What are the front end, middle and back end of the revenue cycle?
Front end, middle and back end are the three phases into which the market commonly divides the revenue cycle. The front end runs from scheduling to admission; the middle (also called mid-cycle or mid-revenue cycle) covers care, documentation and account closing; the back end starts when the claim goes to the health plan and ends when the payment is reconciled.
The split is not a regulation or an official standard. It is a management convention. It appears, for example, in training material published by the Healthcare Financial Management Association (HFMA), a US healthcare finance association. That material separates “front end” (patient access), “mid-rev cycle” and “back end” (patient financial services) [1]. Each hospital can adjust the boundaries, as long as it writes them down.
The eight stages in the Hospital revenue cycle guide fit into the three phases as follows:
| Phase | Pillar stages | Question the phase answers |
|---|---|---|
| Front end | 1. Scheduling and eligibility · 2. Prior authorization · 3. Admission and registration | Is the patient covered and the service authorized? |
| Middle | 4. Care and documentation · 5. Account closing and internal audit | Is everything that was done recorded, supported and on the account? |
| Back end | 6. Billing and submission · 7. Health plan review, denial and appeal · 8. Payment and reconciliation | Did the health plan pay what it owed, on time? |
Authorization does not end at the front end. Stay extensions and items requested during the stay, such as implants and special medical devices (OPME), bring the authorization center into the middle as well.
Each phase has owners, deliverables and a deadline
The split only works when each phase has responsible areas, a clear deliverable and a deadline. Otherwise the account falls between departments: the front desk never sees the denial it caused, and billing cannot control the missing chart entry.
See the phases as a table
| Phase | Stages | Responsible areas | Handoff |
|---|---|---|---|
| Front end | 1. Scheduling and eligibility · 2. Prior authorization · 3. Admission and registration | Front desk and admissions, scheduling center, authorization center | Admission |
| Middle | 4. Care and documentation · 5. Account closing and internal audit | Physicians and nursing, pharmacy, supply chain (OPME), internal audit, billing | Discharge |
| Back end | 6. Billing and submission · 7. Review, denial and appeal · 8. Payment and reconciliation | Billing, denial appeals team, finance and receivables | Submission to the health plan |
| Phase | Responsible areas | Phase deliverable | Deadline it works against | What usually goes wrong |
|---|---|---|---|---|
| Front end | Scheduling center, front desk, admissions, authorization center | Eligible patient, registered exactly as in the health plan's records, with a valid authorization number for the planned service | Health plan's authorization response deadline [6][7] and authorization validity set in the contract [5] | Invalid member ID, inactive plan, missing or expired authorization, missing form |
| Middle | Physicians, nursing, pharmacy, operating room, supply chain (OPME), authorization center, internal audit, billing | Complete account, every item supported by the medical record, closed soon after discharge | Claim submission deadline set in the contract [5] | Item used but not charged, entry without justification, stay extension not requested, account stuck after discharge |
| Back end | Billing, denial appeals team, finance and accounts receivable | Claim sent in TISS format, denial appealed on time, payment matched against what was billed | Payment, appeal and response deadlines set in the contract [5][6] | Rejected batch, late appeal, unnoticed underpayment, open balance without follow-up |
The middle has the most departments: it is where clinical teams, who do not report to billing, decide whether the revenue will exist. See the guide Medical records and billing integration.
Where does the account change hands?
The account changes hands at three handoffs: admission, discharge and submission. Each handoff needs an acceptance criterion: what the next phase checks before taking over the account.
| Handoff | From → to | Suggested acceptance criterion | Sign that the handoff failed |
|---|---|---|---|
| Admission | Front end → middle | Eligibility checked, registration matches the health plan, valid authorization for the planned procedure | Denial for registration or authorization reasons |
| Discharge and closing | Middle → back end | Account closed and audited, with complete clinical and OPME documentation | Growing unbilled accounts, technical denial |
| Submission | Back end → health plan | Batch in the current TISS version, within the contract deadline, with a receipt protocol | Rejected batch, denial for late submission |
These criteria are this guide's proposal, not an industry rule: the next phase only takes over the account when the previous one has delivered what it promised.
Discharge is the handoff that hides the most money. Between discharge and submission, all the time belongs to the hospital. The KPI for that balance is discharged not final billed, explained in the guide Discharged not final billed (DNFB).
Which KPIs should each phase track?
Each phase needs a few KPIs, each with an owner and a written definition. The table uses the KPIs from the Revenue cycle KPIs guide and shows the matching HFMA MAP Keys group as a reference [2].
| Phase | Hospital KPIs (numbering from the KPIs pillar) | HFMA MAP Keys group and examples [2] | Public benchmark in Brazil |
|---|---|---|---|
| Front end | Registration and eligibility denial (1) · Encounters with complete authorization (2) · Authorization response time (3) · Authorization denial rate (4) | Patient Access: pre-registration rate, insurance verification rate, inpatient and outpatient service authorization rates | No public hospital benchmark |
| Middle | Late stay extensions (5) · Technical denial (6) · Discharge-to-submission time (7) · Unbilled accounts (8) · Pre-audit coverage (9) · Value corrected in pre-audit (10) | Pre-Billing: days in DNFB, final billed not submitted, charge lag days | No public hospital benchmark |
| Back end | First-pass acceptance (11) · Batch rejection (12) · Initial denial (14) · Overturn rate (15) · Final denial (17) · Days to payment (18) · Delinquency (20) | Claims, Account Resolution and Financial Management: clean claim rate, remittance denial rate, aged A/R, net days in A/R | Anahp hospitals: initial denial 15.93% and accepted denial 1.72% (2025); days to payment 77.35 (2025) [3]; collection rate 88.61% and delinquency 61.53% (2024) [8] |
Three cautions. MAP Keys are a US reference: the same name may have a different formula in Brazil, and the HFMA page publishes no numeric targets [2]. Anahp figures are averages for member hospitals, mostly large ones: a trend, not a target. And Anahp's initial and accepted denial use different denominators and cannot be divided by each other [3][8].
