“Why Won’t It Let Me Order This?”: Coding, Prior Auth, and the Insurance Maze — IM Bootcamp

Why this session

No medical school on Earth teaches United States billing machinery — and that sentence lands hardest on international medical graduates, who arrive with excellent medicine and zero exposure to ICD-10-driven ordering, prior authorization, formularies, or visit-type coding. This is a system problem, not a knowledge problem, exactly like the presentation-format gap this bootcamp’s introduction describes — and like that gap, it is fixable in an afternoon once someone actually explains it. US graduates need this session too; IMGs need it taught explicitly rather than absorbed by humiliation.

The other reason this session exists: the physician cannot delegate the understanding. Support staff can send forms, but a prior authorization is won or lost on clinical criteria that live in the physician’s head, and a quality dashboard rises or falls on codes only the clinician can choose truthfully. Interns who understand why the machinery exists navigate it in minutes; interns who treat it as arbitrary hazing fight it for years.

What interns leave able to do

Attach a billable, truthful ICD-10 code to every outpatient order — and explain why screening versus diagnostic coding changes what the patient pays.
Navigate a prior authorization: why they exist, what documentation wins them, when to request the peer-to-peer — and what can be delegated versus what cannot.
Check a formulary before prescribing and a network before referring.
Distinguish the visit types — preventive versus problem, new versus established, Medicare Annual Wellness Visit versus annual physical — and set patient expectations before the visit starts.
Explain why quality measures and risk-adjustment coding move payment, and where the ethical line sits: code what is true, to full specificity, never more.

The case

Disclosed one part at a time; interns commit before each reveal.

Part 1 — The first clinic afternoon

Your third week of continuity clinic. In one afternoon: the vitamin D level you ordered bounces back — “order requires a diagnosis code” — and “fatigue” apparently doesn’t count. The pharmacy calls: the inhaler you prescribed is non-formulary on the patient’s plan — $340 to her, unless you switch to the covered alternative or file an exception. And the knee MRI you ordered for Mr. T has generated a prior-authorization request; the medical assistant left it on your desk with a sticky note: “they need clinicals??”

Your co-intern, watching you stare at the screen: “Just wait — open enrollment changes half their plans in January and you learn all their new rules again.”

For each of the three: what is the system actually asking for — and what is the move?

Part 2 — The “annual physical” that isn’t

Next patient: a 68-year-old woman, new to you though the practice has seen her for years, booked for her “annual physical.” She has Medicare. She expects a head-to-toe exam and “the full labs, like every year.” The schedule says: Annual Wellness Visit.

Halfway through the health-risk assessment she mentions, in passing, that her right knee has been swelling and buckling for a month.

What is this visit — and what is it not? What do you say to her before you examine anything? And what happens, for the visit and for the billing, when the knee enters the room?

Part 3 — The year-end view

December. Your clinic preceptor shows you two screens. The first is the quality dashboard: your panel’s diabetes control, blood-pressure control, cancer screenings, vaccines — some green, some red, each tied to the practice’s value-based contracts. The second is a “recapture” list: patients whose chronic conditions were coded last year but not this year — including your patient with diabetes, whose chart says “diabetes without complications” even though you have been managing her CKD stage 3 and her neuropathy all fall.

Why do these two screens exist? What does the incomplete diabetes code actually change — and what is the difference between coding her truthfully and gaming the system?

The machinery, explained once

Every order carries a diagnosis

Outpatient orders — labs, imaging, referrals — are paid against the diagnosis that justifies them, which is why the EHR demands an ICD-10 code before it releases the order. The code must be billable (specific enough to exist in the payer’s world) and truthful (the actual reason you want the test). The distinction with teeth: screening versus diagnostic. A screening colonoscopy in an asymptomatic patient is coded as screening and typically covered as prevention; the same order coded for a symptom becomes diagnostic, with different patient cost-sharing. Coding the visit’s truth is not bureaucratic trivia — it decides what your patient pays.

