How medical codes are chosen is a question many patients ask after seeing a claim, because the claim shows short code descriptions that look nothing like the visit they remember. A coder does not pick a code from a list at random, and does not invent a description of the visit. A coder reads the clinical notes for that visit, checks what the clinician actually documented, and then matches that documented work to the code that best describes it. If the note does not support a code, the coder should not use that code, even if the service feels familiar or routine.
This process matters to patients for a practical reason. The codes on a claim decide how a health plan reads the visit, whether the plan asks for more information, and sometimes whether the claim is paid, denied, or paid at a different amount. Understanding the process helps you read a claim with less confusion, ask better questions when something looks wrong, and see why a billing office sometimes asks the clinician for a clearer note instead of simply changing a code. This guide explains code choice in plain language, without listing code numbers, and it is general information about how claims work, not advice about any individual claim.
What a coder actually reads first
A coder starts with the clinical documentation for the date of service, not with the bill. That documentation usually includes the reason for the visit, the history the clinician recorded, the examination findings, the assessment, and the plan. For a procedure or a test, it also includes what was done, how it was done, and what the result or finding was.
The coder reads that record as evidence. Every code that ends up on the claim should be traceable back to something written in that record. For background on the code sets coders work with, see the American Medical Association overview of the CPT code set and the Centers for Disease Control and Prevention page on ICD-10-CM.
That is why two visits that feel similar to a patient can end up coded differently. One visit may include a detailed discussion and a new plan, while another visit of the same length may be a brief check with no change in plan. The note, not the clock alone and not the patient memory of the visit, is the record the coder must follow. When a patient says the visit was longer or more complex than the claim suggests, the useful question is usually whether the note captured that extra work, because the coder can only code what is documented.
Coders also check the order of the information. The main reason for the visit usually comes first in the diagnosis picture, and supporting reasons follow. A service code then describes what was done during that visit. The pairing between the reason and the service is important, because a plan often checks whether the service makes sense for the reason given. If the pairing looks odd, the plan may ask for the notes or deny the claim until the connection is clear.
Diagnosis codes and service codes do different jobs
It helps to separate the two kinds of codes that appear on most claims. A diagnosis code describes why care was needed. It names a condition, a symptom, or a reason for the visit in a standard form that plans and providers both understand. A service code, often a CPT code, describes what was done. It names the visit, the test, or the procedure in standard form. One answers why, and the other answers what.
A coder chooses each kind from a different part of the note. Diagnosis codes come from the assessment and the documented problems, while service codes come from the description of the work performed. A coder does not choose a diagnosis code simply because a service was performed, and does not choose a service code simply because a diagnosis is present. Both must be supported on their own, and both must fit together. For a fuller explanation of service codes, see What Is a CPT Code. That guide explains how service codes are grouped and how they differ from diagnosis codes and from supply codes.
This separation also explains why a claim can look puzzling at first glance. A patient may expect the claim to read like a story of the visit, but a claim reads like a summary in code form. The story lives in the clinical note. The claim carries only the coded summary that the plan needs to process payment, and the plan can request the fuller note when it needs more detail.
The checks a coder makes before a code is final
Before a code is placed on a claim, a good coder runs through a set of checks. First, is the code supported by the note for this date, not by a previous visit or by a general problem list. A chronic condition that was not addressed at this visit should not drive the coding of this visit simply because it appears in the patient history. Second, is the code specific enough.
Coding systems allow different levels of detail, and a vague description is often not enough when the note contains the detail that would support a more specific code. Third, does the code match the documented work, rather than the work the clinician intended to do or hoped to do. Intent that was not carried out and not documented as performed is not coded as performed work.
Fourth, the coder checks for consistency across the claim. Dates, providers, places of service, and codes should line up with each other and with the note. An inconsistency, such as a service that does not fit the documented setting, is a warning sign that something needs clarification before the claim goes out. Fifth, the coder checks plan and coding rules that apply to that service, such as whether supporting documentation is required or whether two services are normally reported together or separately. Those rules vary by payer, so a code choice that is correct for one plan may need different supporting detail for another.
None of these checks is a judgement about the quality of the care itself. A coder is not deciding whether the clinician made the right clinical choice. A coder is deciding whether the written record supports the coded summary, and whether that summary follows the coding rules that the payer will apply.
Why coders sometimes ask for a clearer note
One of the most common sources of delay is a note that is brief, copied forward from an earlier visit, or written in general terms. A coder who sees a note like that faces a difficult choice. Coding beyond what the note supports risks a denial, a request for records, or a repayment later. Coding only what the note clearly supports may produce a claim that looks smaller than the visit felt. The correct step in that situation is usually to ask the clinician for clarification, through a query, so the record itself is completed before the claim is sent.
A query is not an instruction to change a code, and it should not suggest an answer. It asks what actually happened, what was examined, what was decided, and what the plan was, so the note can state it plainly. Patients rarely see this step, but it is one of the reasons a claim can take time to appear, and it is a reason to be patient when a billing office says it is waiting for documentation. A claim sent quickly with weak documentation is more likely to come back as a denial or a records request than a claim sent a little later with a complete note.
