Blog · Board and management reporting
Compare workers compensation claim counts, paid costs, and lost workdays across employer locations. Use a consistent accident cohort, valuation date, and exposure denominator to separate frequency from cost severity.
Workers comp analytics compares claim counts, paid costs, and lost workdays across an employer's locations using the same incident period and cost valuation date. A useful comparison separates how often claims occur from how much has been paid on each claim. It also adjusts for exposure, because a larger location can have more claims without having a higher claim frequency.
The management question is where recorded costs and work disruption differ, and which differences deserve investigation. This is an aggregate employer review. It does not make clinical judgments, decide claim eligibility, or predict an individual employee's risk.
Choose the accident cohort before requesting totals. For example, include accidents from January 1 through March 31, 2026, valued as of September 30, 2026. That means costs paid through the valuation date for that accident cohort, rather than all payments made during the first quarter.
Those are different views. A cash report would group payments by payment date and could include older claims. Keep it in a separate worksheet rather than using it as the numerator for the quarter's new claim count.
Ask the carrier or third-party administrator for a loss run with a stable claim identifier, accident date, location, status, and separate cost fields. Confirm whether paid costs include claim expenses, recoveries, or only benefit payments. Keep the export's field definitions with the workbook, using versioned metric definitions to document changes between reporting periods.
Use the identifier to count claims once. A transaction export may contain many payments against the same claim; a series of monthly snapshots may repeat the same cumulative balance. Neither is a count of new claims.
Bring workforce hours for the accident period and the same locations. Separately obtain aggregate lost scheduled workdays attributable to those claims through the valuation date. Define how partial days are treated and use the same approach at every location. This operational absence measure should not be presented as an OSHA recordkeeping measure.
For the wider labor-cost context, analyzing headcount costs is a separate exercise. Wages and overtime used to cover absences are not automatically part of the loss run's paid claim cost.
A US distribution company has the following location totals after deduplicating its claim export. These are illustrative inputs for the shared accident cohort and valuation date. Hours worked cover the accident period; lost workdays cover the cohort through valuation.
| Location | Claims | Paid claim cost (USD) | Lost scheduled workdays | Hours worked |
|---|---|---|---|---|
| North | 4 | 24,000 | 20 | 40,000 |
| South | 3 | 30,000 | 36 | 30,000 |
| West | 2 | 6,000 | 4 | 20,000 |
| Total | 9 | 60,000 | 60 | 90,000 |
Use three measures together:
Claim frequency per 10,000 hours = Claims / Hours worked × 10,000
Paid cost per claim = Paid claim cost / Claim count
Lost scheduled workdays per claim = Lost scheduled workdays / Claim count
The frequency scale here is a custom internal measure, not an OSHA incidence rate. Keep the scale in the column title whenever the result is shared.
| Location | Claims per 10,000 hours | Paid cost per claim (USD) | Lost workdays per claim |
|---|---|---|---|
| North | 1.00 | 6,000.00 | 5.00 |
| South | 1.00 | 10,000.00 | 12.00 |
| West | 1.00 | 3,000.00 | 2.00 |
| Company | 1.00 | 6,666.67 | 6.67 |
North's frequency is 4 / 40,000 × 10,000 = 1.00. South has fewer claims, but 3 / 30,000 × 10,000 also equals 1.00. South's larger paid cost is therefore a cost-per-claim difference within this example, rather than a higher observed claim frequency.
Company paid cost per claim is 60,000 / 9 = $6,666.67; lost days per claim is 60 / 9 = 6.67. Round displayed rates and averages to two decimal places and retain unrounded calculations underneath. Do not average the location averages to obtain company results.
Claims sum to 4 + 3 + 2 = 9, paid costs to $60,000, lost days to 60, and exposure to 90,000 hours. The location summaries must match the filtered source totals for the same cohort. Paid amounts should match the loss run; hours should match the approved workforce report; lost days should match the agreed absence extract.
For a Google Sheets detail export with location in column A and paid amount in column B, a location total can use:
=SUMIFS(B2:B1000,A2:A1000,H2)
H2 holds the location being summarized. Apply the cohort and valuation filters consistently before aggregating. The SUMIFS documentation describes conditional summation; it does not deduplicate claims or decide which snapshot is valid. Covirage is not affiliated with Google.
Risk management owns the claim definitions and discusses loss-run questions with the carrier or administrator. Payroll or HR owns workforce exposure and absence reconciliation. Finance owns the reporting bridge and connects confirmed cost movements to the broader cost variance review.
South's figures justify reviewing aggregate claim mix, reporting completeness, and payment timing. They do not prove that its work is less safe or that any particular treatment was inappropriate. Different job hazards and a small number of claims can affect comparisons.
Claim maturity matters too. Washington State L&I's experience-rating explanation explains that claim costs take time to estimate and that larger experience bases receive more weight. That supports caution when interpreting a short internal cohort; its premium rules should not be applied to this location table.
Employers should also keep the management review separate from formal reporting responsibilities. The New York Workers' Compensation Board's workplace injuries guidance distinguishes employer responsibilities, carrier responsibilities, and claim processes. Follow the applicable requirements for the employer's jurisdiction through the responsible specialists.
Bring a sample of your own aggregate loss-run, location, hours, and lost-workday files. Covirage maps the columns; its tools compute and check the comparisons, and the external AI model explains the results with every figure traced to its rows. Create a free Covirage account and bring a sample of your own data.
See cost variance analysis for the supporting business use.
Start with the carrier or administrator loss run, a location mapping, workforce hours for the incident period, and approved absence records. Keep the accident-date range and valuation date explicit. Use aggregate location outputs and exclude personal identifiers and clinical detail from the management review.
No. Paid cost records payments already made under the export's definition. Incurred cost commonly includes paid amounts and outstanding reserves, but field definitions and expense treatment must be confirmed with the provider. Keep each measure separately labeled rather than combining them into an unnamed cost total.
Existing claims may receive additional payments or revised estimates after the incident period closes. Comparing the same accident cohort at later valuation dates captures that development. It should be separated from the appearance of additional claims in a newer accident cohort.
A location summary does not calculate an insurance premium or an experience modification factor. Those calculations depend on the applicable rating rules and insurer inputs. Use this analysis for aggregate employer cost review, then ask the carrier or broker about the relevant rating calculation.