Table 3.
Health Care Claims Databases
| Fair Health | MarketScan | Health Care Cost Institute Database |
Optum Labs Data Warehouse |
|
|---|---|---|---|---|
| Organization that owns the data | Fair Health Inc. (private non-profit) | Truven Health Analytics (private for-profit) | Health Care Cost Institute (private non-profit) | Optum (subsidiary of UnitedHealth Group – public, for-profit) |
| Data contributors | 60–70 private insurance carriers participating in the program (32, 80) | 150 employers (1), 21 commercial health plans (1), Medicare and Medicaid (24). | Aetna, Humana, Kaiser Permanente and UnitedHealthcare (47) | Affiliated and non-affiliated commercial health plans, provider EMR/EHR systems (82) |
| Sample size | 150 million covered lives, data gathered from about. 15 billion claims. Represents estimated 23.4% of national payments by privately insured patients (32). Represents 75% of privately insured population (32). | 230 million unique patients (since 1995), most recent year includes 50 million covered lives (1). | Claims related to 50 million unique people including individual, group, and Medicare Advantage members (44) (25.3% of nonelderly population with ESI) | 150 million unique lives: 19% of US population in commercial health plans, 19% of those in Medicare Advantage plans, 24% of those in Medicare PDP plans, and 7% of U.S. population with any health care utilization (82) |
| Geographic coverage | Covers every locality in U.S. (32) | Wide geographic range, but disproportionately covers South (7). 10–12 unidentified states for Medicaid sample. (1) | Unknown | Relatively geographically representative, concentrated in the South and Midwest (82). |
| Variables of interest | ||||
| Geographic level | Geozip (first three numbers of ZIP code), can be aggregated to State/MSA level | Geozip (first three numbers of ZIP code), can be aggregated to State/MSA level (7) | ZIP Code of with pops greater than 1350 (48). Core Based Statistical Area (only Metro areas with 50,000+ populations included) | Not clear—at least available by Census Region |
| Race/ethnicity | Unknown | Only for Medicaid (61) | No | Yes |
| Age | Yes (DOB) | Yes (DOB) (1) | Yes (DOB) (48) | Yes |
| Other demographics | Some patient and provider information is optional. | Gender, aid category for Medicaid populations (blind/disabled, Medicare eligible), employment status, relationship of patient to beneficiary urban/rural status (64) | Gender, relationship to policyholder | Gender, sociodemographic characteristics (111) Race, income, education, assets, health risk assessment, mortality available via linked secondary data sources (48). |
| Inpatient | ✓ | ✓ | ✓ | ✓ |
| Outpatient | ✓ | ✓ | ✓ | ✓ |
| Pharmacy | ✓ | ✓ | ✓ | ✓ |
| Lab | ✓ | ✓ | ✓ (41) | ✓ |
| Behavioral | ✓ (26) | ✓ (49) | ✓ | |
| Dental | ✓ | ✓ | ||
| Type of claim: fee charged vs. paid claim | All claims contain the fee billed by provider. About 50% of claims report “allowed charge” (80) | Claims represent the allowed amount paid by the plan (7). | Claims represent the allowed amount paid by the plan (48). | Claims represent actual paid amounts (12) |
| Run-out period | Database is updated twice yearly. Claims have a 3-month run-out (31). | Analysts can choose between “Early View data” with no minimum run-out, “Standard Updates” with 3 month minimum run-out, and “Annual File” with at least 6-month run-out (1). | Annual claims submitted at end of CY (44). Claims have a 5–6 month run-out period depending on payer (43). | Unknown |