Everything ED operators and physician groups need to know about billing, coding, and revenue cycle management, from E/M leveling and acuity capture to denials, discovery, and observation.
The emergency department is the only place in healthcare where the patient is seen first and asked about payment second. Federal law requires it. That single fact shapes every downstream step of ED revenue cycle management: the encounter happens before eligibility is known, before demographics are confirmed, and often before the patient can tell you who their insurer is.
Layer on the clinical profile. ED visits span the full acuity range in a single shift, from a sprained ankle to a STEMI. Documentation is written fast, in shorthand, by clinicians who are managing several patients at once. Procedures are frequent and easy to miss. Critical care time has to be captured explicitly or it does not exist for billing purposes. And the payer mix skews toward Medicaid, uninsured, and out-of-network commercial, each with its own rules.
The result is that ED billing rewards specialization and punishes generalists. A billing team that runs office visits or surgical claims well will still leave a meaningful share of ED revenue on the table, because the failure points are different: acuity under-coded, procedures dropped, coverage never discovered, denials that arrive weeks after the chart has gone cold.
This guide lays out the full ED revenue cycle, from the registration desk to the final patient balance, and the operational discipline that separates departments that get paid for the care they deliver from the ones that quietly subsidize it.
Every ED encounter produces two billable events. The professional claim covers the physician or advanced practice provider's work: the evaluation, the medical decision making, and any procedures they personally performed. It is billed on a CMS-1500 (837P) using CPT codes 99281 through 99285 and the relevant procedure codes.
The facility claim covers the hospital's resources: the room, the nursing time, the supplies, the equipment. It is billed on a UB-04 (837I) using the same E/M code range but leveled by the facility's own criteria, and reimbursed under OPPS for Medicare. We cover the facility side in depth in our ED Facility Billing guide.
Why this matters operationally: the two claims are leveled independently, often by different teams, and they frequently disagree. A professional Level 5 alongside a facility Level 3 for the same patient is a signal that one side is wrong. Independent ED groups that only see the professional side, and hospitals that only see the facility side, each miss half the picture. The groups that reconcile both sides catch acuity that was documented but never billed.
Emergency department E/M codes are leveled on medical decision making alone. Since the 2023 revisions, there is no time component for ED visits and no requirement to count history and exam elements. MDM is assessed across three dimensions: the number and complexity of problems addressed, the amount and complexity of data reviewed, and the risk of complications or morbidity from patient management.
99281: may not require the presence of a physician or QHP99282: straightforward MDM99283: low MDM99284: moderate MDM99285: high MDMThe single largest source of ED professional revenue leakage is the gap between Level 4 and Level 5. The reimbursement difference is substantial, and the clinical reality is that a large share of ED patients meet the threshold for high MDM: an acute illness that poses a threat to life or bodily function, a decision about hospitalization, parenteral controlled substances, or drug therapy requiring intensive monitoring. When the chart documents those elements but the coder does not recognize them, or when the clinician managed them but did not write them down, the claim goes out at Level 4.
The fix is not to code more aggressively. It is documentation-aware coding: coders trained on ED MDM specifically, tooling that flags charts where documented elements exceed the assigned level, and a feedback loop that shows clinicians exactly which sentences would have supported the level their work deserved.
Critical care (99291 for the first 30 to 74 minutes, 99292 for each additional 30 minutes) is billed by time, and that time must be explicitly documented. It excludes separately billable procedures and requires that the patient have a condition that acutely impairs one or more vital organ systems with a high probability of imminent deterioration. Sepsis, respiratory failure, unstable arrhythmias, and major trauma routinely qualify. The most common failure is simple: the clinician delivered critical care and never wrote down the minutes.
The ED performs more procedures per visit than almost any outpatient setting. Laceration repair by length and complexity, fracture care and splinting, incision and drainage, foreign body removal, lumbar puncture, central line placement, intubation, cardioversion, procedural sedation. Each carries its own reimbursement, and each is billed alongside the E/M with modifier 25 when the evaluation was significant and separately identifiable.
Procedures leak in predictable ways: documented in the narrative but not in the procedure note, performed by the APP and attributed to no one, supplies not captured, or the E/M level reduced by a coder who wrongly assumed the procedure absorbed the evaluation.
When a patient is placed in observation from the ED, the billing changes. Observation care codes (99221 through 99223 for initial care, 99231 through 99233 for subsequent days, and the same-day admit-and-discharge codes 99234 through 99236) replace the ED E/M for the physician who assumes observation care. The transition point, who owns the patient after it, and the two-midnight rule for Medicare all need to be handled deliberately. Observation is a frequent source of both under-billing and compliance exposure.
Payers that downgrade an ED visit based on the discharge diagnosis are frequently wrong, and the appeals are frequently won. But only if someone files them.
