Utilization Data Employers Can Request From Carriers

Employers have a legal right to see how their health plan dollars are spent, down to the claim level in many cases. What they can actually get their hands on is a different matter entirely, and the distance between the two is where most employer-side healthcare strategy quietly stalls.

The Employer's Legal Right to Utilization Data

The legal entitlement is not in dispute. ERISA places a fiduciary duty on plan sponsors, and that duty is the legal basis for demanding data, not a courtesy carriers extend when they feel like it. The Consolidated Appropriations Act of 2021 reinforced that entitlement: employers can now request claims and utilization information from the vendors who administer their plans. On paper, the employer already has what it needs.

In practice, carriers, third-party administrators, and carve-out partners sit between the employer and its own data, and they tend to control the exits. If an employer doesn't push for them specifically, itemized claims, provider-level performance, negotiated rates, and detailed utilization patterns stay inside vendor systems. Some vendors even use separate data-sharing agreements to wall off access on confidentiality grounds, despite the CAA's explicit ban on gag clauses that would do exactly that. A Health Affairs analysis by Georgetown University researchers found the core problem in plain terms: federal law requires employers to obtain plan information from TPAs, but it does not explicitly require TPAs to hand it over, and it never specifies what form that information has to take. An employer can be fully within its rights and still receive a PDF summary that answers none of the questions it actually has.

The 2026 Pulse of the Purchaser survey points to why this gap matters beyond principle. It found that whether the employer can see its own claims data in the first place is the strongest predictor of whether an employer actually takes action on healthcare spending, more than how much pressure costs are putting on the budget. Cost pressure alone doesn't move employers to act. Visibility does. That finding reframes the whole exercise: getting access to utilization data isn't a compliance checkbox, it's the precondition for every plan design decision that follows.

Storage location and real access

Legal access rights mean little if the data sits somewhere the employer can't reach efficiently, and where that data physically lives turns out to predict, with striking consistency, how much of it an employer ever actually sees. The 2026 Pulse of the Purchaser survey found that three-quarters of employers store their claims data with their health plan or TPA. Of all the storage arrangements the survey measured, that same group had the lowest rate of complete medical claim-level access. So if you keep data in the vendor's own system, you tend to stay dependent on whatever reporting format the vendor decides to offer.

Compare that to the roughly one in four employers who store claims data with an independent data warehouse. Most of that group report complete medical claim-level access, which is the highest rate the survey found. The arrangement itself, not the employer's size or industry, appears to be doing most of the work here.

Self-funded plans compound the problem because they rarely deal with just one vendor. Medical, pharmacy, vision, and behavioral health claims typically run through separate administrators, each with its own system, its own coding conventions, and its own reporting timeline. You have to stitch those feeds together to get a single, plan-wide view, and small inconsistencies in coding or timing can throw off any analysis that tries to treat the data as one coherent set.

A carrier dashboard does not solve this. Dashboards and population health tools give you a general overview, but the carrier picks which metrics get reported and what format they take. They are not a substitute for claim-level detail, and treating them as one leaves an employer analyzing the vendor's summary of its own performance rather than the underlying activity. Before drafting any data request, an employer needs to answer a more basic question: where does this data currently live, and does that arrangement serve the employer's analytical needs or only the vendor's reporting convenience? Auditing storage location is the first step, not an afterthought that follows once a request has already gone out.

What fully insured employers can realistically request from carriers

Fully insured employers get the least access of anyone in the spectrum. The insurer, not the employer, retains the data by default, and the employer's leverage to compel disclosure is weaker than it is under self-funded arrangements. Fully insured employers still have options, though the requests have to be specific, and the odds of success depend heavily on timing and framing.

You can start by asking an insurer or broker for quarterly reports that show claims trends by category, such as ER visits or specialty drug utilization, and most carriers can produce something in this format without much effort. You can also ask for benchmarking data that compares the plan's usage and cost patterns against similar companies, so you get a market reference point even without raw claims. Renewal exhibits and utilization summaries round out the list: they won't show individual claims, but they do identify what's actually driving spend, which is often enough to inform plan design changes even in the absence of granular detail.

Some carriers also offer online dashboards or population health tools, which are worth requesting as a supplement, with the understanding that they show aggregated, carrier-curated data.

Renewal season is the moment when these requests carry the most weight. The carrier wants the business renewed, which gives the employer a counterpart incentive that doesn't exist in the middle of a contract term. Employers whose claims experience runs below the benchmark average have a specific asset to bring to that conversation: US-Rx Care notes that favorable claims experience can be used as explicit leverage in renewal rate and plan structure discussions, not just as a talking point but as a negotiating lever tied directly to dollars. Smaller fully insured employers without internal analytics staff don't have to build that capability from scratch. Brokers can request and analyze insurer data on the employer's behalf. The absence of an in-house data team is not a reason to skip the request.

Data Self-Insured and Level-Funded Employers Can Request and Own

The calculus changes entirely once an employer is self-insured. In that arrangement, the employer is the plan insurer, and that status carries a legal claim to data detail that no fully insured employer can compel from a carrier. Ownership, in this context, is the right. The TPA contract is the mechanism that either honors that right or quietly narrows it, and the two need to be checked against each other rather than assumed to match.

Self-insured and level-funded employers can reasonably request identification of high-cost claimants, de-identified, along with chronic condition tracking across conditions such as diabetes, hypertension, and musculoskeletal issues. Spending breakdowns by demographic factors, including age, gender, and family status, help employers understand whether plan design actually fits workforce composition or was built around assumptions that no longer hold. Utilization patterns matter too, particularly unnecessary ER use and preventive care adherence rates, because both show you where plan design changes could cut avoidable spend. Predictive modeling inputs round out the list, giving employers a way to forecast future claims and shape plan design ahead of the next renewal.

