Long COVID Clinic Billing: Why U09.9 Isn't a Code You Can Just Drop on a Claim

VA Data Shows Documentation Rates Swinging From 3% to 41% Across Facilities on the Exact Same Diagnosis

By Team Rcmaxis  ·  November 9, 2026  ·  10 min read

Here's a number that should bother anyone billing for a long COVID clinic: in a VA study of 388,980 veterans who tested positive for SARS-CoV-2 during the Omicron era, only 5.28% had the post-COVID diagnosis code documented within 12 months of infection. A separate VA cohort of nearly 200,000 veterans found the same code showing up in just 13.5% of patients overall, but that average hid something stranger: documentation ranged from 10.8% to 18.1% across VA regions, and from 3.0% to 41.0% across individual medical centers. Same virus, same health system, a thirteen-fold swing in how often the diagnosis got written down.

That's not a quirk of VA data collection. It's a preview of what's happening in your own billing queue if your long COVID clinic doesn't have a consistent rule for when and how to use ICD-10-CM code U09.9, Post COVID-19 condition, unspecified. Get the sequencing wrong and you're either under-supporting the E/M complexity you're billing, or handing a payer a coding pattern that looks inconsistent enough to pull for review. Neither is where you want to be heading into 2027.

What U09.9 Actually Is, and the Sequencing Rule That Trips Up Billers

U09.9 became a billable, specific ICD-10-CM code effective October 1, 2021, built specifically to capture long COVID, also called post-COVID conditions, in claims data. It sounds like a simple add-on. It isn't billed that way. Per the ICD-10-CM Official Guidelines for Coding and Reporting, section I.C.1.g.1.m, the rule is explicit: assign a code for the specific symptom or condition related to the previous COVID-19 infection first, then add U09.9 as the secondary, additional code. The specific manifestation leads. U09.9 follows.

In practice, that means a patient coming in with persistent post-viral fatigue gets R53.83 (other fatigue) first, then U09.9. A patient with lingering shortness of breath on exertion gets R06.02 first, then U09.9. A patient with confirmed postural orthostatic tachycardia syndrome gets G90.A, the dedicated POTS code that's been billable since October 1, 2022, then U09.9. The one thing U09.9 is not designed to do is stand alone as the reason for the visit. It also should never be used for an active, current COVID-19 infection; it exists purely for the lingering aftermath once the acute illness has resolved.

The Documentation Gap: What the VA's Own Data Shows

The VA runs one of the most heavily studied health systems in the country for exactly this question, because it has the patient volume and the data infrastructure to track it. What two separate peer-reviewed studies found should worry any billing office running a long COVID service line. The 2022 study, covering 198,601 veterans, found long-COVID diagnostic codes documented in just 13.5% of the cohort within three or more months of infection, with facility-level documentation rates ranging from a low of 3.0% to a high of 41.0%. The researchers pointed to differences in local screening practices, not differences in how sick patients actually were, as the likely driver. Some facilities ran systematic phone-based symptom screening for every COVID-19 patient. Most didn't.

The follow-up study, covering a larger 388,980-veteran Omicron-era cohort, found cumulative U09.9 documentation at just 4.79% by six months and 5.28% by twelve months after infection, well below most published estimates of how common lingering post-COVID symptoms actually are. The study identified older age, female sex, Hispanic or Latino ethnicity, higher comorbidity burden, and more severe acute infection as factors independently associated with the code actually getting documented, with vaccination appearing protective against developing symptoms in the first place. A separate multicenter analysis spanning the VA, Beth Israel Deaconess, and UPMC found that even among patients who did get coded U09.9, the share who met a formal WHO clinical definition of long COVID ranged from 23.2% to 62.4% depending on the health system. That's the coding gap running in both directions at once: plenty of genuine long COVID patients never get the code, and some patients who do get the code wouldn't meet a strict clinical definition if you checked.

For a billing team, the lesson isn't abstract. If your clinic's intake and documentation habits aren't standardized, your U09.9 usage rate says more about your own charting discipline than about your patient population, and a payer's data analytics team can see that just as easily as a researcher can.

