Dental Insurance Verification: What the CAQH Index Actually Measures (2026)

A manual dental eligibility and benefit verification costs the provider $6.52 per transaction, against $2.53 done electronically, and going electronic saves a dental provider 8 minutes per transaction. Those figures come from the CAQH Index, which is the only measurement of this that exists with a stated method. The dental industry ran 1.2 billion eligibility verifications in 2023, up 24% in a year, and CAQH puts the remaining annual savings opportunity for dental at $580 million.

Why is verification the job everyone wants to stop doing?

Before a patient sits in the chair, someone confirms their insurance is active, reads their benefits - annual maximums, deductibles, copay percentages by procedure category - checks frequency limitations, and works out what the patient will owe. Get it wrong and the practice absorbs a denied claim, or sends a patient a bill they were told they would not receive.

The task is not hard. It is repetitive, it is slow, and a meaningful share of it still runs through phone trees and hold queues. CAQH prints a dental practice's own words on the state of the alternatives: "When it comes to eligibility and benefits, I don't have an automated tool that I can trust, so I don't use it."

That sentence is the actual problem, and it is a trust problem rather than an availability problem. The infrastructure has largely been built. What has not been built, in many practices, is confidence that the automated answer is complete enough to quote a patient from.

What does the CAQH Index measure, and how?

CAQH publishes an annual index of US healthcare administrative transactions, and it breaks dental out separately from medical, which almost nothing else does. The 2024 Index covers data year 2023 and states its own scope: "this year's report reflects the collaborative efforts of over 600 provider organizations across the medical and dental industries, and health plans covering 63 percent of insured lives".

It is important to read the cost definition before quoting the cost. CAQH states that its per-transaction costs "include the labor time required to conduct the transaction, not the time and cost associated with gathering information for the transaction and follow-up" and that they do "not include system costs". So the figures below are labour on the transaction itself, and they exclude both the preparation around it and what the software costs you.

What does one verification actually cost?

CAQH splits the cost three ways - what the plan spends, what the provider spends, and the industry total - across three modes. The provider column is the one that lands on your payroll.

Three things are worth pulling out of that table. First, the provider carries roughly two-thirds of the manual cost and virtually all of the electronic cost: the plans have already automated their side almost completely, at two cents a transaction. Second, the portal is not the answer. A web portal saves the plan almost everything and saves the provider only $2.15, because somebody is still sitting at a screen typing. Third, the full saving available to a provider moving from phone to fully electronic is $3.99 per verification, which is the number to multiply by your own volume.

CAQH's 2025 Index, published in February 2026, reports "a 9 percent reduction in medical administrative spend and a 4 percent reduction in dental administrative spend", and notes that "the dental opportunity dropped 10%, as the cost to conduct transactions decreased across modes". The detailed dental per-transaction figures in that edition sit behind registration and a paid chartbook, so the 2024 numbers above are the most recent ones we could open and read in full. Expect the true 2025 gap to be slightly narrower than $3.99.

How much of dental verification is still manual?

Much less than the category claims, and this is the correction that matters most on this page. CAQH reports dental eligibility and benefit verification adoption for 2024 as 82% fully electronic, 15% partially electronic (web portals and IVR), and 3% fully manual, up from 75% fully electronic in 2022 and 79% in 2023.

The claim that "28% of dental practices use automated verification" and that "30-40% of plans still require phone verification", both of which appeared on an earlier version of this page, had no source and do not sit comfortably next to the CAQH adoption series. They have been removed.

What is the dental claim denial rate?

Nobody publishes one. This is worth stating plainly because the figure circulating most widely is attached to a document that does not exist.

The claim in circulation is that 12-16% of dental claim denials trace to eligibility errors, "according to the ADA 2024 Dental Practice Economic Survey". There is no ADA product with that name. ADA Health Policy Institute's dental practice research consists of the Survey of Dental Practice and The U.S. Dentist Workforce (ADA HPI), and neither measures claim denials.

The 19% figure that also circulates for dental is a misapplied medical number. It comes from KFF's analysis of ACA marketplace plans, which found that "insurers of qualified health plans (QHPs) sold on HealthCare.gov denied 19% of in-network claims in 2024" - and that brief explicitly excludes stand-alone dental plans (KFF, March 2026).

The closest thing to dental-specific evidence is a vendor survey that measures direction rather than level. Zentist surveyed "160+ dental revenue cycle and insurance billing professionals" and reports that "78% of respondents say claim denials and payer scrutiny have increased over the past 12 months" and that "seventy-one percent pointed to insurance verification as their top operational challenge" (Zentist, February 2026). Self-selected respondents, run by a company that sells dental RCM software, and a measure of perceived change rather than a denial rate. Useful as a signal, useless as a benchmark.

How does AI verification actually work?

How do you size this for your own practice?

Two of the three inputs are now published, which is unusual for this subject. The third is yours.

The reason the mode split matters so much is that a practice already running most verifications electronically has already banked most of the saving. The $3.99 applies to the calls, not to the whole book. A practice that still phones a third of its carriers has three times the opportunity of one that phones a tenth, and no industry average will tell you which you are.

What does this do to the people at the front desk?

Insurance verification is consistently described by dental front desk staff as the task they would most like to stop doing. It is repetitive, it involves hold queues, and the consequences of getting it wrong land on somebody else weeks later.

The training effect is the underrated one. Teaching a new front desk employee to verify accurately across dozens of carriers takes weeks, and it is weeks during which they are also learning everything else. With verification running as an exception queue, that training time goes to exception handling and patient interaction instead.

What should you check before buying?

Practice management system compatibility

Integration depth varies more than feature lists suggest. The leading dental platforms - Eaglesoft, Dentrix, Open Dental, Curve Dental - all have verification products that connect to them, but some write verified benefits directly into the patient record while others produce a result a human must review and post. The second kind captures a fraction of the time saving, because the review is the part that costs eight minutes.

Payer coverage against your own mix

No verification system reaches every payer. Government programmes vary by state and smaller regional plans may have no electronic eligibility at all. The question is not what percentage of national payers a vendor covers; it is what percentage of your patients' carriers it covers. Take your top twenty payers by patient volume and check them individually.

What lands in the exception queue

Ask the vendor what proportion of verifications it expects to flag for manual handling on a payer mix like yours, and what happens to those. A system that automates cleanly and then hands you a large exception queue with no context has moved the work rather than removed it.