Patients can help this process in a small but useful way. At a visit, it is worth stating the main reason for the visit clearly, mentioning new symptoms or changes since the last visit, and asking the clinician to note any decision that matters to you, such as a change in treatment or a referral. You do not need to use coding language, and you should not ask for a particular code. Clear, plain language about what brought you in and what was decided makes the note easier to code accurately.
Common reasons a code choice looks wrong to a patient
A code can look wrong for several ordinary reasons that are not errors. The description attached to a code is often broad, so it may not mention your exact situation. A visit code, for example, describes a level of visit in general terms, not the specific conversation you had. A diagnosis description may use a formal medical term that sounds different from the words your clinician used with you. In both cases, the code may still be correct even though the wording feels unfamiliar.
A code can also look wrong because the claim shows only part of the visit. Some parts of care, such as care coordination or a brief discussion, may not appear as separate lines, because they are included in the main visit code under the coding rules. That does not mean the work did not happen. It means the coding system groups it with the main service rather than listing it separately.
Real errors do happen, of course. A code can be chosen from the wrong note, a date can be mistyped, or a service can be coded at the wrong level. When you spot something that does not match your visit at all, such as a service you did not receive or a date you were not seen, contact the provider billing office and describe the mismatch in plain terms. Ask what the note says, and ask whether a corrected claim is needed. For the wider claim journey, from visit to payment, see What Is Medical Billing. That guide explains where code choice sits in the full process.
What code choice means for payment and denials
Plans do not pay a code. They pay a claim that meets their coverage rules, their documentation rules, and their payment rules. Code choice affects each of those steps. A well supported code, paired with a matching diagnosis and a complete note, is more likely to pass the plan automated checks without a request for records. A code that is vague, inconsistent, or unsupported is more likely to be held, denied, or paid at a lower level than the provider expected.
That is also why a denial is not always a statement about the care. A denial can mean the plan could not see, from the coded summary alone, why the service was needed or what exactly was done. In that case the fix may be a clearer note, a corrected code, or supporting records, rather than a dispute about whether the care was valuable. Understanding this distinction helps patients avoid the frustration of arguing about the care itself when the real problem is the documentation behind the code.
It also helps to know that coding rules change over time. Code sets are updated, guidance is revised, and payer rules are adjusted. A code choice that was correct last year may need different documentation this year. That is a normal part of the system, not a sign that an earlier claim was wrong, and it is one reason billing offices review their coding regularly rather than reusing the same choices without checking.
How to read your own claim with this in mind
When your claim or explanation of benefits arrives, read it in three passes. First, check the basics: the date, the provider, and the service description in words, if a description is shown. Second, check whether the reason for the visit, as shown in words, matches the visit you remember. Third, check the amounts and the patient responsibility figure, and compare them with any bill that follows.
If the basics are right but the code wording feels broad or formal, that alone is rarely a reason to worry. If a service, a date, or a provider is plainly wrong, raise it promptly with the billing office, with the statement in front of you.
Keep your visit notes, your appointment confirmations, and your statements together for each episode of care. That small habit makes it much easier to spot a real mismatch and to explain it clearly. It also helps if a plan asks for information later, because you can point to the date and the provider without searching through months of paperwork.
A coder check you can mirror at home
When a claim line looks wrong, mirror the coder check in this order. First, read the visit note summary you received and write down, in plain words, what the clinician said was done and why. Second, compare that plain words summary with the short description on the claim line. Third, ask one focused question of the billing office, naming the date of service and the line you mean: does the note support this description, or is a clearer note needed.
This order keeps the conversation on documentation, which is the only basis on which a code can change in the US system. If the billing office says a note needs to be clarified, ask who will request that clarification and when you should expect a corrected statement. Do not accept a code change that no note supports, and do not pay a disputed line until the description and the note agree.
Frequently asked questions
Why does my claim use codes instead of plain words
Codes give every provider and every plan a shared language for diagnoses and services. Plain words vary from clinician to clinician, while a code means the same thing to every system that reads it. The clinical note still holds the plain language story, and the plan can ask for it when it needs more detail than the code provides.
Can I ask for a different code if I think mine is wrong
You can ask the billing office to review the coding, and you should describe what looks wrong in plain terms. You should not ask for a specific code, because the coder must follow the note and the coding rules. If a review finds an error, the fix is usually a corrected claim, not a change made simply because a patient requested it.
Does a coder decide whether my care was necessary
No. A coder decides whether the note supports the code and whether the code follows the rules. Medical necessity is a separate question, decided from the clinical picture and the plan coverage rules. A well coded claim can still be denied on necessity grounds, and that denial follows an appeal path rather than a simple coding correction.
Why do two similar visits sometimes get different codes
Because the notes differ. One visit may include a new problem, a change in plan, or more detailed work, while another visit may be a brief check. The coder follows the note for each date, so similar visits can quite properly carry different codes.
Sources
- American Medical Association, CPT code set overview
- Centers for Disease Control and Prevention, ICD-10-CM
For more on the claim journey, read What Is Medical Billing, and for the service code itself, read What Is a CPT Code. Together they explain how a note becomes a code, and how a code becomes a claim.
Cura Editorial Team
Contributor at Cura Health Journal.