What a CRA Actually Does

Part of Clinical Research Associate (CRA) Training: Foundations for the Monitor.

About 6 minutes. Free, in your language, no account needed.

Why this matters

The job title says "associate" and the job is oversight. New CRAs frequently misunderstand the role in one of two directions: they behave like an auditor, or they behave like an extra pair of hands for the site. Neither works.

This lesson sets out what the role is, what it is not, and what the week actually looks like.

The three objectives

Everything a CRA does serves one of three goals.

Participant safety. That the rights, safety and wellbeing of participants are protected. That consent was obtained properly. That adverse events are being identified, documented and reported.

Data integrity. That the data reported to the sponsor are accurate, complete and verifiable from source records. That the site enters data promptly and corrects it properly.

Compliance. That the trial is being conducted according to the protocol, GCP and applicable regulation, and that problems are detected early enough to fix.

When you are unsure whether something is your business, ask which of the three it touches. If it touches none of them, it probably is not.

What monitoring is, in the guideline's terms

ICH E6(R3) section 3.11.4 describes monitoring as one of the principal quality control activities, and requires it to be performed by persons not involved in the clinical conduct of the trial at the site being monitored. That independence is what makes your findings worth anything.

The guideline distinguishes two kinds.

• Site monitoring covers activities at the investigator site, including its pharmacy and local laboratories. It may be performed on site, remotely, or both, and may include secure read-only remote access to source records.
• Centralized monitoring is the sponsor's evaluation of accumulated data by qualified staff such as a medical monitor, data scientist or biostatistician. It can complement site monitoring, reduce its extent or frequency, or be used on its own.

One hundred percent source data verification is not required. Section 3.11.4.5.4(b) allows checking reported data against source records using samples, supported by data analytics as appropriate, and section 3.11.4.3 says the monitoring plan should focus on aspects critical to quality. Module 8 covers what this means in practice.

The monitoring cycle

Before a visit. Review the site's data in the electronic system; look at enrollment, missing data, open queries and safety events. Review the previous visit report and its open actions. Read the monitoring plan for what this visit requires. Agree the agenda and the documents needed with the site.

During a visit. Open the visit with the coordinator and, where possible, the investigator. Verify consent and eligibility. Compare reported data with source records for the sampled participants. Review investigational product storage and accountability. Review the site file. Check safety reporting timeliness. Discuss findings as you go. Close with a formal meeting that agrees actions, owners and dates.

After a visit. Write the monitoring visit report. Send the follow-up letter. Track the actions to completion. Escalate anything that needs a decision above you.

Modules 7 and 8 take each of these apart in detail.

What a CRA is not

You are not the site's staff. You do not enter their data, complete their logs, or write their notes. Doing so destroys your independence and creates a record nobody can defend.

You are not an auditor. An audit is an independent test of the system, performed by someone outside the trial team. Monitoring is routine quality control of this trial. Different purpose, different person.

You are not the site's manager. You have no authority to direct their staff. Your influence comes from clarity, consistency and being right.

You are not the medical monitor. Medical questions — is this event related, may this participant continue — go to the investigator and, where needed, to the sponsor's medical monitor.

The skills that actually decide performance

• Reading closely. Most findings are found by comparing two documents carefully.
• Asking open questions. "Walk me through how you consent someone" reveals more than "do you follow the consent SOP?"
• Writing precisely. A finding that a site cannot act on is not a finding.
• Judging severity. Distinguishing a filing inconsistency from an unreported serious adverse event, and treating them differently.
• Managing a relationship without losing independence. Sites work harder for monitors they trust and hide things from monitors they fear.
• Organization. Several sites, several sets of open actions, and travel.

The difficulties nobody warns you about

• Delayed data entry, so you arrive to review data that does not exist yet.
• Missing or incomplete source records.
• Under-reported adverse events.
• High staff turnover at sites, so you retrain the same site repeatedly.
• Investigators who are not engaged with the trial.
• The tension between the sponsor's timelines and what a site can actually do.

None of these is solved by being tougher. Most are solved by finding the cause and escalating early.

In practice

At one VT-301 site, Leah finds that data entry is consistently three weeks late. The contract says five working days.

Her first instinct is to write a finding about late data entry and move on. Instead she asks the coordinator to walk her through her week.

The site has one coordinator running three trials. VT-301 visits are booked back to back on Thursdays, with no time afterwards for entry. Nothing is going to change by asking her to try harder.

Leah documents the cause, not just the symptom, and escalates it to her project manager. The sponsor agrees to fund a part-time data entry assistant at that site, and the site changes its Thursday scheduling to leave the last hour free.

Entry comes within contract within a month. A finding that said only "data entry is late" would still be open a year later.

Key takeaways

• A CRA's work serves three objectives: participant safety, data integrity, and compliance.
• ICH E6(R3) requires monitoring to be performed by persons not involved in the clinical conduct of the trial at that site, and describes both site monitoring and centralized monitoring.
• One hundred percent source data verification is not required; sampling supported by data analytics is expressly allowed, and the monitoring plan should focus on what is critical to quality.
• A CRA is not site staff, not an auditor, not the site's manager and not the medical monitor, and blurring any of those roles damages the independence the role depends on.
• Findings that name a cause get fixed; findings that name only a symptom recur, which is why asking a site to walk you through their process is the highest-value monitoring skill.