Executive finding
SIRA’s public report displays 455 metric rows across seven periods. WIA captured every displayed cell and independently reconciled every public numeric result to the underlying Power BI model. The model also exposed values behind some displayed suppression symbols. WIA therefore used the visible report as the binding publication mask: exact model precision was used only where SIRA displayed a number.
Scope
The current report contains five quarters from 2025-Q1 to 2026-Q1 and two rolling annual periods ending March 2025 and March 2026. WIA publishes the Nominal Insurer, specialised insurer and self-insurer groups. System and Treasury Managed Fund records are excluded from WIA’s charts and integrated JSON/CSV. Their displayed source cells remain only in the downloadable XLSX audit sheet, clearly marked as outside the WIA comparison scope.
Every visible cell accounted for.
The matrix reconciles the complete visual report. Numeric values are limited to results SIRA displayed. Symbols and blanks are retained as source text and are never converted into estimates or zeros.
| Measure | Cells | Numeric | * | ^ | - | Blank |
|---|---|---|---|---|---|---|
| Injuries notified | 637 | 604 | 21 | 0 | 0 | 12 |
| Claim count | 637 | 550 | 67 | 0 | 8 | 12 |
| RTW | 637 | 546 | 0 | 75 | 0 | 16 |
| SAW | 637 | 550 | 0 | 75 | 0 | 12 |
| WR | 637 | 550 | 0 | 75 | 0 | 12 |
| Total | 3,185 | 2,800 | 88 | 225 | 8 | 64 |
Marker meanings
| Marker | Treatment |
|---|---|
| * | Fewer than 10 claims counted in the period; the count is suppressed. |
| ^ | Insufficient data or fewer than 10 claims in the relevant metric cohort; the rate is suppressed. |
| - | Zero claims or claims excluded in that period. |
| Blank | No current SIRA definition is published. WIA retains the blank and does not treat it as zero. |
Source precision check
All 2,800 displayed numeric cells reproduce from the model: 604 injury counts, 550 claim counts, 546 RTW rates, 550 SAW rates and 550 WR rates. Rate checks used RTW = returned-to-work count / time-lost count, SAW = stayed-at-work count / all claims, and WR = working claims / all claims. No ratio mismatch exceeded 1e-12.
What was technically extractable.
The public Power BI semantic model included exact source fields used to calculate the displayed results. Those fields were technically retrievable even where the visual replaced a result with a suppression symbol. This is a suppression-control finding, not permission to republish the hidden values.
| Model field | Meaning |
|---|---|
notifications | Injuries notified |
allClaims | Claim count and SAW/WR denominator |
returnToWork | RTW numerator |
timeLost | RTW denominator |
stayAtWork | SAW numerator |
workingClaims | WR numerator |
rtw / saw / wr | Calculated recovery rates |
insurer and span fields | Insurer identity, group and reporting period |
Values present behind displayed masks
| Display | Cells | Raw numeric present | Public treatment |
|---|---|---|---|
| * | 88 | 88 | Exact counts present internally; all remain * publicly. |
| ^ | 225 | 188 | Some exact rates present internally; all remain ^ publicly. |
| - | 8 | 8 | Underlying zero or excluded values present; all remain - publicly. |
| Blank | 64 | 0 | No numeric value extracted; all remain blank publicly. |
| Total | 385 | 284 | No hidden numeric value is published. |
Controls applied
| Control | Implementation |
|---|---|
| Visible-report mask | 455 visual rows captured; every one of 3,185 cells retained. |
| Exact-value join | 2,800 numeric visual cells matched to the exact model value. |
| Suppression override | All 88 *, 225 ^, 8 - and 64 blanks override raw content. |
| Exclusion scope | System and TMF excluded from WIA charts and integrated JSON/CSV; displayed source cells are retained only in the XLSX audit sheet and marked out of scope. |
| Private artifacts | Raw response and query payload kept outside public site assets. |
| Regression checks | Counts, keys, periods, entity rosters and output hashes fail closed on change. |
Current results, retained history.
The 19 August 2026 source is a refreshed dataset, not merely a new reporting period. WIA uses the new vintage for repeated periods and keeps the older archive so former and reclassified insurers remain available in the periods where SIRA published them. That includes DXC and Catholic Church Insurance in historical views.
| Public group | Latest roster | Historical treatment |
|---|---|---|
| Nominal Insurer | 7 current | DXC remains historical |
| Specialised insurers | 5 current | Catholic Church Insurance remains historical |
| Self-insurers | 71 current | 40 higher-volume and 31 lower-volume source cohorts |
| Total public entities | 83 current | Former and reclassified entities retained by period |
Provenance record
| Item | Record |
|---|---|
| SIRA page updated | 19 August 2026 |
| Power BI model refresh | 19 August 2026 04:57:05.237 (timezone not declared) |
| WIA retrieval | 20 August 2026 |
| Visible extraction | 455 rows; SHA-256 090f9a362ba224068d6ab1a8007ac274d1b0fa9e4c53b673eb518804dae45465 |
| Internal exact response | 625 records; SHA-256 c68f6df2eef5deabe53789993677ef673fa52e1758fdeb2082152b7c422bc8b6 |
| Public display overlay | 455 rows; SHA-256 48ec76ab4e07cb8045fb6ec40d9a5a86ae4b3c2d1bf78940f8168b07f8531c96 |
Interpretation limits
Recovery-through-work results describe one 13-week outcome measure. Portfolio size, industry and injury mix, suitable-work opportunities, claims practices and external conditions differ. Suppressed or unavailable results are excluded from rankings rather than treated as zero. Historical values can change when SIRA republishes revised data.