# ICHRA Report original reporting protocol

Version 1.0 · September 12, 2026 · Publisher: ICHRA Report

Status: prepared; zero interviews conducted for these series. No outreach has been sent. Recruitment and document intake are not open. This is a reporting framework, not findings or a representative survey protocol.

## Shared preparation

Define the reporting question, intended population and selection method before recruitment. Seek employers that adopted, rejected or discontinued ICHRA, and retain unsuccessful implementations in the reporting universe. Record how each participant was introduced and relevant commercial relationships. Participation does not buy coverage, rankings or control over conclusions.

Before any interview, explain the proposed publication, attribution, recording if any, handling of sensitive material and boundaries of consent. Confirm the agreed terms in writing. Do not promise anonymity that the remaining details could undermine. Collect redacted or aggregate evidence rather than employee health information. Establish an appropriate private intake method and a retention/deletion process before accepting documents.

## Series 1: The employer decision

1. What problem initiated the benefits review, and on what date?
2. What were the eligible and enrolled populations, geography, coverage period and current benefits arrangement?
3. Which alternatives were considered, including retaining or modifying group coverage?
4. Which actual proposals and employee-impact analyses were available at the decision date?
5. What contribution, participation and service assumptions drove the comparison?
6. What did the employer choose or reject, and what evidence most influenced it?
7. Which employee groups faced potential tradeoffs, and how were they addressed?
8. Which projected outcomes can now be compared with actual records?
9. What went differently from the plan, and what would change the decision today?

Evidence request: redacted renewal/proposal, dated decision memo, contribution assumptions, aggregate headcount definitions and comparable spending records. Never turn projected savings into realized savings. Document premiums, reimbursement spending, administrator fees and internal costs separately.

## Series 2: The first 90 days

1. What was the intended effective date, and who owned each milestone?
2. When were design, census, communications and required notices completed?
3. How was plan selection distinguished from payment completion and active coverage?
4. How many people were expected to enroll, how many had confirmed starts, and on what date?
5. Which exceptions occurred, how were they discovered, and who resolved them?
6. What support was available to employees and what unresolved needs remained?
7. How were payroll, premium payments and reimbursements reconciled?
8. What changed after launch, and what was the operational cost of the correction?
9. Which claims can the employer, administrator or insurer substantiate?

Evidence request: redacted project timeline, aggregate status/exception reports, dated issue-resolution records and a scope-of-services document. Preserve incomplete cases at the reporting cutoff. Avoid selecting only successfully resolved cases.

## Series 3: Questions for administrators

1. What are the fee schedule, billing population, minimums and additional charges for a defined employer scenario?
2. Which services are included, optional or the employer's responsibility?
3. Who handles initial premium payments, ongoing payments, reconciliation and failed transactions?
4. How is active coverage confirmed, and how are exceptions escalated?
5. Which integrations are operational for the requested workflow, and what remains manual?
6. What licensed assistance and support hours are available to employees?
7. How are implementation time, response time, retention and enrollment metrics defined?
8. What population, period and denominator support each performance claim?
9. What contract terms apply to renewal, cancellation, export and transitions?
10. Which answers can be published, and what independently checkable evidence supports them?

Give every provider the same core questions. Track sent date, deadline, response status and follow-ups only after outreach is authorized. Nonresponse is a reporting state, not evidence of poor service. Provider-confirmed claims remain separate from independently corroborated findings.

## Evidence record template

| Field | Required entry |
| --- | --- |
| Case ID | Non-identifying internal identifier |
| Selection | Why this participant and how introduced |
| Dates | Decision, implementation, measurement and review dates |
| Population | Eligible/enrolled definitions, denominators and missing records |
| Claim | Exact proposition being evaluated |
| Attribution | Participant, company, document or editorial calculation |
| Evidence | Record type, date, locator and permitted use |
| Status | Supported, participant-reported, inferred, unresolved or contradicted |
| Calculation | Inputs, units, formula and comparison baseline |
| Limitations | Selection bias, missingness and alternative explanations |
| Response | Factual response opportunity for materially criticized parties |
| Consent | Attribution, redaction and publication terms |
| Revision trigger | Evidence that would change the conclusion |

## Publication gate

Confirm the central claim against the records and document disagreements. Check arithmetic and consistent populations/periods. Obtain appropriate consent and remove unnecessary identifying information. Clearly label projections, individual experiences and company claims. Give parties a focused opportunity to respond to material criticism without editorial veto. Publish methodology, limitations, source dates and a correction path.

One case cannot establish national averages or causation. Do not publish a benchmark until a separate sampling, consent, measurement and disclosure protocol can support it. Do not use fabricated quotations, respondents, outcomes or anonymous composites presented as real people.

## Research sequence

Prepared now: questions, evidence template and publication gate. Required before collection: working contact channel, intake and handling process, participant selection and explicit outreach authorization. Required before findings: actual interviews, source records, reconciliation and editorial review.