What You Need to Know About Medical Claims

Click the title above to view in your browser

Why this matters

Practice, Consult, Discover and Research all run on real-world claims data, not your EHR or internal billing system. A claim moves through several steps and hands before it ever reaches Jiro, and gets described using specific procedure and diagnosis codes along the way.

A handful of concepts (how long that takes, who's involved, what kind of claim it is, and what the dollar amounts actually mean) show up repeatedly across the platform. This page walks through those broad ideas once, and calls out which features each one touches.

Who's involved: clearinghouses, payers, and data vendors

Two different chains matter here, and it's easy to conflate them.

The claim's own path:

Medical claims follow the path of Submission → Adjudication → Remittance.

When you submit a claim, it typically routes through a clearinghouse first. A clearinghouse is a third-party intermediary that checks a claim for errors and translates it into the format each specific payer requires, then forwards it on, similar to a sorting hub that sits between your practice's billing system and the many different payers you bill. Payment decisions on these claims haven’t been made yet, which is why they are considered open.

Once a payer receives it, they take on adjudication, the actual decision on what to pay, deny, or adjust.

Once a decision has been made, the payer sends back remittance advice, often in the form of an ERA (Electronic Remittance Advice) for your practice, or an EOB (Explanation of Benefits) for the patient. This tells you what was paid, denied or adjusted and why, via remark and adjustment codes. Once this decision is communicated, these claims are considered closed.

How Jiro gets its data:

Separately, Jiro itself doesn't sit inside that pipeline. It sources de-identified, aggregated claims data from a variety of sources including:

Third-party claims data providers and vendors
CMS provider registry (to link claims to your NPI)
NDC/drug databases (for pharmacy claims)
Facility/location data vendors (for practice location and specialty enrichment)

This includes a mix of open and closed claims, which Jiro sorts and clarifies before displaying the information. This is also why Jiro's coverage is a percentage, roughly 60-70% of U.S. medical encounters and 50-60% of pharmacy encounters, not a complete record. Coverage is dependent on which payers and data sources participate.

Where this shows up:

Every feature's data completeness disclaimer traces back to this. If a patient's payer doesn't participate in Jiro's data sources, or a claim never reached a participating clearinghouse, it simply won't appear anywhere in Practice, Consult, or Discover, regardless of whether the visit happened.

Claims lag: why nothing you see is "live"

Understanding claims lag is crucial to interpreting the insights Jiro provides into your practice. After a claim goes through submission, adjudication and remittance, it is closed, also called a complete claim.

Claims lag is the length of time it takes for different types of claims to go from submission to complete.

Lag by claim type:

Pharmacy claims move fastest, with 90% reaching completeness in about 2 weeks.
Open (in-process) claims are usually still moving through clearinghouses, with 90% reaching completeness in about 8 weeks.
Closed (fully adjudicated) claims, usually from the payer itself, take about 6 months to complete.
QE (Qualified Entity) claims, meaning Medicare and Medicaid claims specifically, take the longest, about 9 months to complete, because government adjudication runs longer.

While some Jiro features can use recent claims data, others must wait for the full lag cycle to complete. This depends on whether they use open or closed claims.

Open claims carry a charge (billed) amount, but the paid amount may not be finalized yet. These provide insights with high recency but limited information.
Closed claims have finished adjudication and carry an allowed or paid amount instead of just a charge. These claims have high lag, but provide the full picture.

Where this shows up:

Jiro automatically filters by claim type to provide accurate insights into your practice. This takes a slightly different form in each feature:

Clinical Tab: When viewing clinical metrics, you can select a Data Quality filter from Complete, Recent, and Latest views. You can choose between fully-verified-but-older data (Complete) or fresher-but-less-complete data (Recent/Latest).

This is also why a metric might look different this week than it did last month even though nothing about your practice changed, more claims simply finished adjudicating.

Financial Tab: All features use a 12-month rolling window, defined above the information window

Referrals Intelligence: Uses open and closed claims, as long as referring provider was noted

Reimbursement Intelligence: Uses exclusively closed claims, as it requires information about the payers decision

Denial Intelligence: Uses exclusively closed claims as they are all linked to remittance data

Encounter Coding Intelligence: Uses open and closed claims, from professional claims only

Consult/Discover/Research: Use your specialty, open and closed claims to deliver personalized material. Patient level insights will follow the same lag.

Claim caveats that change the numbers

Differences in claims type, source, and how they are processed determine the scope of what information is included in each of Jiro’s features

Professional vs. Institutional claims

Professional claims cover office and clinic visits, and are attributed to the rendering provider on the claim (or the billing provider if that's blank). Institutional claims cover hospital and ED encounters, and are attributed to the attending clinician instead.

Where this shows up: Denial Intelligence, Reimbursement Intelligence and Encounter Coding Intelligence all work from professional claims only. If a lot of your billable work happens in a hospital setting, some of that activity won't be reflected in any of the three. All other features use a mix of institutional and professional.

