Policy Lab · anchored to California AB 1199 · sources checked

Credentialing modernisation without lowering competency standards

Almost every argument about credentialing delay is really an argument about duplication. The same licence, the same diploma and the same training record are verified from scratch by each hospital and each payer a physician encounters — and the verification is nearly identical each time. Removing duplication is not the same as removing scrutiny, and conflating them is what stalls reform.

Six gates, routinely discussed as one

The most common error in this area, made by clinicians and legislators alike, is treating credentialing as a single process. It is six, with different holders, different purposes and different failure modes. A reform that fixes the wrong one changes nothing.

The credentialing stack. Each row is a separate determination by a separate body.
GateWho decidesWhat it establishesWhere it fails
LicensureState medical boardLegal authority to practise medicine in the state. Without it nothing else is available.Board investigation, discipline, or lapse. The only layer that can end a career outright.
Board certificationPrivate certifying bodyEvidence of specialty training and examination. Not a licence and not legally required to practise.Continuing-certification requirements imposed as a condition by institutions that did not set them.
CredentialingInstitution (verification function)Verification that the licence, training, certification, references and history are genuine. Primary-source verification is the actual work.Duplicated across every institution and payer a clinician touches; the delay is the harm.
Medical-staff membershipHospital medical staff / governing boardMembership in the organised medical staff, with its bylaws, peer review and due-process protections.Membership can be conditioned on things unrelated to competence.
PrivilegingInstitution / departmentPermission to perform specific procedures and care for specific populations at that institution.Privilege decisions are clinical judgements made administratively, and reviewability varies enormously.
Payer enrolmentEach health planBeing eligible to be paid for care delivered. Separate from every layer above.A fully licensed, certified, credentialed, privileged physician who cannot bill is not practically employable.

Where the delay actually is

Primary-source verification is the step everyone points to and the step least worth accelerating: confirming with the issuing body that a degree, licence and training actually exist is cheap, fast when the source responds, and the whole basis of the system’s integrity. What is expensive is doing it again at the next institution, and then again for each payer, from the same primary sources, with no shared record.

That is a structural problem with a structural answer: verify once to a standard, and let subsequent institutions rely on the verification rather than repeat it. Reliance is exactly the mechanism regulators use across borders for drug assessments — the same logic, applied to a clinician’s file.

The layers that genuinely require independent institutional judgement are different: privileging is a clinical decision about what this physician should do at this hospital with these resources, and medical-staff membership carries peer-review and due-process consequences. Neither should be centralised, and neither is what causes the delay.

What must not change — stated first, deliberately

Any proposal in this area has to survive one objection: that streamlining is how unsafe clinicians move between institutions. The history that makes that objection serious is real, and the answer is not reassurance but design.

So: nothing here reduces the substantive verification standard, shortens the history that must be examined, weakens adverse-action reporting or reduces the institution’s obligation to review it. A shared verification record must carry adverse findings with the same fidelity as favourable ones, or it becomes a laundering mechanism — and the National Practitioner Data Bank analysis on this site is the reason that risk deserves naming rather than a footnote.

The questions this Lab entry opens

  • Should primary-source verification be performed once to a published standard and relied upon by subsequent institutions, with re-verification triggered by event rather than by calendar?
  • Who should hold a shared verification record, and what makes that holder trustworthy to both clinicians and hospitals?
  • Should payer enrolment be permitted to duplicate institutional credentialing at all, given that it verifies the same facts for a different purpose?
  • Should a credentialing decision be reviewable, and on what standard — noting that privileging decisions genuinely require clinical judgement that a reviewing body may not hold?
  • Should board certification be permitted as a credentialing condition when the institution has not stated the criteria a credential must meet? This connects directly to parallel certification recognition, which addresses recognition criteria rather than verification procedure.
  • What is the measured harm of credentialing delay — in unfilled posts, delayed patient access, and locum cost — and does anyone collect it?

No KPSGILL position is registered yet. The last question is why: this site has no published record of measured credentialing delay and its consequences, and a recommendation to restructure verification without that measurement would be an opinion wearing a proposal’s clothes. This is opened as a Lab entry on the same basis as coverage architecture.