Prior authorization — why it exists, and whose job it is

Prior authorization is the payer’s utilization gate on expensive things — advanced imaging, brand-name drugs, procedures, some referrals: a checklist of criteria that must be documented before they agree to pay. Hating it is permitted; understanding it is required, because the navigation is learnable:

Know the triggers at your clinic — the MRI, the biologic, the sleep study — and anticipate: the authorization request written the same day as the order beats the one discovered two weeks later at the imaging center.
Win it in the note. Authorizations approve when the documentation answers the criteria: what first-line treatment was tried and failed, for how long, with what findings. If your note tells that story, the form is transcription. If it does not, no one downstream can fix it.
Delegation has a boundary. Your medical assistant can submit the form and chase the fax number — and should. But the clinical justification cannot be delegated: the MA does not know why conservative therapy failed or which guideline anchors the request, which is precisely why authorizations stall when they are handed off whole. The physician who understands why the gate exists, and feeds it the criteria it runs on, clears it in minutes.
The denial is not the end. Request the peer-to-peer, make the clinical case doctor-to-doctor, and document the outcome. A denial accepted silently is a treatment decision made by default.

Formularies and networks — the patient’s insurance is part of the prescription

Every plan carries a formulary — its tiered list of covered drugs — and a prescription written against it generates a phone call, a delay, or a bill the patient cannot pay. The workflow fix is small: check coverage at the point of prescribing where the EHR shows it, default to the covered generic when clinically equivalent, and treat the formulary exception as a mini prior-auth for the cases where the non-preferred drug genuinely matters. The same logic governs referrals: plans have preferred labs, imaging centers, and specialist networks, and an out-of-network referral — however excellent the specialist — can become the patient’s four-figure problem. Asking “who is in her network?” is part of the referral, not an extra.

Visit types — what the appointment actually is

Distinction	The teaching
Preventive vs problem	A preventive visit exists for screening, risk assessment, and the prevention plan; a problem visit exists for evaluating and managing complaints. They are coded — and often paid — differently. When a significant problem is addressed inside a preventive visit, both can legitimately be billed — but only when the documentation shows two distinct pieces of work, and the patient should hear that two things happened.
New vs established	By the standard convention, a patient seen by you — or a same-specialty colleague in your group — within the past three years is established; otherwise new. New-patient visits are valued differently. “New to me” is not the test; the practice’s history is.
Medicare AWV vs annual physical	The collision every clinic lives weekly: the Annual Wellness Visit is not a physical. It is a structured prevention visit — health-risk assessment, cognitive and depression screening, the screening schedule, a personalized prevention plan — and it explicitly does not replace a head-to-toe examination; a routine yearly physical is, in traditional Medicare, simply not a covered service.¹ The skill is expectation-setting before the visit: what today covers, what it doesn’t, and what a problem visit today would mean for billing.
Visit level	Since the 2021 overhaul, office-visit levels ride on medical decision-making or time — problems addressed, data reviewed, risk managed — not on the length of the review of systems.² Document what you actually weighed and decided; the level follows the thinking.

Quality measures and RAF — why the codes move the money

Quality measures are how payers see your care: diabetes control, blood pressure, screenings, vaccines, captured from claims and the chart. They are imperfect proxies — and they are increasingly how practices are paid, through value-based contracts and incentive programs, which is why the dashboard in Part 3 exists and why closing a care gap counts twice: once for the patient, once for the practice that keeps its lights on.

Risk adjustment is the other half: payers estimate how sick a panel is from its diagnosis codes. Many ICD-10 codes map to hierarchical condition categories (HCCs), which — with demographics — produce each patient’s risk-adjustment factor (RAF) score, and risk-adjusted payment models pay against that expected burden of illness.³ Two consequences for the intern in clinic. First, specificity is honesty: “diabetes without complications” on the patient whose CKD and neuropathy you actively manage understates the truth — her chart, her risk score, and the resources her care is credited with all read wrong. Chronic conditions also must be documented and coded each year to remain on the record the payer sees. Second, the ethical line is bright: code everything that is true, to its full specificity, supported by your documentation — and nothing that is not. Upcoding is fraud; undercoding quietly starves the practice and misrepresents the patient. The target is the truth, precisely stated.

Six encounters, coded out loud

Rapid fire — two minutes each. The facilitator reads the setup; the room names the visit type, the coding logic, and the trap.

