CRA Applicants:
You were trained to run trials, not to run fraud investigations as you search for CRA applicants. And yet, here you are.
From what my team sees at intake, roughly 60% of CRA applicants have falsified all or part of their credentials, another 20% are AI-generated ghost candidates, and only about 20% are real people with real qualifications. That is a craresources intake observation, not a published statistic, but it is the water you are swimming in. This checklist is built to help you find that real 20% faster, and to spot the other 80% before they reach your team or, worse, your study.
One distinction before you start. This is about fraud, the wholesale falsification of employment history, degrees, or references, not the ordinary polishing every candidate does. You are hunting for fabrication, not for embellishment.
Step 1. Start in the top third of the resume
The top third of a professional’s resume will typically include their professional summary, core competencies, and career highlights. This is where a real CRA proves themselves and where a fake one gives themselves away. Read it first, and read it for proof.
The summary names specifics, not adjectives.
A real senior CRA names therapeutic areas, phases, and the kind of studies they have monitored. A fabricated one leans on words like “detail-oriented,” “results-driven,” and “GCP-compliant” with nothing concrete underneath.
Core competencies are verifiable, not a keyword wall.
A list that reads like it was reverse-engineered from your job posting, matching your every requirement a little too perfectly, is a flag, not a match.
Career highlights show proof, not duties.
“Monitored oncology sites” is a duty. “Managed 12 sites on a Phase III oncology trial, resolved a recurring consent deviation that was risking an FDA finding” is proof. Real experience carries the study, the phase, the therapeutic area, and the specific problem solved. Fabrication stays vague because vagueness is safe.
The language sounds generated – because it is.
This is the newest half of the problem and the one the old playbooks miss entirely. Roughly one in five applicants is not a real person at all, but an AI-generated profile built to pass a keyword scan. When the language is fluent, generic, and strangely weightless, with no specific studies, sites, or problems that a real person would mention, that is a flag that this may be an AI-generated profile.
The principle underneath this step: proof over polish. A beautifully written resume that collapses the moment you ask one specific question is the single most common tell. Polish is easy to manufacture. Proof isn’t.
Step 2. Check the dates before anything else
Dates are the easiest to check and usually the fastest fraud flags you have, but most managers skim right past them.
Every experience entry shows month and year, not just years.
A year-only range (2019 to 2022) is purposeful, intended to hide gaps, overlaps, and invent tenure. It can also help the candidate misrepresent their qualifications. For example, an oncology position listed as 2019 to 2020 could mean that the candidate only has two days’ experience in oncology. And this is a problem if you require two years.
In clinical research, details matter. Therefore, a year-only resume is a professional-standard violation and a flag. My agency will not advance a candidate without month and year, and you shouldn’t either.
The employment dates fit the company’s real dates.
Know when the companies on the resume actually existed. A candidate who lists a company before it was founded is claiming to have worked somewhere that didn’t yet exist. That is fabrication, not a typo.
Acquired, merged, and rebranded companies line up.
When you see a trend where a candidate has listed multiple companies on their resume that did exist but are either deceased, merged with another company, or have otherwise rebranded, dig deeper. These details are where fabrication hides, because the lineage is confusing. Two real examples to show my point:
-
- Beryllium was founded in 2009 and acquired by UCB Pharma in June 2017. A candidate claiming Beryllium before 2009 is fabricating. One claiming it well after the acquisition needs a hard second look.
- Syneos Health has a long lineage (SFBC to PharmaNet in 2004, PharmaNet to inVentiv, inVentiv merged with INC Research in 2017, became Syneos in January 2018, and more). Someone claiming PharmaNet after it stopped being PharmaNet is either careless or fabricating. Dig either way.
The claimed experience matches the math.
Three common mismatches, each a quick check rather than a verdict:
-
- Graduation date versus years’ experience claimed. A summary bullet claiming a certain number of years, set against the graduation date, that would have them starting their career at age twelve. The subtraction does the work for you.
