I think that it is helpful to understand the specific concerns that relate to prescribe a stimulant to an adult, how much of a concern you consider them to be, and then have a strategy for each. I see these as the common issues:
1) You may believe that your patients are entirely honest, but find the phenomenology difficult to disentangle - this is in some ways the lest significant issue, and rating scales and more experience with assessment may be helpful. Neuropsychological testing could include 1) simply administering rating scales that are already available to you anyway (and may not add that much more information than a clinical interview) and 2) assessment of cognitive task performance for attention and executive function. The latter may be helpful in determining whether the patient has the underlying brain-based vulnerabilities that seem to be most common in those who end up with the clinical syndrome of ADHD. Note that many ADHD 'clinics' don't actually assess this and you may need to make sure you request this specifically.
2) You believe that your patient may be exaggerating symptoms - neuropsychological testing may help here, as it can be more obvious when someone is faking poor performance on a cognitive test versus reporting symptoms. Getting to know the patient over a number of visits may be helpful, and likely they will stop coming to see you if create barriers in the form of insisting on more than one or two visits to establish the diagnosis.
3) You believe the patient may intend to sell the medication (+/- exaggerating symptoms) - I don't actually know how one would detect this clinically, but it is clearly extremely common and large groups of young people have admitted to it on surveys. Again, simply being extra thorough may reduce the likelihood that you get a patient like this. Doing a UDS to ensure the patient tests positive for whatever you have prescribed is another way but it is not particularly reliable and I'm not sure it constitutes 'medicine' to do it for this purpose.
4) You believe the patient has a substance use disorder (+/- exaggerating symptoms OR being phenomenologically complex) - if you in general have the belief that people who have a substance use disorder shouldn't be prescribed stimulants, than UDS at some interval can address this concern. You could also assess the patient clinically for evidence of behavioral or physical manifestations of an active substance use disorder. It is of course possible that patients may have clinically significant attentional deficits and do a lot better with stimulants AND have difficulties with active substance use disorders.
I take the order of precedence as being that I first need to be confident that this is something that won't harm the patient. Then I need to believe it will also help the patient. I don't know that I can do much about it if they then also secretly sell some. I don't believe that patients with addictions shouldn't get any medications.
Some basic strategies that I believe allow me to help the most while minimizing risk is that I ONLY prescribe long acting stimulants (there are probably some patients who could do incrementally better if I gave them an afternoon dose of IR but they are suffering for the benefit of the overall practice being safer); I only give 1 month at a time for a fairly long period, I do a detailed assessment, and I try and develop a therapeutic relationship even though these are patients that could technically have very brief visits.