Part 1 was about whether the diagnosis actually holds. Part 2 was about building a process around the evaluation. Part 3 was about expectations, functional targets, red flags, monitoring and boundaries.
This is finally the medication post.
There are plenty of ADHD medication charts online and nobody needs another table telling them Vyvanse is lisdexamfetamine and Concerta is methylphenidate. The problem is not that the information is unavailable. The problem is that there are now an absurd number of stimulant formulations, and most of us prescribe from a much smaller toolbox than the market actually offers.
I certainly do. I live in Adderall IR, Adderall XR and their generics, Vyvanse or generic lisdexamfetamine, methylphenidate IR, Concerta or another familiar methylphenidate ER, and occasionally Focalin. Those are the medications I know well. I know how they behave, how I titrate them, what problems I am likely to hit, and what my patients can usually actually get.
Then there are Mydayis, Adzenys, Dyanavel, Xelstrym, Jornay, Azstarys, Quillivant, QuilliChew, Cotempla and Daytrana, and for years I looked at some of those in a prescribing screen and thought: is this genuinely something different, or is this another expensive branded version of a stimulant I already prescribe?
Sometimes it is another way of administering a familiar molecule. Sometimes it solves a genuinely different problem. Working out which is which is what this post is for.
Why we stay in a small toolbox, honestly
Part of it is familiarity. We use what we trained on, what our colleagues use, what has been marketed to us, and what has worked before. Then familiarity reinforces itself, because the medication you know well is the one you can titrate confidently and troubleshoot quickly.
Part of it is insurance, and I think that part gets less acknowledgment than it deserves.
I hear clinicians talk constantly about fighting prior authorizations, and not only for ADHD medications. The same conversation happens about Vraylar, about Trintellix, about whatever has been marketed hardest to us most recently. I genuinely do not deal with that as often as many people seem to, and I do not think that is because I am better at appeals. It is because I mostly prescribe medications insurers already expect to see.
Whatever is newest tends to be both the most advertised to us and the most expensive to obtain, and those two facts arrive together. Generic mixed amphetamine salts, lisdexamfetamine, methylphenidate and dexmethylphenidate are simply easier operationally than asking for the newest branded delivery system, and the same pattern holds across the rest of psychiatry.
That does not make them the clinically perfect choice every time. It means friction shapes prescribing, and it shapes mine too. Familiarity produces coverage and coverage reinforces familiarity, and the loop closes quietly enough that you can practice inside it for years without noticing.
So the point here is not that everyone should start prescribing the entire stimulant aisle. It is two questions:
What problem does one of these other formulations solve that my usual medication does not? And immediately after that: can this patient realistically get it? A brilliant formulation that perfectly matches somebody's day is not a treatment plan if their insurer wants multiple documented failures, a prior authorization, and a copay they cannot afford.
Amphetamine or methylphenidate
I am deliberately not spending three pages here.
Both families raise dopamine and norepinephrine, but not in the same way. Methylphenidate mainly keeps more of what the brain already released available. Amphetamine does that too, and also increases how much gets released.
Does that tell me which one will work better for a given patient? Not really.
On average, amphetamines appear somewhat more effective than methylphenidate for adults with ADHD. That is a group-level finding, not a patient-selection rule. NICE treats either lisdexamfetamine or methylphenidate as a reasonable first-line option in adults, and recommends trying the other family when an adequate trial of the first does not deliver enough benefit.
So what actually makes me start with one rather than the other?
Prior response comes first. If this patient has already clearly done well on one family, I care far more about that than anything I can infer from mechanism. A first-degree relative's response is interesting, but I treat it as a soft clue rather than something I prescribe by. Either way it is the piece most often sitting unrecorded in an old chart, because somebody wrote "failed Adderall" instead of what happened.
Then tolerability. In my own practice, amphetamines more often produce the two problems I worry about most: appetite suppression, and a more physically activated or edgy feeling. Both families can do both. This is a tendency I use to choose the first trial, not a dependable rule.
If anxiety or activation is already a major concern, I usually start on the methylphenidate side. Low dose, slow titration. That is my preference, not evidence that methylphenidate is the anxiety-safe stimulant. Anxiety does not reliably predict which family will work, and treating the ADHD sometimes improves anxiety that was downstream of it all along.
Sleep usually does not choose the family for me. That is more often a timing and duration problem than a molecule problem.
Cardiovascular history usually does not choose the family either. I want a reasonable baseline sense of pulse and blood pressure, and I monitor them during treatment. In telehealth that does not necessarily mean an in-office set of vitals. A reliable home reading, recent documented vitals from a pharmacy or a PCP visit, or another reasonable source may be what I actually have. A concerning personal or family cardiac history is a different matter, and may change whether I start a stimulant at all or whether something needs evaluating first.
Substance use risk usually changes the formulation more than the molecule. If misuse or diversion is a meaningful concern, I favor a long-acting formulation over immediate release and match the monitoring to the actual risk, using the structure from Part 3.
