Self-Funded Health Plan Analytics: What Employers Should Measure

When you self-fund, every claim is paid from your own balance sheet, and that arrangement is now the norm rather than the exception: 67% of covered workers are enrolled in self-funded plans, rising to 80% at large firms. The average family premium reached $26,993 in 2025 after a 6% increase, and employers are being quoted renewals near 9.1% for 2026.

Despite carrying the risk, most employers see a single quarterly report from their carrier or TPA. It documents what was spent without explaining why, and self-funded health plan analytics is the discipline of closing that gap.

Why the standard report is not enough

A quarterly summary will tell you that spending rose 9%, but it will not tell you whether the increase came from three catastrophic cancer cases, a newly approved specialty drug, or a steady accumulation of emergency room visits that never needed to happen. Those three problems call for entirely different responses, and only one of them can be negotiated.

There is a compliance argument as well. Federal rules require plan sponsors to file an annual Gag Clause Prohibition Compliance Attestation, certifying that no contract prevents them from accessing their own claims and pricing data. As an ERISA fiduciary you are expected to understand what your plan actually pays, which makes data access an obligation rather than a convenience.

What should self-funded employers measure?

Six measures account for most of what moves a self-funded plan's cost, and they are worth establishing before anything more sophisticated is layered on top.

1. PMPM, split between medical and pharmacy. Track cost per member per month rather than total spend, so that changes in headcount do not distort the trend line. Separate the two categories, because pharmacy claims a larger share every year: drug spend reached 29.5% of total claim spending in 2025, up from 27.2%.

2. High-cost claimant concentration. Five percent of members generate 56% of spending, and the top 1% account for 28%. Track how many members crossed the $100,000 threshold, which conditions drove those claims, and how many were identified before the bill arrived rather than afterward. The same analysis should inform your stop-loss attachment point.

3. Price paid for identical services. Private plans reimburse hospitals at 254% of Medicare on average, though the state-level range extends from under 170% to above 300%. Pull your allowed amounts by facility for the ten shoppable services your members use most; the variation is almost always wider than anyone anticipates.

4. Avoidable utilization. Measure non-emergent emergency room visits, 30-day readmissions, duplicate imaging, and low-value testing, and express each as a rate per 1,000 members so that quarters remain comparable as enrollment shifts.

5. Chronic condition prevalence and open care gaps. Establish how many members carry a diagnosis of diabetes, hypertension, or a behavioral health condition, then how many currently have a gap open. An A1c that was never drawn or a prescription never refilled becomes a claim you pay for later, usually at a higher price.

**6. Pharmacy detail. **Monitor GLP-1 medications, specialty drugs, and your generic dispensing rate as separate lines. GLP-1s now represent 20.3% of prescription spend at roughly $7,400 per member annually, while brand and specialty drugs account for just 14% of prescriptions filled but 87.6% of pharmacy spend.

Two further measures are worth adding once those six are running: whether members are seeking care where your network is strongest, and how much of your spend sits with providers you hold no direct contract with. Both surface as leakage, and both respond to plan design rather than negotiation.

Three rules that make the numbers usable

Effective employer healthcare analytics depends less on the metric list than on the discipline applied to the underlying data.

• Risk-adjust before benchmarking. An older or sicker workforce will always appear expensive in absolute terms, and unadjusted comparisons teach you nothing you can act on.

• Use matching periods with adequate run-out. Setting a freshly closed quarter against a fully settled one makes costs look better than they are, so allow at least three months of claims run-out.

• Join your data sources. Medical claims, pharmacy claims, eligibility files, and laboratory results must resolve to a single member record, or the same person appears four times and every count is wrong.

Where to start

Every measure above depends on one clean, joined data set. Health Compiler consolidates claims, eligibility, pharmacy, laboratory, and EHR data into a single member record, then surfaces cost drivers, risk, and quality gaps in one place. Schedule a call to see what your own claims data reveals.