EMTALA requires a medical screening exam before any inquiry about payment. Registration in the ED therefore happens in pieces: a quick-reg at arrival to open the chart, then a fuller registration during or after the visit, sometimes at the bedside, sometimes never completed at all. Every gap in that process becomes a denial or a write-off 45 days later.
The common failure points:
ED practice groups feel this most acutely because they inherit the hospital's front desk. The registration mistake did not happen on the group's watch, but the denial lands in the group's AR. The operational answer is independent validation: run every record through demographic verification and real-time eligibility at the point of registration, and again before claim submission, regardless of what the hospital already did. Belt and suspenders, deliberately.
Upstream data quality is the highest-leverage investment in ED RCM. Every downstream process, from coding to denials to patient collections, degrades when the demographic and coverage data are wrong.
A meaningful share of ED patients who present as self-pay are not. They have Medicaid managed care they did not know was active, a marketplace plan they enrolled in and never used, coverage through a spouse or parent, retroactive Medicaid eligibility that attaches after the visit, or Medicare they assumed did not apply. They arrived in distress and did not bring a card.
Insurance discovery queries payer eligibility databases against the patient's demographics after the visit, surfaces coverage the registration process missed, and converts a write-off into a paid claim. For departments with high self-pay percentages, discovery routinely reclassifies a significant fraction of self-pay accounts to active coverage. Retroactive Medicaid alone, when worked systematically, recovers dollars that most departments never see.
Discovery has to run on every self-pay account before a statement goes out, and it has to run again at intervals, because coverage that was not active on the date of service can attach retroactively. The departments that treat self-pay as a final status leave that money behind permanently.
ED clinicians are not billing experts, and they should not have to be. But documentation determines reimbursement, and small, consistent changes in how a clinician records their decision making move the E/M distribution materially.
What does not work: generic coding education delivered once a year. What does work: per-clinician scorecards built from their own charts, delivered monthly, showing the specific encounters where documentation fell one element short of the level the encounter clearly supported. "You documented X; adding Y would have supported Level 5." Real examples, their patients, their words.
Paired with that: a documentation-intelligence layer that flags charts in near-real time, before the claim is coded, so the clinician can amend while the encounter is still fresh. That is where a platform like EventCare earns its place, by making the feedback loop measured in hours instead of quarters.
The compliance posture matters here. The goal is never to document what did not happen. It is to make sure the record reflects the complexity the clinician actually managed.
ED denials arrive late and cluster in predictable categories:
Downgrades deserve special attention. The prudent layperson standard, adopted in most states and in federal law for many plans, says an emergency is judged by the presenting symptoms, not the discharge diagnosis. A chest pain workup that rules out MI is still an emergency. Payers that downgrade based on final diagnosis are frequently wrong, and the appeals are frequently won, but only if someone files them with the clinical argument attached.
The operational distinction that matters most is real-time versus aged denial work. Denials worked within days of receipt recover a large share of denied dollars. Denials worked after 30 days recover far less, because filing windows close and clinical context evaporates. Every payer also has a payment cadence. When a claim ages past its expected pay date, follow-up should begin automatically rather than waiting for a denial that may never come.
Patient responsibility in the ED is large, unpredictable, and emotionally charged. The patient did not choose to be there. High-deductible plans mean commercially insured patients can owe thousands. And the first statement often arrives weeks after a visit the patient would rather forget.
The approach that works is the same one that protects the brand: reach the patient early, while they still remember the care and are grateful for it; explain the balance in plain language; offer payment plans and financial assistance screening proactively; and use the channels people actually respond to, which means text and email before paper.
For hospital-based departments, financial assistance and charity care policies must be applied consistently and documented. For independent groups, the balance-billing landscape has changed under the No Surprises Act and state laws, and collections practices need to be reviewed against them. In both settings, sending a self-pay account to collections before discovery has run is leaving money and goodwill on the table at the same time.
Five numbers tell you most of what you need to know about an ED revenue cycle:
Track all five monthly, by site and by clinician where relevant. The trend matters more than the point estimate.
Not every department needs to outsource billing. But every department should ask whether its current process is built for emergency medicine specifically or adapted from something else.
Signals that it is time to talk to someone:
QueueLogix was built for this environment. Our team has spent two decades inside emergency medicine revenue cycles, and EventCare runs the full cycle as one continuous system: validating data at registration, discovering coverage, flagging documentation gaps before the claim is coded, and following up with payers on their own schedule. Whether that means running your full billing operation or layering coding, CDI, and discovery onto the team you already have, the engagement is shaped around your objective, and the cost is tied to what you collect.
30 minutes. No slides, no sales pressure. Tell us about your department, your volume, and your current billing setup, and we'll tell you honestly whether we're a fit.