DESKGATE, INC. · OPERATIONAL SPECIFICATION
← POSITION DOWNLOAD PDF
WHITE PAPER · RESEARCH SPECIFICATION & METHODOLOGY · SCHEDULE M

The Non-API Residual:
An Empirical Methodology for Measuring Coordinator Throughput Across Payer Portals After CMS-0057-F

Issuer: DeskGate, Inc. (Delaware C-Corporation)
Authors: Hans J. Schulte & Jolanda C. M. Wevers Schulte
Date: September 2026
Reference: DG-WP-2026-09-M
Status: Methodology + specimen derivation — not an observed live desk
Abstract. CMS-0057-F requires impacted health plans to implement FHIR prior-authorization APIs by 1 January 2027. Market consensus treats that mandate as the end of portal work. The statutory text does not. Self-funded employer plans under ERISA, specialty drug prior authorization, clinical-review handoffs, denials, and appeals remain outside the API. This paper defines a measurement protocol for that residue: one named portal family, a frozen eligible cohort, a certified written baseline, and a 14-day window scored as completed cases per coordinator-day. The runtime under test is a client-side dual surface (embed.js on owned pages; a browser extension on third-party payer tabs), multimodal visual grounding, and a mandatory human gate before Submit. Numbers in Section 4 are a specimen derivation that shows how the ruler works. They are not a live Availity observation.

1. The Regulatory Landscape: CMS-0057-F

Utilization management and prior authorization concentrate a large share of U.S. revenue-cycle administrative time. CMS-0057-F requires Medicare Advantage organizations, Medicaid and CHIP agencies, and QHP issuers on Federally Facilitated Exchanges to expose HL7 FHIR prior-authorization APIs by 1 January 2027.

Four carve-outs leave high-friction work on the desk after that date:

Workflow / segment CMS-0057-F scope Desk impact
ERISA self-funded employer plans Exempt. Department of Labor ERISA (29 U.S.C. § 1002), not CMS title authority. A large share of commercially insured workers. Portals and fax remain operating channels.
Specialty drug prior authorization Excluded from the rule’s PA API provisions for pharmacy-benefit drugs. GLP-1s, oncology infusions, and biologics stay manual.
Clinical review handoffs API can carry a structured request. It does not assemble the chart packet. Notes, labs, and attachments remain human-collated when review is not instant.
Denials, rebuttals, status chase Decision-reason obligations do not automate the appeal desk or the IVR. Packet assembly and phone trees persist.
2026–2028 mixed book Staggered payer rollouts. Staff context-switch between FHIR lanes and portal lanes. A mixed book is more expensive than a uniformly manual one.

DeskGate does not contest the API. It measures and stages the residue the statute leaves behind.

2. Governance: Spear of Medicine, Human Gate, Zero-Retention

Clinical AI stalls when the vendor cannot say who is the filing actor. Scope is limited as follows.

SPEAR OF MEDICINE
▪ Tip — diagnosis, prescribing, patient-facing care. Excluded. DeskGate never touches.
▪ Near-tip — medical-necessity determination. Excluded. The EHR and physician already decided.
▪ Mid-shaft — portal field staging, PDF merge, attachment collation. Workstation.
▪ Shaft — session inheritance, DOM micro-actions, IVR chase, REA hashing. Runtime.

Human gate. The runtime stages fields and attachments, then stops. The licensed coordinator inspects and clicks Submit. Legal filing authority does not transfer.

HIPAA / 45 CFR § 164.312. ePHI is processed in workstation RAM for the session and is not written to a DeskGate database. Production screen-reads route only to BAA-eligible providers (Anthropic Claude on a HIPAA-ready organization, or Vertex AI under a Google Cloud BAA). Developer-tier Gemini keys are synthetic-only. Zero-retention describes DeskGate storage, not a provider void: Anthropic Covered Models retain approximately 30 days. Each completed case can emit a SHA-256 hashed Residual Evidence Artifact (REA v1.2).

Two runtimes. embed.js mounts on pages the practice owns. Third-party payer tabs (Availity, CoverMyMeds) require the operator extension. The website overlay cannot see Epic or Availity from deskgate.app.

3. Occupational Mapping and Price-Plus-Burden

Tasks are mapped to USDOL/ETA O*NET codes so the boundary is occupational, not marketing.