Level-funded plans sit in a hybrid position. They offer more granular access than fully insured plans typically allow, but how much claim-level detail you actually get depends on the TPA's contract terms, not the funding structure alone. That makes the contract itself the document that matters most. Employers should confirm that their TPA agreement explicitly grants claim-level data access, real audit rights, and the right to receive data in a machine-readable, portable format, rather than settling for summary reports dressed up as transparency. NIS Benefits points out that self-funded and level-funded plans typically provide claims-level reporting through the carrier or TPA, and that the real analytical value comes from reviewing patterns across multiple years rather than reacting to a single claims period in isolation.

Pharmacy claims need to be handled as their own category. PBMs often supply only high-level drug use and rebate reports instead of detailed claims data, so if you lack claim-level pharmacy detail, you have no way to check whether your drug pricing lines up with Medicare or commercial market benchmarks. An employer that has secured strong medical claims access but never pushed on pharmacy data has only solved part of the problem.

Prior authorization data as an underused utilization signal

Claims data shows what already happened. Prior authorization data shows what a carrier is choosing to block or approve before care ever happens, and it gives employers a second, largely untapped signal for evaluating carrier behavior. CMS-regulated insurers now have to publicly post their prior authorization approval and denial rates, so you finally have a market benchmark to measure your own carrier's practices against.

A KFF analysis looked at the first year of public reporting under the 2024 CMS regulation and found that Medicare Advantage, Medicaid managed care, and ACA Marketplace insurers each denied a meaningful share of standard prior authorization requests, with denial rates that differ by market segment. That analysis established, for the first time, actual baselines against which any single carrier's practices can be judged. The same KFF analysis found that denials are rarely appealed, but when they are, a substantial share get overturned, 67% in Medicare Advantage alone, with significant overturn rates in Medicaid managed care and the ACA Marketplace as well. So initial denials aren't always clinically grounded, and that matters if your workforce is experiencing delays in approved care.

Employers should request three specific things from carriers on this front: the full list of services requiring prior authorization, the plan's own approval and denial rates measured against the newly available public benchmarks, and response time data. KFF found median response times of roughly one day for standard requests across market segments, so you get a concrete yardstick for judging whether your plan's process is faster or slower than the norm.

One limitation deserves acknowledgment. The publicly posted metrics are aggregated across all services and don't break out denial rates by service category, so you can't tell from public data alone whether the specific services your workforce uses most are facing above-average denials. That requires a direct, plan-specific request to the carrier. KFF also found that the metrics are often hard to locate and interpret on insurer websites, and no standardized reporting template exists across carriers. Public data establishes the baseline. Getting a precise read on a specific plan still requires asking the carrier directly.

Transparency in Coverage machine-readable files as a benchmarking complement to internal claims data

Internal claims data tells an employer what it paid. It does not, on its own, say whether that price was competitive. Federal Transparency in Coverage rules fill that gap: they create a public layer of pricing data employers can check their own numbers against, no matter what the carrier chooses to disclose directly.

Together, the Hospital Price Transparency and Transparency in Coverage rules have produced machine-readable files, and these hold millions of in-network negotiated rates and out-of-network allowed amounts across providers and payers. The Hospital Price Transparency rule also requires hospitals to publish cash prices and payer-specific negotiated charges. US-Rx Care notes this combination lets employers analyze and compare real prices for the first time, instead of relying on a carrier's summary characterization of its own network discounts.

On their own, these files show the range of prices charged for the same service across different providers and payers, which is useful for high-level network and benefit design decisions. The real value appears when these files are paired with an employer's own claims data. Together, they let you audit TPA and carve-out vendor performance against actual market rates instead of the vendor's self-reported discount percentages. They allow benchmarking of projected claims spend against regional and national averages. They also strengthen an employer's position in vendor contract negotiations, especially in cases where the carrier's own reporting has been restricted or incomplete.

A finalized Utilization File rule, issued by the Departments of Labor, HHS, and Treasury, extends this framework further. It requires plans and issuers to publish a machine-readable annual Utilization File alongside each In-network Rate File, which will show which providers are actually serving enrollees as opposed to providers who remain listed in a directory without seeing patients from that plan.

One limitation applies here as well. Aggregate "discount off charges" figures hide provider-level detail, and a single high-volume hospital in a network can skew an entire average, leading an employer to believe its network pricing is stronger than it actually is. Rate files alone can mislead if read as a complete picture. Combining them with claim-level data from the employer's own plan is what corrects for that distortion and turns a market reference point into an actual audit tool.

The legal framework that backs up every data request an employer makes

Every request described above rests on the same legal foundation. ERISA imposes a fiduciary duty on plan sponsors, obligating them to act in the interest of plan beneficiaries. That duty includes confirming the plan isn't overpaying for care and that the vendors administering it are being properly monitored, not simply trusted by default. Data access, under that standard, is a requirement the fiduciary duty itself imposes.

That legal grounding is what turns everything in this guide from a wish list into something an employer can actually insist on. A fully insured employer asking for quarterly trend reports, a self-insured employer demanding claim-level pharmacy data, an employer pressing a TPA for machine-readable files rather than PDF summaries: each of these requests draws its authority from the same fiduciary obligation that made the legal right real. The gap between entitlement and access closes only when employers treat these requests as what they are, obligations owed to them under law, not favors a vendor may or may not choose to grant.