The gold standard for treating the "classic" bipolar symptoms is still lithium but valproate is also very effective. It depends on the prescribing doctor and other things like lifestyle factors and how that person is managing their symptoms; because lithium has a especially narrow therapeutic window, at least in the mainstream literature, this means that in order to get an effective dose you have to be skimming just below a dose that is toxic.

Obviously if it was incredibly risky they wouldn't prescribe it at all but if you have someone who is manic/hypomanic and you aren't able to supervise them closely, or if they are in the pits of crippling depression, then you can risk a person not eating and drinking enough (or getting blind drunk and dehydrating themselves, for example) or accidentally taking too much lithium because they're on a bit of a rampage or because they can't remember if they took their dose and suddenly you can have someone who is quite seriously sick from lithium. Of course as you're titrating up you need to carefully monitor the lithium levels of the patient and so for someone whose mood is way too high or way too low, that can be difficult to manage.

(Caveat to say that there's at least some indication that sub-"therapeutic" doses of lithium may be effective especially for depressive symptoms, but I wouldn't expect to take a low dose of lithium and have my depression cured or anything like that.)

So yeah, I wouldn't be surprised if a patient comes in with well managed bipolar, or what the practitioner presumes to be bipolar but they're in their 20s with no history of hospitalisations for manic episodes, and for them to prescribe valproate because it's easier to manage and if you're working with someone who isn't at imminent risk because they aren't on that really textbook Bipolar I rollercoaster ride then you probably don't have the urgent need for lithium to stabilise them and so valproate is likely a decent choice.

Lamotrigine is really useful but I still think that the best first line treatment is really lithium and then next is valproate.

Obviously it depends on the individual and their particular flavour of bipolar but the broad brushstrokes are that lithium is the first port of call for Bipolar I whereas for Bipolar II or BD-NOS etc. it seems as though valproate tends to be preferred and then lamotrigine is really good especially for treating bipolar depression, so if someone is probably bipolar but they are mostly experiencing depressive symptoms or if it's a person whose depression isn't responding to conventional treatments like antidepressants (where there's suspicion that it might be bipolar depression rather than the typical depression) then that's where lamotrigine can really be effective, as well as where there is insufficient response to lithium/valproate or the side effects are not tolerable.

But it really depends on different factors and how a person responds to the meds in question. (Obviously with lamotrigine there's a slim chance of causing SJS/TEN 😬 so if anyone's considering taking it make sure that you're aware of those symptoms and that you have made your prescribing doctor is aware of any allergies.)

I guess it's also a tricky thing because once you step outside the classic Bipolar I then it gets very murky with differentiating between the other bipolar diagnoses and MDD or schizoaffective disorder etc. and so the treatment is often a bit of guesswork and a lot of understanding that the map =/= the territory; just because a patient doesn't report symptoms that indicate something like psychotic depression or schizoaffective disorder doesn't mean that isn't what they're dealing with; my armchair expert opinion is that there's a significant amount of blurring and there's sub-clinical symptoms or symptoms that go unreported and so you might have a person whose diagnosis is bipolar but who doesn't respond to the typical mood stabilisers much/at all but who responds really well to an anti-psychotic. And then you're left to ask whether the diagnosis is accurate, whether there's something else like comorbidity that hasn't been identified, or whether there's something else that the antipsychotic is hitting that the lithium wasn't which would explain the response. (And I think a good psychiatrist is one who treats a patient not as a diagnostic label but who works to understand the symptoms that a person is dealing with and to determine their etiology, and who takes a very strategic and scientific approach to how a person responds to meds rather than having a very mechanistic "Bipolar In -> Lithium Out" sort of approach.)

But yeah, that's a long way of saying that lithium really is the closest we have to an ideal medication for textbook Bipolar I and that outside of Bipolar I it quickly gets very murky trying to know what will work/how well it will work/what's going on.