Multi-System Complexity vs. What Your Documentation Can Actually Support

Long COVID visits are exactly the kind of encounter that tempts a coder toward a high-level E/M code, and sometimes rightly so. The symptom picture is genuinely multi-system: persistent fatigue, post-exertional malaise, cognitive dysfunction that patients describe as brain fog, cardiovascular symptoms including POTS-type dysautonomia, respiratory symptoms, and a long tail of less common complaints. Research from the NIH's RECOVER Initiative identified a core set of symptoms, including post-exertional malaise, fatigue, brain fog, and dizziness, that best distinguish long COVID patients from those without it, out of a symptom list researchers say now runs past 200 documented presentations.

Here's the problem: a 99214, 99215, or a new-patient 99204/99205 doesn't get earned just because the visit was long or the symptom list was extensive. It gets earned through documented data review, number and complexity of problems addressed, and risk, the same three MDM pillars that govern every E/M claim. Long COVID visits lean heavily on subjective patient-reported symptoms and comparatively light on objective findings, which is exactly the combination that invites a payer to ask whether the chart actually supports the level billed, not just whether the visit felt complicated to the clinician in the room.

Build your templates so the coder has to pick a real manifestation code before U09.9 ever becomes selectable, and train providers to document the specific finding, test result, or functional limitation behind each symptom claim, not just a running list of complaints. That's what turns a defensible level-4 or level-5 E/M claim into one that survives a records request.

Payer Medical-Necessity Review: What's Coming for Extensive Long COVID Workups

Long COVID clinics tend to order broad diagnostic panels: autonomic testing, extended cardiac workups, pulmonary function tests, cognitive screening, and referrals across three or four specialties for a single patient. That pattern is increasingly drawing payer medical-necessity scrutiny, and it's a predictable kind of scrutiny once you understand how it's triggered. A broad standard panel ordered for every long COVID patient, regardless of their specific symptom profile, reads very differently to a payer's review algorithm than the same tests ordered with a documented, symptom-specific rationale for each one.

This isn't a named OIG Work Plan item yet the way RAC audits target skilled nursing or hospice billing, but the underlying exposure is the same shape: claims with a thin link between the diagnosis coded and the services actually performed are the ones that get pulled first when a payer runs a utilization review on a specialty with rapidly growing claim volume, which is exactly what long COVID clinics represent right now. Document the clinical reasoning behind every test order against the specific symptom it's investigating, not against a one-size panel, and you've already addressed most of what a reviewer is looking for.

Frequently Asked Questions About Long COVID Clinic Billing

Can you bill U09.9 as the only diagnosis code on a long COVID claim?

No. Per ICD-10-CM Official Guidelines section I.C.1.g.1.m, U09.9 is sequenced as an additional code, after the code for the specific symptom or condition tied to the prior infection, such as R53.83 for fatigue, R06.02 for shortness of breath on exertion, or G90.A for postural orthostatic tachycardia syndrome. Reporting U09.9 alone, with no manifestation code identifying what the patient is actually being treated for, doesn't reflect correct sequencing and leaves the claim without the clinical detail a payer needs to support medical necessity.

Why does long COVID documentation vary so much between clinics and facilities?

A 2022 VA study of 198,601 veterans found long-COVID codes documented in only 13.5% of the cohort, ranging from 10.8% to 18.1% across VA regions and from 3.0% to 41.0% across individual medical centers. A later VA study of 388,980 veterans found just 5.28% had U09.9 documented within 12 months of infection. Researchers tied the swing to local screening practices, such as whether a facility ran systematic phone-based symptom screening, rather than to how sick patients actually were. For a billing team, that means your own clinic's documentation habits, not just your patient population, largely determine how often this code shows up on your claims.

Does postural orthostatic tachycardia syndrome (POTS) have its own ICD-10-CM code for long COVID patients?

Yes. G90.A, effective October 1, 2022, is the dedicated billable code for postural orthostatic tachycardia syndrome. It sits in the nervous system chapter (G00-G99) and should be used, alongside U09.9 as a secondary code, whenever a long COVID patient's documented workup supports a POTS diagnosis rather than a vaguer autonomic-symptom description.