COOKED
The modal worker abstracts data from charts into registries (cancer, trauma, EHR quality databases), audits coding accuracy, runs reports, and maintains system documentation — text-in, structured-data-out work that NLP over clinical notes already does at usable quality. Certifications (RHIT/RHIA, CTR) are employer-preferred credentials, not licensure with personal liability, so there is no legal requirement that a human touch the record. What survives is the accountability layer: defending registry data quality to CoC/state accreditation surveyors, resolving ambiguous or contradictory documentation, and owning HIPAA release and system-configuration decisions.
Roughly flat across the period, with year-to-year wobble.
This line is counted by the Bureau of Labor Statistics — the one figure on this page that isn't a judgement of ours. Headcount moves on demand, offshoring, demographics and the business cycle, and automation is one term among several, often not the loudest.
So a falling line is not evidence that AI did it, and a rising one is not evidence that it won't. Both happen in this register: some occupations resist automation and shrink anyway, others are highly automatable and keep growing.
BLS projection, 2024–2034
+14.7% 41,900 → 48,100 on the projections basis
Exposed, but growing
AI can already do a lot of these tasks, and the BLS still expects +14.7% more of these jobs by 2034. Demand for the output is growing faster than the work is being automated away — the mechanism BLS gives for software developers, and the combination people most often misread as an error.
Different clocks. The score is what current AI could do to this work today. The projection is how many of these jobs will exist in 2034. Everything between the two — how fast employers actually adopt, whether demand grows in the meantime — is why they can point opposite ways without either being wrong.
~3,200 openings a year on average, including replacing people who leave.
Tumor RegistrarCancer RegistrarClinical AnalystApplication AnalystHealth Data AnalystMedical Data AnalystMedical Records ClerkCancer Tumor RegistrarCompliance CoordinatorHealthcare Data AnalystMedical Records AnalystPublic Health RegistrarClinical Data SpecialistMedical Records DirectorData Integrity SpecialistMedical Record ConsultantCertified Cancer RegistrarDigital Health TechnologistHealth Information SpecialistHealth Information TechnicianMedical Information SpecialistUtilization Review CoordinatorCertified Tumor Registrar (CTR)Severity of Illness Coordinator
Holding it up: judgment & accountability . Weakest point: embodiment .
Core tasks are already automatable Abstracting stage, histology, and treatment dates from pathology reports into a NAACCR-formatted registry record, running quality reports, and reconciling coding to the chart are exactly the extraction tasks that clinical NLP handles end-to-end — a 6 rather than lower only because contradictory or missing documentation still forces a human to go chase the physician for a clarification.
Fully desk- and screen-based The job is a workstation, a tumor registry application, and the EHR; the 2 rather than 0 reflects walking a floor for paper charts, scanning legacy records, or sitting in on a tumor board rather than any physical task that resists a remote worker.
Certification preferred, not legally required RHIT, RHIA, and CTR are hiring filters and CoC staffing requirements, not state licenses — no one loses a personal credential-to-practice when a registry field is wrong, and the covered entity absorbs the HIPAA penalty, which puts this at certification-preferred rather than the 11+ of a licensed clinician signing an order.
Meaningful discretion Assigning stage when the op note and path report conflict, deciding whether a records request meets the minimum-necessary standard, and calling reportability edge cases are genuine discretion — but you work inside NAACCR/AJCC/FORDS coding manuals and HIPAA rules that specify the answer for most cases, which caps this at the low end of the discretion band rather than the 14+ of someone making an unscripted call.
The verdict above describes this occupation as a whole. Almost nobody does the typical version of a job — tick what's actually in your week and see how your own mix sits.
Your task mix speaks to task resistance (6/20 here) — how much of the day's work current AI already does. That is the dimension the boxes above are about.
It cannot move the other three. Liability shield (6/20) is whether the law requires a licensed human to sign. Trust premium (4/20) is whether buyers specifically pay for a person. Judgment and accountability (7/20) is whether the role exists to own consequential calls. Those are facts about the occupation's standing, not about which tasks are in your week — a paralegal who does only trial exhibits still holds no licence. Together they are 17 of this occupation's 25 points (68%).
Embodiment (2/20) is also a property of the work rather than the worker, but we don't tag individual tasks as physical or not, so the picker can't tell you anything about it. That's a limit of this tool, not a claim.
Did we get the list right? Tell us what's missing — the tasks are written from the outside, and you're reading this from the inside.
