Trial Data Reviewed Against the Protocol.Every Finding Traced to Its Row.
CPAR reads listings, case report forms and the protocol together. It finds out-of-window visits, safety signals without a matching report and data that breaks a protocol rule, and links every finding to the row, the rule and the protocol section behind it.
| Subject | Site | Visit | Day | ALT (U/L) | AST (U/L) | × ULN | AE / SAE |
|---|---|---|---|---|---|---|---|
| 1031 | 104 | Week 4 | 29 | 34 | 28 | 0.9 | — |
| 1038 | 104 | Week 4 | 30 | 41 | 33 | 1.0 | — |
| 1047 | 112 | Week 4 | 36 | 148 | 96 | 3.7 | None |
| 1052 | 112 | Week 4 | 28 | 29 | 25 | 0.7 | — |
| 1058 | 118 | Week 8 | 58 | 38 | 30 | 1.0 | — |
| 1063 | 118 | Week 8 | 71 | 52 | 41 | 1.3 | — |
| 1071 | 121 | Week 8 | 57 | 127 | 88 | 3.2 | AE 0412 |
| 1079 | 121 | Week 8 | 56 | 31 | 27 | 0.8 | — |
A sample liver function listing in CPAR. All subjects, studies and numbers are made up.
Four Reviews, One Platform
CPAR runs the reviews your clinical, safety and data teams do by hand today, on every row and every visit, as the data arrives.
- DataListing ReviewEvery listing checked against the protocol.
- DeviationsProtocol DeviationsDeviations found, classified and routed to sites.
- SafetySAE ReconciliationClinical and safety databases kept in agreement.
- ReportingReview SummariesMedical review summaries drafted, every finding cited.
- RecordAudit TrailEvery finding and decision on the record.
Deviations Found, Classified and Routed
CPAR reads visit dates, dosing, eligibility and procedures against the protocol and the schedule of assessments. Each deviation is classified by importance and routed to the site and the study team.
From the Schedule of Assessments
Visit windows, required procedures and dosing rules are read from the protocol and its amendments.
Classified by Importance
Each deviation is classified as important or not, using your deviation plan.
Routed to the Site
Site-level deviations go to the site with the evidence, and trends go to the study team.
Subject 1047 · Site 112
Event termDrug-induced liver injuryAI
Onset date2026-03-18AI
OutcomeRecoveringAI
CausalityRelatedAI
Narrative: ALT rose to 3.7× ULN at the Week 4 visit. Study drug was interrupted, and liver tests were improving at follow-up.
Clinical and Safety Data in Agreement
Serious adverse events are recorded in the clinical database and the safety database, and they drift apart. CPAR compares them field by field and sends every mismatch to the right person to resolve.
Field by Field
Onset dates, terms, seriousness, outcome and causality are compared for every SAE.
Both Sources Shown
Each mismatch shows the clinical and safety values side by side, with their sources.
Resolved by People
Data managers and safety staff resolve each mismatch, and the resolution is recorded.
Medical Review Summaries, Drafted and Cited
At each data cut, CPAR drafts the medical review summary from the findings, the listings and the protocol. Every statement cites the listing or finding behind it, and the medical monitor edits and signs the final version.
Drafted at Every Data Cut
Summaries are ready when the data cut is, not weeks later.
Every Statement Cited
Each sentence links to the listing rows and findings it summarizes.
Signed by the Monitor
The medical monitor reviews, edits and signs every summary.
Medical Review Summary · Data cut March 31
3. Hepatic Safety
Two subjects had ALT above 3× ULN at Week 4 or Week 8, compared with none at the previous data cut.1 Subject 1047 had no adverse event recorded, and a query has been raised with site 112.13
Both visits for subject 1047 and subject 1063 fell outside the protocol window.24 No case met the protocol definition of potential drug-induced liver injury.4
This record is append-only. Nobody can change or delete an entry.
- Listing loaded from EDCListing 16.2.6, 412 rows
- Rule applied by CPARProtocol 8.3.2, ALT above 3× ULN
- Finding raised by CPARSubject 1047, no AE or SAE
- Assigned by Jamie TorresMedical monitor Sam Okonkwo
- Query raised by Sam OkonkwoSite 112, response due in 5 days
- Site responded by Site 112AE entered, AE 0418
- Finding closed by Sam OkonkwoResolved, AE confirmed
Every Finding, From Data to Decision
Every finding keeps its rule, its source rows and every decision made about it. Inspectors and auditors see exactly what was found, who reviewed it and how it was closed.
Inspection-Ready
Findings, decisions and signatures export for an inspection or an audit.
Every Data Cut Kept
Findings are kept for every data cut, so changes over time are clear.
Study Oversight
Findings by site, type and age show where a study needs attention.
Built for Clinical Development
ThinkTrends runs in your environment, keeps a qualified person on every decision, and keeps your data out of model training.
People Close Every Finding
CPAR finds and explains. Monitors, data managers and safety staff decide, and every decision is recorded under a name.
Your Data Never Trains a Model
Cases, labels, trial data and documents are used to do the work and nothing else.
Runs in Your Environment
Deploy in your cloud or data center, with patient and product data kept inside your boundary.
Validation Evidence Included
Each agent version is tested against your acceptance criteria, and the results are ready for your validation file.
21 CFR Part 11 Aligned
An audit trail no one can edit, electronic signatures with their meaning, and access by role.
Configured, Not Coded
Rules, routes, checks and thresholds are settings your administrators can read and change.
Related Modules
The platform modules these solutions are built on.
Frequently Asked Questions
Common questions about ThinkTrends CPAR — Clinical Trial Review.
What is CPAR?
+
CPAR is ThinkTrends’ agentic AI for clinical trial data review. It reads listings, case report forms and the protocol together, finds out-of-window visits, unreported safety signals and protocol violations, and links every finding to the row, the rule and the protocol section behind it.
How is CPAR different from EDC edit checks or manual listing review?
+
EDC edit checks catch field-level errors as data is entered, and manual review covers a sample. CPAR reviews every row of every listing at every data cut against rules read from the protocol, and reconciles safety data across systems.
Where does CPAR get the protocol rules?
+
From the protocol and its amendments. Visit windows, eligibility, dosing and safety thresholds are read from the document and confirmed by your team before they run.
Can CPAR reconcile SAEs between the clinical and safety databases?
+
Yes. It compares the clinical and safety databases field by field and sends every mismatch to the right person to resolve.
Does CPAR replace our EDC?
+
No. CPAR reads from your EDC, safety database and listings, and sends findings and queries back to them.
Is our data used to train AI models?
+
No. ThinkTrends uses trial data only to do the work, inside your own environment. Your data is never used to train, tune or evaluate a model.