O*NET Role Staged by runtime Kept human
43-6013.00 Medical secretaries & admin assistants Form completion, note compilation, field map Phone triage, scheduling, greeting
29-2072.00 Medical records specialists Retrieval, diagnostic/CPT transcription, collation Release authority, training, governance
43-9041.00 Insurance claims & policy processing Portal field entry, uploads, status check Insured communication, complex exceptions
43-3021.00 Billing & posting clerks Statement assembly, error-revision staging Ledger reconciliation, counseling
NEW TASK Review gate (Time-at-Gate) Inspection of staged output before Submit. Product-created drag. Specimen median ~1:52.

Price-plus-burden

Gross “minutes saved” that ignore review time are not a ledger. Time-at-Gate is deducted before net capacity is claimed.

4. Schedule M Protocol and Specimen Ledger

Schedule M is a 14-day measurement on one named portal family against a certified written historical baseline. A completed unit requires human-confirmed Submit and a payer tracking reference captured in the same session.

Label. Tables below are a methodology specimen. They show how rate uplift converts to coordinator-days avoided. They are not an observed 14-day Availity desk. The first signed design-partner cohort replaces this appendix.
Rate uplift is not labor reduction. Moving from 4.88 to 6.40 completed cases per coordinator-day is a +30.6% rate uplift. Coordinator-days for equal volume fall by 1 − (4.88 / 6.40) = 23.8%. Quote labor as coordinator-days avoided per 1,000 completed cases. Do not invent dollar savings.

Specimen derivation

Value Metric Source class
39 requests Average PA requests per physician per week (all cases) AMA physician survey benchmark
13.0 hours Staff and physician time per physician per week on PA AMA physician survey benchmark
20.0 min Implied average staff time per routine request Derived from AMA figures
80.0 min Touch-time parameter for a residual non-API specialty case DeskGate stated baseline — not measured on a named desk
390 min Productive minutes per coordinator-day Operating assumption (6.5 hours)
4.88 cases Baseline completed cases / coordinator-day (390 ÷ 80) Derived specimen baseline
6.40 cases Illustrative staged throughput Illustrative staged cohort — replace with observed

14-day cohort specimen

Layer Dimension Pre-period 14-day specimen Delta
A Outcome Completed cases / coordinator-day 4.88 6.40 +30.6% rate
23.8% days avoided
A Allocation Coordinator-days in window 78.0 equiv. 39.5 allocated 4-coordinator illustration
B Contribution Gate acceptance, zero material edit n/a 70.8% (179) Verification metric — specimen
B Accepted after coordinator edit n/a 22.1% (56) Field adjustment
B Pre-gate return (exceptions) n/a 7.1% (18) Returned to clinic queue
C Diagnostic Median time-to-gate ~80 min touch 0 min 41 sec Specimen staging clock
C Median time-at-gate 0 (new task) 1 min 52 sec Deduct as drag
D Integrity Unauthorized autonomous submits 0 0 (100% gated) Required invariant
D Cohort / eligible / excluded 261 / 253 / 8 Pre-registered exclusions — specimen

5. Design-Partner Reference Desk Charter

First two reference desks run under these covenants so the specimen can be replaced by an observation.

DESKGATE, INC. (Issuer)

By: _____________________________________
Hans Johannes Schulte
Systems & Product Architect
Date: ___________________________________
DESIGN-PARTNER PRACTICE / RCM AGENCY

By: _____________________________________
Authorized operations / executive lead
Title: ___________________________________
Date: ___________________________________

6. References

  1. CMS-0057-F, Interoperability and Prior Authorization Final Rule, 89 FR 8758 (42 CFR Parts 422, 431, 435, 438, 440, 457).
  2. Employee Retirement Income Security Act of 1974, Pub. L. 93-406, 29 U.S.C. § 1002 et seq.
  3. HIPAA Security Rule, 45 CFR § 164.312.
  4. American Medical Association, Prior Authorization Physician Survey (recent iterations). Used here as published benchmarks, not as DeskGate field measurement.
  5. U.S. Department of Labor / ETA, O*NET OnLine, SOC 43-6013.00, 29-2072.00, 43-9041.00, 43-3021.00 (CC BY 4.0).

DeskGate, Inc. · https://deskgate.app · hans@deskgate.app · +1 (302) 244-7953
Capital paper: /pitch · Buyer measurement: /access · This document: /whitepaper