If you are the only person abstracting cases at your facility, you already know the quiet math of the job. You read the pathology, the operative note, the imaging, the clinic visits. You code the primary site, the histology, the stage, the first course of treatment. Then you move the case forward, and no one looks at it again. That last part is the problem. Strong cancer registry quality assurance depends on a second reviewer, and a lone registrar rarely has one. This post makes the case for building that second set of eyes into your workflow, explains why abstraction gets more accurate with review, and shows what dependable cancer abstracting QA actually looks like in practice.
Here is the takeaway up front: abstraction is skilled work that improves measurably with training and a second review. Wanting a peer to check your abstracts is not a sign that you are struggling. It is how good registries are supposed to run.
The Lone Registrar Problem Is Real, and It Is Common
Plenty of cancer programs run on one person. A single abstractor at a community hospital. A manager who abstracts between meetings. A contractor covering a backlog with no colleague in the same file. When the registry is a department of one, several protections that larger teams take for granted simply do not exist.
There is no peer to catch a miscoded histology before it locks in. There is no one to talk through an ambiguous stage. There is no second reader when a rule changes and you are not sure your interpretation matches the intent. And there is very little time to keep up with the annual updates to coding manuals, staging systems, and accreditation standards, because the case list does not pause while you read.
The workload data backs up what the day feels like. A 2024 workload study of registry leads and registrars published in the Journal of Registry Management found a mean caseload of roughly 441 cases per full-time equivalent and recommended somewhere around 1.6 to 2.1 FTEs per 1,000 cases to do the work well. The same study reported that only about 71 percent of registrars felt they had time to complete high-quality abstracts. Read that the other way and roughly 29 percent did not feel they had the time. That is not a story about effort. It is a story about capacity. When one person carries the whole registry, quality assurance is usually the first thing that gets squeezed, because it feels optional next to a growing case list. It is not optional. It is the part that protects everything else.
Abstraction Is Accurate When It Is Reviewed, and That Is a Feature
There is a persistent and unfair assumption that a good registrar should get every field right, every time, alone. The evidence on medical record abstraction tells a more honest story: accuracy is field-dependent, and it climbs with training and review.
A peer-reviewed abstraction study by Gianinazzi and colleagues, published in PLOS ONE in 2015, measured how consistently trained abstractors pulled the same data from the same records. Inter-rater reliability ranged from substantial to excellent, with agreement of about 86 percent to 100 percent and Cohen’s kappa between 0.70 and 0.83. Reliability varied by data field, and, importantly, accuracy against a gold standard improved significantly after training. This was a general medical record abstraction study rather than a tumor registry study, so treat it as illustrative rather than a registry benchmark. The pattern still matters. Some fields are inherently harder and more variable than others. Training moves the needle. A second reader catches what a first reader, working fast and alone, does not.
That is the whole argument for a review step. Not because registrars are careless, but because abstraction is exactly the kind of interpretive, detail-heavy work where a second perspective reliably improves the result. Airlines use checklists. Radiologists get second reads on hard films. Cancer abstracting QA is the same principle applied to registry data.
What the Data Actually Has to Meet
Quality assurance is not an abstract virtue. The data you produce has to clear real standards, and those standards are strict.
The CDC’s National Program of Cancer Registries sets a national data-quality benchmark of 95 percent completeness and expects 99 percent of records to pass a prescribed set of standard computerized edits, along with other criteria. You can read the full set on the CDC NPCR standards page. Those numbers leave almost no room for uncaught error. A registry running without a review layer is betting that a single reader hit those thresholds on the first pass, every quarter, across every field.
Accreditation adds another layer. The Commission on Cancer Standard 4.3 requires that abstracting be performed by an Oncology Data Specialist, and there are more than 6,000 active ODS-certified professionals nationwide, per the NCRA. Credentialing sets the floor for who is allowed to abstract. It does not, on its own, create the peer review structure that keeps a solo abstractor’s work consistent over time. A registrar can be fully certified and still be working without anyone checking the output. Credential and QA are two different protections. You want both.
What Good Cancer Registry Quality Assurance Looks Like
When people hear “audit,” they often picture a punitive, once-a-year event that finds problems too late to fix them. Useful cancer registry QA is the opposite. It is continuous, specific, and built to teach as much as to catch. A few things separate real quality assurance from a box-checking exercise.
A genuine second review of abstracts. Not a spot check of demographics, but a qualified reviewer re-reading the source documentation and confirming the hard fields: primary site, histology, laterality, stage, and first course of treatment. This is the tumor registrar review step that a solo registry structurally cannot provide for itself.