No occupation passed every test: close enough to health information technologists and medical registrars on skills and subject matter, at least 10 points more resistant, no big jump in training, no new licence, no pay cut, and not shrinking on its own. That happens for 223 of the 654 occupations here that aren't SAFE, and it is worth stating plainly rather than leaving the section off.
The usual reason is that exposure travels with the skill profile. The jobs most similar to yours tend to be exposed for the same reasons yours is, so the near neighbours don't clear the gap — and the ones that do are a different kind of work, not a transfer of what you already know. Read that as a limit of this method, not a verdict that you're stuck: it only compares whole occupations, and it cannot see specialisation, industry, or anything you'd bring that isn't in a federal skill survey.
The moves above are yours to make. This is the other half: what would have to change in the world for the occupation itself to score higher. None of it is in any one person's gift, but it is where the floor actually comes from. Scores here are not a one-way ratchet. Only two of the five dimensions — task resistance and embodiment — track what machines can do. The other three track law, what buyers will pay for, and who is answerable, and those move in both directions, often in response to the same pressure AI creates. If every lever below landed, this occupation would score around 45/100 — EXPOSED.
ONC HTI-1 predictive-decision-support transparency obligations, plus state AI-in-utilization-review laws, being operationalized inside hospitals as a standing duty to validate and locally re-test any NLP abstraction/coding tool — with the HIM department named as the validating owner in the compliance plan. That converts 'runs reports' into 'certifies the model's output against sampled charts and documents the drift'.
Task-mix shift is genuinely available here: the occupation has a routine tier (chart-to-field abstraction, report generation) and a judgment tier (reconciling contradictory pathology vs. clinical documentation, casefinding edge cases, defending data quality to CoC/state surveyors, resolving coding disputes with physicians). If the routine tier is fully automated and headcount collapses, the residual job is mostly the judgment tier and per-worker resistance rises — note this is a smaller-occupation effect, not a safer one.
Commission on Cancer / NCDB accreditation standards (or state central cancer registry statutes and their reporting regs) amended to require that each abstracted case be attested by a credentialed CTR — i.e., AI-generated abstracts are non-reportable unless a named certified human signs. CoC already requires CTR staffing ratios and reviews casefinding audits, so tightening from 'staffed by' to 'signed by' is a small step on an existing hook.
CMS quality-program rules (eCQM/MIPS, Promoting Interoperability) requiring a named individual — in practice the HIM/registry lead — to attest personally to the accuracy of machine-extracted measure numerators, with False Claims Act exposure for knowingly submitting unvalidated AI output. Analogous attestation-with-liability already exists for PI attestations.
HIPAA Privacy Rule enforcement trends (reproductive-health privacy attestations, right-of-access initiative penalties) pushing organizations to designate a specific HIM/registrar role as the accountable decider on ambiguous release-of-information and re-disclosure requests, rather than routing them to automated portals.
The limit. No plausible route on trust_premium: patients and payers never see the registrar and no buyer selects a hospital for human-abstracted data. Embodiment is fixed near zero. Even with every lever above, this is an attestation-and-audit role whose headcount tracks the volume of human review regulators demand — the score can rise while the job count still falls.
| New York-Newark-Jersey City, NY-NJ | 2,340 | $77,860 +14% |
| Los Angeles-Long Beach-Anaheim, CA | 1,200 | $87,530 +29% |
| Charlotte-Concord-Gastonia, NC-SC | 990 | $42,480 -38% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 990 | $60,080 -12% |
| Atlanta-Sandy Springs-Roswell, GA | 880 | $79,800 +17% |
| Dallas-Fort Worth-Arlington, TX | 820 | $78,960 +16% |
| Baltimore-Columbia-Towson, MD | 740 | $66,750 -2% |
| Columbus, OH | 730 | $83,580 +23% |
| San Jose-Sunnyvale-Santa Clara, CA | 260 | $126,670 +86% |
| San Francisco-Oakland-Fremont, CA | 280 | $111,480 +64% |
| Sacramento-Roseville-Folsom, CA | 190 | $101,500 +49% |
We have no reported case of a named organisation automating this occupation. Not one deployment, not one announcement.
That is worth saying out loud next to a score of 25. The verdict above is about what the work exposes — what current AI could do to these tasks. It is not a claim that anyone has done it. For this occupation those two things have come apart completely: the capability argument is on this page, and the evidence column is empty.
Has AI actually changed your work? One tap, anonymous, and the running tally is public. Nothing else is asked of you.
Rather than check back: get the digest and we'll tell you what changed — or watch a single occupation from its own page.