- Therapeutic-area years versus actual history. A claim of X years monitoring a specific therapeutic area when they do not have that much monitoring experience overall, or when the companies they list were not running those trials for that long. The claimed depth outruns the history that could have produced it.
- Two roles, one schedule. Two overlapping perm roles should always be poked at. Additionally, while sometimes a contract and a perm role at the same time can be fine, that is only acceptable if the contract work can be done on a different schedule than a daytime role. The flag for both of these types of overlaps is when both positions require the same working hours.
- The resume, the LinkedIn profile, and any earlier version line up. Real people have one story. Fabrications have seams. Compare the resume to the LinkedIn profile, and to any earlier version of either you have on file. A real history doesn’t move. A fabricated one drifts, because a lie told years ago is hard to remember exactly, so the details shift with each new resume. Sometimes you will even watch a fabrication slowly turn true, as someone who faked their way in accumulates real experience and quietly corrects the record. Either way, movement is the tell.
Step 3. Confirm the companies are real
Do not assume a company is real because it has a website.
Don’t assume that the company is real just because there is a website or a location listed on Google. Fabricated companies can come with a fabricated web presence, so they are hard to spot.
Verify unfamiliar companies.
If you aren’t familiar with a company used on a resume, research it. OpenCorporates is the fastest single source, since it scrapes state registrars and international listings. State registrars and Manta work too.
Watch for a pattern of unverifiable companies.
One deceased company on a resume is normal. But several that cannot be confirmed, stacked together, is a pattern that may point to a falsification of employment.
Step 4. The rest of the resume tells
The digital footprint is thin or wrong.
In this industry, a genuine senior CRA leaves a broad trail: LinkedIn, clinicaltrials.gov, site directories, citations. A missing footprint is a flag. A footprint that shows the person driving a truck or working outside the industry entirely is a bigger one.
A below-market rate request.
Real senior CRAs know what the market pays. A rate well under it usually means desperate, underprepared, or from outside the industry. Not proof of fraud, but a reason to probe hard.
Self-employment or consulting with no clients or projects.
Genuine contract CRAs are willing to discuss their clients, studies, therapeutic areas, as well as start and end dates. A candidate who will not name or discuss any of it is usually hiding a reason: no real clients, or contracts that ended badly.
Step 5. Interview for proof, not polish
The interview is where fabrication finally breaks, but only if you interview for proof. Create an interview template, use behavioral questions, and refuse to accept a surface answer.
The interview feels scripted or off.
Watch for answers that sound read rather than recalled, and for audio or video that lags, glitches, or never quite syncs. Deepfake interviews are real now. Verify identity.
Ask for the whole story: Situation, Action, Result, Benefit.
Have them walk you through a real study. What was the situation, what did they specifically do, what was the result, and what was the benefit to the trial/project/company?
Probe one level deeper than they expect.
“You mentioned resolving a consent deviation. Walk me through exactly what you found, who you escalated to, and what changed in the SOP afterward.” Real experience deepens under pressure. Fabrication runs out of detail.
When the story doesn’t hold, keep probing, or stop.
When the situation, action, result, or benefit doesn’t line up with the competency you are hiring for, that is your signal to press harder or to end the interview. A polished talker who cannot go one level deeper is telling you what you need to know.
One last thought
You are looking for trends, not perfection. There are gray areas here, and genuine CRAs can trip a flag or two. A thin digital footprint alone wouldn’t stop us from progressing a candidate. But a fabricated company on the resume would. Weigh the whole picture.
And know this: doing all of the above, on every applicant, is a real job. It is the job that gets bolted onto yours when you were hired to lead a study, not to screen out fraud at scale. That is the exact work my team does before a single resume reaches you, so the only candidates you ever see are the real 20%, already verified.
When you want that off your plate, or you want to get sharper at doing it yourself, that is what I am here for.
To Your Success,
Angela Roberts, Managing Partner of craresources