Field-level feedback, not just a pass or fail. Because reliability varies by field, good QA tells you where the disagreements cluster. If staging is where errors concentrate, that is where training and attention should go. A score with no detail teaches nothing.
Attention to the edits and completeness thresholds that matter. QA that keeps the NPCR standards in view, the 95 percent completeness target and the 99 percent edit pass rate, catches the issues that actually threaten your data submission and your accreditation, rather than chasing trivia.
Current survey and accreditation intelligence. Standards change. Manuals get revised. A strong QA partner knows what surveyors are asking about right now and folds that into the review, so a lone registrar is not the last person to learn about a change that affects their coding.
A tone that supports rather than shames. The point of a second set of eyes is to make the work better and to take weight off the person carrying it. Reviewers who treat findings as teaching moments build accuracy that lasts. Reviewers who treat findings as failures just make people defensive and tired.
Documentation you can show a surveyor. A defensible QA process produces a record: what was reviewed, what was found, what changed. That trail is worth a great deal when accreditation time comes.
How Medovent’s ODM Team Becomes Your Second Set of Eyes
Medovent Solutions is a US-based revenue cycle consulting and staffing company, and its Oncology Data Management division exists to provide exactly the layer a lone registry is missing. The brand line, “large enough to scale, small enough to care,” is a fair description of how the work is meant to feel: enough depth to bring a real QA program, small enough that you are working with people who know your registry rather than a ticket number.
The Oncology Data Management team is built around experienced, ODS-credentialed cancer registrars working in domestic, managed models with QA included. That last point matters. QA is not sold as an afterthought. It is part of how the abstraction service is delivered, which means a second reviewer is looking at the work as a matter of routine rather than as an emergency measure after a problem surfaces.
For programs that need a focused, defined engagement rather than ongoing coverage, Medovent’s Specialty Solutions handle QA and audit projects: a structured cancer registry audit, a review of a specific site or time period, or a second read on a backlog before it goes out the door. This is the option when you do not need a full staffing model but you do want qualified eyes on the data before it becomes permanent.
And because staying current is half the battle for a registrar working alone, Medovent also offers continuing education to help registrars keep pace with changing rules and maintain their credentials. Combining review with education addresses both sides of the accuracy problem the PLOS ONE study pointed to: a second reader catches what one person misses, and training raises the baseline over time.
The through-line across all three is the second set of eyes. Whether it arrives as managed abstraction with built-in QA, a standalone audit, or education that keeps your interpretation aligned with current standards, the goal is the same. You stop being the only person responsible for catching every error, and you get survey and accreditation know-how from a team that does this work every day.
Frequently Asked Questions
What is cancer registry quality assurance, exactly?
Cancer registry quality assurance is the set of review activities that confirm abstracted data is accurate, complete, and consistent with current coding and staging rules. In practice it centers on a qualified second reviewer re-reading source documentation and verifying the key fields, backed by attention to the completeness and edit standards the data has to meet. Good QA is continuous and specific, and it gives field-level feedback rather than a single pass-or-fail grade.
We only have one registrar. Isn’t a QA program overkill for a small program?
It is the opposite. A one-person registry is the situation where a second set of eyes matters most, because there is no built-in peer review at all. Small programs face the same NPCR completeness and edit standards and the same accreditation requirements as large ones. A right-sized QA arrangement, whether a periodic cancer registry audit or a managed model with review included, gives a small program the protection a larger team gets automatically.
Does having an ODS-credentialed registrar mean we do not need separate QA?
Credentialing and QA solve different problems. Commission on Cancer Standard 4.3 requires abstracting by an ODS-certified registrar, which sets the standard for who does the work. Quality assurance is about checking the output over time, regardless of how skilled the abstractor is. Even excellent registrars benefit from a second reader, because abstraction accuracy varies by field and improves with review. You want the credential and the QA layer, not one instead of the other.
How does a second review actually improve abstraction accuracy?
Abstraction is interpretive work, and different fields carry different levels of difficulty and disagreement. The Gianinazzi PLOS ONE study found that abstraction reliability varied by field and improved significantly with training measured against a gold standard. A second reviewer catches the specific fields where a first reader, moving quickly and alone, is most likely to diverge from the correct answer, and the feedback from that review makes the next round of abstracting better.
You Should Not Have to Carry the Whole Registry Alone
If you have been holding the quality of an entire cancer registry on your own shoulders, wanting a second reader is not a weakness in your work. It is the structure your work has been missing. Abstraction gets more accurate when someone reviews it, the standards your data must meet leave little margin, and the person doing the abstracting deserves support rather than pressure. Medovent’s ODM team can be that second set of eyes, through managed abstraction with QA built in, a focused audit, or education that keeps you current.
Contact Medovent Solutions