For a long time I used a pretty simple coding shortcut:
Stable = 99213.
Medication change, worsening symptoms, or something more complicated = 99214.
It felt conservative. It also made me undercode a lot of visits.
The problem is that stability is not what decides whether a visit supports 99213 or 99214 by medical decision making. A patient can be doing well, have no medication changes, and still have moderate-complexity MDM.
That matters in psychiatry, because a very ordinary follow-up can look deceptively simple:
Routine. Stable. 99213.
That may be wrong.
I am not going to turn this into a full MDM tutorial. There are other posts for that, linked at the end.
For this one you need a single rule and two facts about psychiatric practice.
The rule: 99214 by MDM requires two of the three MDM elements to reach the moderate level. The three are Problems, Data and Risk.
In many routine psychiatric follow-ups, the path to 99214 does not involve Data at all. It is moderate Problems plus moderate Risk, and that combination can come up surprisingly often in psychiatry. Data can also be your second element, but it has its own counting rules and I am deliberately leaving it to the posts below.
One boundary before we start. This post is about selecting the level by MDM. For an established outpatient visit, 99214 can also be selected by qualifying time when that pathway is available to you. And if you bill a psychotherapy add-on, the E/M level is selected by MDM, not time.
This is the first piece people miss.
One stable chronic illness is generally low. Two or more stable chronic illnesses that you actually address during the visit can support moderate Problems.
So there is a real difference between stable ADHD alone, and stable ADHD plus stable MDD with both actively addressed. The patient does not need to be deteriorating. You do not need a new diagnosis. You do not need a dramatic clinical event.
This is also why the pattern is so common in psychiatry rather than being an edge case. Our patients frequently carry more than one chronic condition: ADHD and depression, depression and anxiety, bipolar disorder and insomnia, PTSD and GAD. Multi-condition follow-ups are the normal shape of the work.
The load-bearing word is addressed. Having MDD, ADHD, insomnia, GAD, PTSD and migraines sitting on the problem list does not give you moderate Problems if today's visit was really only about the ADHD.
This is the second piece people miss, and it hides inside another bad shortcut: medication changed means management happened, medication unchanged means not much did.
Prescription drug management can include starting a medication, raising or lowering a dose, stopping or switching one, or making a reasoned decision to continue the current prescription after evaluating the treatment.
That last one is the one that gets left on the table. If your patient is doing well on sertraline, and you assess symptoms, benefit, tolerability, adverse effects and whether anything should change, and then decide to continue it, you made a medication management decision. No medication change does not mean no medication management.
What it is not is this:
Medications reviewed.
The medication has to be part of your management decision, not an item you looked at.
Now go back to the patient from the top. MDD stable, ADHD stable, both addressed, response and tolerability assessed, and a documented decision to continue the current treatment.
Problems: moderate. Two stable chronic illnesses addressed.
Risk: moderate. Prescription drug management.
Two moderate elements. 99214 by MDM.
The patient was stable. Nothing got worse. You changed nothing. It can still be a 99214, and that is the part I think most psychiatric prescribers are leaving behind.
Now take the depression away.
Stable ADHD on Adderall XR. You assess symptom control, function, sleep, appetite, adherence and tolerability, review the PMP, and decide to continue.
That is real clinical work. But:
Problems: low. One stable chronic illness.
Risk: moderate. Prescription drug management.
One moderate element. 99213 by MDM. Whether the same encounter supports 99214 by time is a separate analysis using the time pathway.
The lesson is not "if I manage a prescription, bill 99214." That is just another bad shortcut pointing the other way.
The second frequent psychiatric pattern is one condition that is no longer stable.
A patient with MDD reports worsening symptoms and declining functioning despite treatment. You evaluate the worsening illness, review response and tolerability, weigh increasing against switching against augmenting, and make a decision.
Problems: moderate. A chronic illness with exacerbation or progression.
Risk: moderate. Prescription drug management.
99214 by MDM again.
Which is why stable versus unstable is such a poor proxy. Two stable chronic illnesses plus prescription drug management, and one worsening chronic illness plus prescription drug management, arrive at the same place by different routes.
Not "is the patient stable." Not "did I change a medication." And definitely not "did this visit feel complicated enough."
Where are my two moderate elements?
For many routine psychiatric 99214s the answer is Problems plus Risk, and usually one of two shapes: two stable chronic illnesses addressed with prescription drug management, or one worsening illness with prescription drug management.
If only one element reaches at least moderate, you are not at 99214 by MDM. If two do, you are. That is the whole distinction.
Everything above argues that you are probably undercoding. So the opposite error deserves saying plainly: do not go looking for ways to make a visit moderate.
A long note does not create a second element. Neither does listing every possible risk of a medication, and neither does a visit that was emotionally difficult. If only one of the three MDM elements reaches moderate or higher, you do not have 99214 by MDM. The fix is not more adjectives.
Document the work you did. Then code what the documentation supports.
I kept this post to one pattern on purpose. For the rest:
The 60-Second Code Check works through Problems, Data and Risk properly, including a fuller treatment of prescription drug management, and gives you a way to run it while you are charting.
Choosing Time vs MDM covers which pathway applies to the encounter in front of you.
Minimum Defensible Documentation covers what the chart actually needs once you have picked the code.
And if you would rather practice than read, MDM Foundations is live in the Practice Lab under Billing and Coding: classifying the domains, translating a real note into what it actually supports, and a decision sim where you pick the pathway and the code.
But the takeaway here is smaller than any of that.
Stable does not automatically mean 99213. No medication change does not mean no medication management. And before you default down, ask where your two moderate elements are.
Your turn. How often do you still catch yourself using "stable equals 99213" as the shortcut?