Fully denied vs. Partially denied claims

A claim isn't just "paid" or "denied," it can be a mix. A fully denied claim was rejected in its entirety. A partially denied (line-item denial) claim had multiple service lines, and only some were denied while others were paid, so the claim as a whole isn't treated as a denial. Contractual adjustments (the write-downs already built into your payer contracts) aren't treated as denials either.

Where this shows up: Denial Intelligence only counts fully denied claims. A claim with one denied line and the rest paid won't appear there, even though it reflects a real, partial loss.

Billed charges vs. Allowed amount vs. Contracted (negotiated) rate

Three different dollar figures can describe the same claim:

Billed charges: What was invoiced to the payer. Usually the highest number, and not what you'd expect to actually collect.
Allowed amount (reimbursed): What the payer actually paid. Closer to what landed in your account, though it still doesn't account for patient responsibility or write-offs.
Contracted (negotiated) rate: What you're entitled to be paid per your contract with a payer, regardless of what any single claim happened to pay out.

Where this shows up: Denial Intelligence's dollar figures are billed charges, so the real recoverable amount is typically lower. Reimbursement Intelligence's "Reimbursed" tab uses allowed amounts, while its "Negotiated" tab uses contracted rates, which is why the two tabs can tell different stories about the same payer. Encounter Coding Intelligence’s RVU differential figures are Medicare-payment estimates based on wRVUs, not your actual contracted reimbursement.

Peer comparisons and CBSA

Many features compare you to "peers," physicians in your specialty matched to your geography, starting as local as possible and widening out if there isn't enough data. The exact geography levels used vary slightly by feature. See How We Build Your Comparison Group for the full breakdown.

Where this shows up: Every feature on the practice page use some version of this peer logic.

Claims Terminology

A quick reference guide of the codes and identifiers that show up across features, and what each one actually means for your data.

Procedure and Activity codes

Procedure and Activity codes describe the specific procedures and services billed on a claim, from an office visit to a surgical procedure. Jiro uses them for more than just identifying what happened at a visit: they also determine whether a patient counts as "new" or "established," which affects several Metrics.

Where this shows up: Encounter Coding Intelligence benchmarks the complexity level of your office/outpatient E&M coding (billed via specific procedure codes) against specialty and geography peers, surfacing whether you tend to code higher or lower than peers for similar visits. Reimbursement Intelligence's Negotiated tab breaks your contracted rates out by code. The Procedures category in Clinical Performance Metrics is also built around procedure code-level volume.

Diagnosis codes (ICD-10 / CCSR)

Diagnoses are recorded on claims using ICD-10 codes, the standard diagnosis code set in U.S. healthcare, but at a level of detail too granular for meaningful comparisons. Jiro groups related ICD-10 codes into broader categories using CCSR (Clinical Classifications Software Refined), a standard maintained by AHRQ (Agency for Healthcare Research and Quality). This keeps comparisons meaningful without requiring exact code-for-code matches between you and your peers.

Where this shows up: Referrals uses your CCSR-grouped diagnosis mix to narrow "missed opportunity" comparisons to your actual clinical scope, so a dermatologist isn't shown missed cardiology referrals. Research also draws on ICD-10-coded claims to surface relevant patterns. Patients category in Clinical Performance Metrics is also built around diagnosis codes.

CARC codes

Every denied or adjusted claim comes back from the payer with a standardized code explaining why, a Claim Adjustment Reason Code (CARC). It's used industry-wide, across Medicare, Medicaid, and commercial payers, so the same code means the same thing no matter which payer sent it. Jiro rolls the most common codes, by volume, into a handful of plain-language categories so you see a readable pattern instead of a list of codes, with less-common reasons grouped into "Other."

Where this shows up: Denial Intelligence's reason categories (like "Missing information or billing error") are built from CARC codes. The category label is what's shown by design; clicking or hovering over a slice of the breakdown will show the specific CARC code behind it.

RVU (Relative Value Unit)

An RVU is a standardized value assigned to each procedure code, maintained by CMS as part of the Resource-Based Relative Value Scale (RBRVS), measuring the physician work, practice expense, and malpractice risk involved in a service, independent of what any specific payer actually paid. The physician-effort component, work RVU (wRVU), is the piece most commonly used to compare coding and productivity across providers, since it strips out payer-specific pricing and reflects only the relative work involved.

Where this shows up: Encounter Coding Intelligence converts your procedure-level coding mix into wRVUs to show how many RVUs above or below peers you are, along with a complexity comparison, broken out separately for new and established patients since the two are coded and valued differently.

NPI (National Provider Identifier)

Your NPI is what ties claims data back to you specifically. Professional claims attribute to the rendering provider's NPI; institutional claims attribute to the attending clinician's NPI.

Where this shows up: Every feature built on claims data uses your NPI, and that extends beyond Practice. Consult uses your NPI-attributed claims history to personalize its "Personalized for you" section with your own patient population and practice patterns, and Discover uses the same data (specialty, diagnoses, procedures) to tailor which literature and clinical trials show up in your feed. If your patient counts or personalization look off anywhere in the app, NPI attribution is usually the first thing worth checking.

Still need help?

Reach out to support using the help button in the app or by emailing support@jirohealth.com