Established patient, follow-up of hypertension and diabetes, medications adjusted.
A problem visit, leveled on the decision-making: two chronic diseases actively managed, data reviewed, regimen changed. The trap is documenting a wall of auto-pulled data while never stating the assessment and plan that actually justify the level.
“New patient” establishing care — but she saw your practice partner two years ago.
Established, by the convention: same group, same specialty, within three years. The trap is coding “new” because she is new to you — a routine downstream correction that teaches the rule better than any lecture.
Medicare patient booked for her “annual physical.”
It is an Annual Wellness Visit — prevention plan, risk assessment, screening schedule — not a head-to-toe exam, which traditional Medicare does not cover as a routine service.¹ The trap is discovering the mismatch mid-visit; the move is the expectation-setting sentence at the door.
Commercial-plan preventive visit — and “while I’m here, this mole has been changing.”
Preventive visit plus a significant, separately documented problem — legitimately both, when the note shows two distinct pieces of work. The traps run both directions: absorbing real problem work into the preventive visit for free, or nickel-and-diming a trivial aside into a second charge.
Screening colonoscopy referral for an asymptomatic 46-year-old.
Screening coding — which is what makes it prevention, with prevention’s cost-sharing. The trap is reflexively attaching a symptom code “to make it go through,” which converts a covered screening into a diagnostic study with a bill — the same test, the same scope, a different truth and a different price.
Knee MRI for Mr. T — the prior-auth sticky note from Part 1.
The criteria live in the note: six weeks of documented conservative therapy, the exam findings, the guideline anchor. The trap is handing the whole thing to the medical assistant — who can submit the form but cannot supply the clinical story — and then blaming the payer for the stall. Feed the gate its criteria; request the peer-to-peer if it still says no.

Key teaching points

The machinery is learnable, and understanding beats resenting. Every gate runs on criteria; feed it the criteria.
Every order carries a truthful, billable diagnosis — and screening versus diagnostic coding changes what the patient pays.
Prior auth: the form is delegable; the clinical justification is not. Win it in the note; escalate to the peer-to-peer.
The insurance is part of the prescription: formulary before prescribing, network before referring.
Know what the visit is before it starts — preventive vs problem, new vs established, AWV vs physical — and say it to the patient.
Quality measures and RAF move real payment. Code the truth, to full specificity, never more.

Pocket card

Carry this

Every order gets a truthful, billable code. Screening ≠ diagnostic — it changes the bill.
Prior auth: anticipate the trigger, win it in the note, delegate the form — never the criteria.
Formulary before prescribing. Network before referring.
AWV ≠ physical. Say what the visit is at the door.
Preventive + real problem = two documented pieces of work.
Code true and specific — chronic conditions, every year. Never more than the truth.

Variations

The angry-statement version: a patient calls about a surprise bill from encounter #5 coded wrong — run the service-recovery conversation and the correction workflow. Bridges to Day 2’s de-escalation session.
The EHR-live version: run the six encounters inside your actual EHR’s ordering and coding screens with a superuser present — the localization this session most rewards.
The IMG-focused run: for programs with large international cohorts, open with ten minutes of “what surprised you most about US clinic?” — the answers write the emphasis for the rest of the hour.

Notes

Every patient on this page is a fictional composite. The evidence for the bootcamp’s design is in the introduction.

Sources

American Academy of Family Physicians. (n.d.). How to implement and code Medicare’s Annual Wellness Visit (AWV). https://www.aafp.org/practice-operations/billing-and-coding/annual-wellness-visits ↩₁ ↩₂
Centers for Medicare & Medicaid Services. (2021). Physician Fee Schedule (PFS) payment for office/outpatient evaluation and management (E/M) visits fact sheet. https://www.cms.gov/files/document/physician-fee-schedule-pfs-payment-officeoutpatient-evaluation-and-management-em-visits-fact-sheet.pdf ↩
American Academy of Family Physicians. (n.d.). Hierarchical condition category coding. https://www.aafp.org/family-physician/practice-and-career/getting-paid/coding/hierarchical-condition-category.html ↩

This page is a teaching scenario for facilitated small-group education. Every patient in it is fictional. It is educational material, not billing, coding, or legal advice: payer rules, coding conventions, and documentation requirements change and vary by plan, state, and institution — teach from your compliance office’s current guidance, and localize before teaching. Last reviewed July 2026.

“Why won’t it let me order this?”:
coding, prior auth, and the insurance maze