Clinical Data Management (CDM) Interview Questions – Section Wise Guide
🧠 Practice Before Interview
Test your CDM database, coding, and validation logic with our interactive quiz.
🚀 Take CDM Practice QuizClinical Data Management (CDM) is a highly critical phase in drug development. This guide provides the most important CDM interview questions and answers for freshers, organized section-wise to help you clear technical rounds with confidence.
CDM is the process of collecting, cleaning, and managing clinical trial data in compliance with regulatory standards. Its main goal is to generate high-quality, error-free, and complete data for statistical analysis.
The lifecycle is divided into three key phases:
✔ Setup Phase: Database design, eCRF creation, and Data Management Plan (DMP) finalization.
✔ Conduct Phase: Data entry, discrepancy (query) management, medical coding, and data review.
✔ Closeout Phase: Data reconciliation, blind review, database freeze, lock, and archiving.
A CRF is a paper document used to record protocol-required data on each trial subject. An eCRF (Electronic CRF) allows site coordinators to enter patient data directly into an Electronic Data Capture (EDC) system.
The DMP is a master document that defines the strategies, procedures, and tools used for handling data throughout the study lifecycle. It serves as a compliance guideline for data managers.
ECS is a document containing the logical programming rules used to detect data entry errors. For example, an edit check will fire a query if the 'Discharge Date' is entered earlier than the 'Admission Date'.
UAT is the process of testing the newly built clinical database and eCRF screens using test patient data before the study goes live. It ensures that edit checks and data entry paths function correctly.
CDISC (Clinical Data Interchange Standards Consortium) is a global standard for structuring clinical trial data. It ensures that data submitted to regulatory authorities (like FDA) is uniform and easy to review.
A discrepancy is an error, inconsistency, or missing value in the clinical database that violates validation rules. Resolving discrepancies is called query management.
✔ System Query: Automatically generated by the EDC database edit checks (e.g., missing mandatory field).
✔ Manual Query: Raised manually by a Data Manager during data review (e.g., illogical clinical comments).
A DCF is a paper document used in paper-based trials to send queries to clinical sites and receive written corrections from the investigator. (In electronic trials, this is handled online via the EDC).
DDE is used in paper studies where two independent operators enter the same paper CRF data into two separate databases. The system compares the records to identify and resolve transcription differences.
A query generated and resolved by the system programmer or coordinator at the site without requiring external intervention from the Data Manager.
SDV is the process of checking the data entered in the eCRF against the patient's original hospital charts (source documents). This is typically performed by Clinical Research Associates (CRAs) or monitors.
Medical coding is the translation of unstructured medical terms (Adverse Events, illnesses, medications) into standardized codes using global medical dictionaries.
MedDRA is used to code Adverse Events. It is structured into 5 hierarchical levels:
1. System Organ Class (SOC)
2. High Level Group Term (HLGT)
3. High Level Term (HLT)
4. Preferred Term (PT)
5. Lowest Level Term (LLT)
WHODrug (WHO Drug Dictionary) is used to code concomitant medications (all medications taken by patients during the clinical trial).
It is the process of comparing Serious Adverse Event (SAE) data in the clinical database against the safety database (managed by Pharmacovigilance) to ensure both datasets are identical.
Key data points matched include: Patient ID, Event Term (Onset/Resolution date), Severity, Causality relation, Seriousness criteria, and Outcome.
Reconciling external data files sent directly by central laboratories with the lab results entered by clinical sites in the eCRF.
DBL is the final process of changing database access permissions to read-only. This locks all clean data permanently, preventing further edits before statistical evaluation.
✔ All data entry completed by sites.
✔ All queries resolved and closed.
✔ Medical coding finalized and approved.
✔ SAE and Lab reconciliation completed.
✔ Investigator signatures obtained on all eCRFs.
A temporary lock where write permissions are revoked for sites but remain active for data managers to run final checks. If errors are found, the database can be unfrozen easily.
Yes, but only under exceptional circumstances. Unlocking requires a formal request, regulatory justification, and senior management/sponsor signatures.
A final review of the database before lock, conducted without knowing which treatment group (active drug or placebo) patients belong to. This prevents evaluation bias.
This regulation defines the FDA rules for using electronic records and electronic signatures in clinical trials, ensuring they are as reliable and secure as paper records.
ALCOA+ defines the guidelines for clinical data integrity:
✔ Attributable, Legible, Contemporaneous, Original, Accurate.
✔ + Complete, Consistent, Enduring, and Available.
An audit trail is a secure, system-generated log that records every change made to database values. It logs: the original value, the updated value, who changed it, and when (timestamp).
✔ Audit: A systematic QA review conducted by the company (or an external firm hired by the sponsor) to check internal compliance.
✔ Inspection: An official evaluation conducted by regulatory agencies (like FDA, EMA) to verify study compliance with legal mandates.
A protocol deviation occurs when the study is conducted differently from the approved protocol document (e.g., patient taking prohibited medication or missed clinic visits).
CAPA stands for Corrective and Preventive Action. It is a structured process to identify deviations, investigate the root causes, fix the issue (Corrective), and implement steps to prevent it from happening again (Preventive).
.png)