Only the back end has public benchmarks in Brazil; the front end and middle are compared against the hospital's own history.
Front-end errors show up in the back end
Denial is reported in the back end, but its reason code often points to another phase. The health plan returns the claim review statement with each denial reason coded in table 38 of TUSS (Terminologia Unificada da Saúde Suplementar, the unified terminology of Brazil's private health sector) [4][9]. Reading that code by phase is the simplest way to send the error back to whoever caused it.
| Example table 38 code [9] | Phase where the error usually starts | Area that gets the feedback |
|---|---|---|
| 1001 — invalid member ID number | Front end | Front desk and registration |
| 1904 — length of stay inconsistent with the authorized procedure | Front end and middle | Authorization center and concurrent review |
| 3217 — no justification or insufficient justification | Middle | Physicians and internal audit |
| 3209 — duplicate procedure/item | Middle and back end | Billing |
| 3214 — charge submitted after the contract deadline | Middle and back end | Billing coordination |
Code descriptions are translated from Portuguese. Linking codes to phases is a management reading proposed by this guide; the same code can have different causes in different hospitals. See the guide TISS denial codes.
Feedback needs a routine: a monthly revenue committee, with one person from each phase, reviews denial by reason and by health plan and decides who fixes what. See the guide Denials committee.
How does the split differ in a Brazilian hospital?
The logic of the three phases is the same; what goes into each one changes. In the US, the middle revolves around coding diagnoses and procedures in ICD-10 and CPT [10], and the back end includes patient billing [1]. In a Brazilian private hospital the main payer is the health plan: health plans accounted for 76.47% of gross revenue at Anahp hospitals in 2025 [3].
Three differences shape who owns what:
The middle is about review, not coding. The account is built item by item with TUSS codes for daily rates, supplies, drugs and procedures [4]. In addition, 58.6% of health plans' medical expenses were paid fee-for-service in 2025 [3]. That is why internal account audit weighs more in the Brazilian middle.
The back end follows regulated contract rules. ANS Normative Resolution 503/2022 (RN 503/2022) requires the contract to describe the audit routine, the grounds for denial and the deadlines to appeal and to respond, with the same deadline for both [6]. Appeals go through the TISS standard itself [4]. See the guide Denial appeal.
The front end depends on health plan channels. Electronic exchange of eligibility and authorization is optional in TISS [4], and channels vary by health plan. Authorization response times follow RN 623/2024: up to 5 business days, or 10 for high-complexity procedures and elective admissions, with the shorter deadline of RN 566/2022 prevailing where it applies [7]. See the guides Eligibility verification and Prior authorization.
Who coordinates the three phases?
One person accountable for the whole cycle, with authority over the targets of all three phases, usually in the finance or administrative leadership. This is the same recommendation as the Hospital revenue cycle pillar. Each phase has a coordinator; the cycle owner arbitrates the handoffs.
Dividing gives each team focus; fragmenting is when each phase chases its own KPI without looking at the effect on the next, as when the front desk speeds up admission and leaves the registration denial to the back end. Managing the cycle as a single process, with an owner and KPIs, is what the market calls RCM; see the guide Revenue Cycle Management (RCM).
What we see in operations
At Accrua, a hospital assessment starts with the history of claims it has already sent to health plans. We read each denial claim by claim and trace it to the phase where it started: registration, eligibility and authorization in the front end; documentation, charge entry and claim closing in the middle; submission, collection and reconciliation in the back end. That mapping, not the denial code alone, shows which team needs to change its routine.
Day to day, Accrua OS runs on top of the HIS/ERP the hospital already uses and reads the whole claim before submission: medical record, prescriptions, items, OPME (implants and devices) and bundles, checked against the contract, the current price table, the authorization rule and that health plan's denial history for that item. When an alert points to a registration or authorization error, the fix goes to the front-end team instead of staying on the billing desk.
Agents handle what has a defined rule, such as XML validation, submission and deadline tracking. Anything that changes the value of a claim goes to a person, with the source in view, and what we learn in each appeal feeds into the next claim.
Where to start
- Write down the boundaries. List the eight cycle stages, mark the phase of each, and name the responsible area and the coordinator of each phase.
- Set acceptance criteria for the three handoffs (admission, discharge, submission) and agree with each area what happens when an account fails them.
- Pick two or three KPIs per phase, with a written formula and an owner, and measure for three months before setting any target.
- Classify denial by phase of origin using table 38 codes and send each group back to the area that caused it.
- Set up a monthly revenue committee with one representative from each phase and the cycle owner.