The practical answer
Compare each prepared Part IV row with its approved twelve-month coverage pattern. Verify partial boundary months, the all-year indicator and the identity of the person receiving those marks. Matching an annual total is insufficient when different months are checked.
This guide is for coverage-provider reviewers checking generated Forms 1095-B. It begins with enrollment facts already confirmed by the source team and tests whether the form preserves them. It uses the final 2025 instructions and an original output-comparison worksheet. Source teams should resolve disputed coverage dates before the reviewer approves the resulting month boxes.
Fix the person, provider and output version
Identify the provider, reporting year, statement reference and covered-person row. The responsible individual's Part I name alone does not identify which Part IV row needs review. A statement can contain several people with different coverage periods.
Record the approved enrollment-data version and generated form version together. If the source owner has supplied a later retroactive update, establish whether the draft reflects the earlier or later decision. Comparing mismatched versions can create an apparent output defect that is actually an unfinished handoff.
Use internal person references in the comparison sheet and keep the complete identifying records in controlled storage. Preserve continuation-sheet associations so a person on an additional page receives the same month review as a person on the main form.
Apply the correct meaning of a month box
The 2025 Part IV instructions use coverage for at least one day during a month. Column (d), the all-year indicator, applies when that condition is met in every month. Otherwise, column (e) identifies the applicable individual months.
A partial first or last month can therefore be marked. The reviewer should not remove a month merely because the source period begins after its first day or ends before its last day. Retain exact dates because monthly reporting does not establish daily coverage throughout every marked month.
Use the confirmed coverage period, including any reviewed retroactive change. A payment date, event-entry timestamp or letter date is not automatically the coverage boundary. If its meaning is unresolved, return a specific source question rather than choosing boxes to make the row look complete.
Compare patterns before comparing totals
| Check | Compare | Typical finding |
|---|---|---|
| Person association | Approved member reference and generated Part IV row. | One child's pattern copied to another person. |
| Boundary months | Confirmed first and last covered dates against month marks. | A partial starting month omitted. |
| Interior pattern | All twelve reviewed month decisions against output. | A gap filled by an all-year default. |
| Compression | All-year flag against twelve confirmed monthly conditions. | Eleven months represented as twelve. |
| Version | Approved monthly data against the current generated draft. | Old source pattern retained after regeneration. |
Count changed cells only after locating the differences. A total can help reconcile the review, but it should not replace a month-by-month comparison.
Worked example: six correct months, two omitted boundaries
Fictional example: Alder Cove Coverage approves a 2025 period for person AC-18 from March 19 through August 7. The source owner confirms that these dates mean actual coverage began and ended on those days.
The expected pattern marks March, April, May, June, July and August: one starting month + four intervening months + one ending month = six marked months. A draft instead marks April through July because its transformation incorrectly keeps only complete calendar months.
The reviewer identifies two omissions, March and August, and references the approved source interval. Preparation corrects the transformation for the affected records and generates a new version. The reviewer confirms all six expected marks and verifies that January, February and September through December remain unmarked for this source.
The review does not convert six monthly marks into an assertion of coverage before March 19 or after August 7. It verifies the form's monthly meaning against a known interval.
Look for equal-count changes and shared mapping defects
A different defect can preserve the number of marks. Suppose another fictional row should show August through November but instead shows September through December. Both patterns contain four months. The output still needs two cell changes: August added and December removed.
Group findings by the transformation or source path that caused them. A full-month-only calculation may affect every person with a partial boundary. A shifted month index may affect all rows in an imported file. Identify that affected scope before declaring the problem resolved from one repaired example.
Compare revised and unchanged people after regeneration. Preserve the old draft and difference report. A systematic repair should have a predictable effect, including fields and rows that should remain unchanged.
Record the release decision and later correction referral
Save expected pattern, prepared pattern, affected cells, source version and actual reviewer with the final output. The complete field review checks the remaining parts of the statement before release.
If the problem concerns an already filed or furnished version, preserve that history and route it through provider correction intake. The instructions distinguish those states and require the applicable correction procedure. Updating the working comparison does not itself complete IRS or furnishing follow-up.
Keep unresolved source periods visible. A reviewer can approve an accurately explained pattern only when its supporting facts are established; an unchecked box must not silently stand in for missing data. The final comparison should make the decision reproducible without access to the preparer's memory.
From approved period to reviewed month boxes
Read the workflow as text
- Approved AC-18 period. March 19 through August 7, 2025, with verified boundary meanings.
- Expected monthly pattern. March-August gives six marked months, including two partial boundaries.
- Draft comparison. April-July alone omits March and August.
- Revised output review. Confirm six marks, unchanged other months and the correct person row.
Put this guide to work
1095-B generated month-box comparison
Save the editable text worksheet and use it with your own records. Keep completed copies in your secure working files.
Download the worksheet TXTCommon questions
Should partial boundary months be removed from a draft?
No. Apply the at-least-one-day rule to confirmed coverage dates. A month can correctly be marked even when coverage lasted only part of it.
Can an all-year indicator represent eleven covered months?
No. It represents the applicable coverage condition in every one of the twelve months. Preserve the actual monthly pattern when one month does not qualify.
Why compare actual months when the totals already match?
Equal totals can conceal a shifted pattern. August-November and September-December both have four marks but differ in two cells.
Who should resolve an unclear termination date?
The source owner should establish whether the date means the last covered day, the first uncovered day or something else. The output reviewer should not infer that meaning from the field label alone.
Does fixing the workbook complete a correction?
No. Determine the prior filing and furnishing state, preserve the original output and assign the applicable follow-up. The comparison sheet documents a finding and its approved facts.
Official sources and scope
Sources checked September 5, 2026. Use the edition for the tax year and filing method you are working with; later instructions may change thresholds, fields, or procedures.
- IRS: 2025 Instructions for Forms 1094-B and 1095-B
Part IV monthly indicators, all-year treatment and corrections.
