry with independent verification against the scanned interim source.
- Perform 100% reconciliation for all interim‑period data using a reconciliation log; acceptance threshold = zero unresolved differences at final sign‑off (after cutover); escalate exceptions with documented CAPA.
- Maintain operational visibility during interim (enrollment, visit status, open AEs/SAEs, expected‑vs‑received forms) using simple KPIs and an expectedness matrix; brief study leadership and PV weekly (each SAE still reaches PV directly from the site).
- Go‑live exit criteria: UAT (including edit checks) passed and signed off, study training complete before user access is provisioned, expectedness matrix configured, and critical integrations verified; announce a clear cutover window.
- Archive all interim records: scan, index by Page ID, and lock the archive to keep the record enduring and findable (ALCOA+).
- Governance & escalation: DM does not unilaterally suspend/close a site; escalate investigator refusal or persistent data gaps through study governance and document decisions.
- Post‑cutover assurance: confirm 100% back‑entry and reconciliation closed and obtain final sign‑off.