And sometimes the delivery menu chooses for me. The two families do not offer identical formats. If I already know this patient needs a liquid, a chewable, an orally dissolving tablet, a patch, unusually early morning coverage or a very long day, that is worth knowing before I pick a family rather than after.
None of that tells me which one will work. It tells me which trial makes the most sense to run first, and that is a more honest thing to be deciding.
A quick note on the names, because they are confusing
Adderall is mixed amphetamine salts, and what the patient ends up with is roughly three parts dextroamphetamine to one part levoamphetamine. You will never see levoamphetamine listed anywhere on the label, because it is not a separate ingredient, which is why most of us have never heard of it. Plain dextroamphetamine drops that portion. Vyvanse ultimately becomes dextroamphetamine too.
And one trap. Lisdexamfetamine is not the l form of amphetamine. The lis refers to lysine, which is attached to dextroamphetamine. The body has to remove that lysine before any active dextroamphetamine is available, which is why Vyvanse behaves differently from a bead-based extended-release product.
On the methylphenidate side, Focalin is basically the more active half of regular methylphenidate. Why that matters comes up when we get to Focalin later.
Before choosing the next drug, name what actually failed
This is the part I would keep if you threw away the rest of the post.
"I failed Vyvanse" tells me almost nothing. What actually happened?
Those conclusions send you in completely different directions, and a medication change should solve a named problem. If I cannot say what I expect the next formulation to do differently, I am mostly expensive-guessing.
This is also the single biggest thing that determines whether the next clinician inherits anything useful. More on that at the end.
Same drug, same dose, different manufacturer
This is a variable I ask about far more than I used to.
Technically nothing changed on the prescription. The manufacturer did.
Someone has been stable for months. The pharmacy switches manufacturers. Suddenly the patient reports it lasts four hours instead of eight, barely seems to work, feels much more activating, or simply does not feel like the same medication. I seem to see this disproportionately with stimulants, and mixed amphetamine salts XR is where I hear it most often in my own practice.
FDA-approved generics have to demonstrate sufficiently similar drug exposure to the reference product. That matters and I am not interested in generic conspiracy thinking. It also does not require me to pretend every individual patient will experience every manufacturer's product identically.
There is a real ADHD precedent for taking these reports seriously. FDA initially allowed two generic Concerta products to be substituted for brand Concerta. After receiving reports that some patients were not getting the same effect, FDA reexamined them and concluded their later-day methylphenidate delivery might not match Concerta closely enough. FDA stopped treating them as automatically substitutable, and subsequent testing supported the concern with at least one of them.
That does not prove every generic complaint is about the drug itself, and it certainly does not prove one current Adderall XR generic is better than another. I have to watch my bias in both directions here. If several patients complain about one manufacturer, I can start expecting the next patient to complain and quietly turn my own experience into a rule that is not there. But I can make the opposite error just as easily, hearing "FDA-approved generic" and deciding in advance that the manufacturer cannot possibly have anything to do with what the patient is describing.
In my own practice I have had a disproportionate number of patients report poorer efficacy or tolerability with Camber mixed amphetamine salts XR, while Actavis has generally gone smoothly. That is my observation from my panel. It is not evidence that Camber is inferior for everybody, and another clinician may see the opposite pattern.
What actually persuades me is reproducibility in one patient. Stable six months. Different manufacturer. Sudden change. Back to the previous manufacturer, back to baseline. Same manufacturer reappears, same problem returns. At that point I do not need to prove anything about a manufacturer nationally. I have learned something useful about this person.
And the practical consequence is the one that matters: I do not want to increase somebody's stimulant dose to compensate for a problem that appeared the same month their formulation changed. Ask what was dispensed before you titrate.
The study clock is not the patient's clock
Stimulant charts make duration numbers look far more precise than they are, so it is worth saying plainly what those numbers mean.
"Benefit at twelve hours" means the treated group still differed from placebo at that measurement. It does not mean the effect was as strong at hour twelve as at hour four. It does not mean every patient gets twelve hours. And it absolutely does not mean the patient who reliably gets seven good hours has an abnormal metabolism or needs a higher dose.
If somebody tells me Concerta gives them eight very good hours, they get eight hours. I do not diagnose them with rapid metabolism because a chart on the internet says twelve.
Use those numbers for what they are actually good for: understanding what a formulation was designed to accomplish. Then believe the patient about what happens in their life. The study clock tells me what the formulation is capable of. The patient's clock tells me whether it fits.
If it works but causes side effects, ask when
The same discipline applies to tolerability. "Adderall makes me anxious" is not enough information.
When? As it first comes on? At the strongest part of the day? All day? Only when they have not eaten? Only on high-stress days? As it wears off?
A side effect tied tightly to the peak may be solved by dose or by release profile. A side effect that follows someone through several formulations in the same class makes the molecule more suspicious. One that follows them across both methylphenidate and amphetamine makes me question whether another stimulant formulation is the answer at all. And something that appears only as the medication wears off sends me back to rebound versus wear-off before I blame a medication that was working fine six hours earlier.
I went through that distinction properly in Rebound, Withdrawal, or Side Effect?, so I will not repeat it here.