Tuesday, 9 March 2010

Medicines Management

I had an email request from Xavier asking a number of questions about medicines management. The main theme was, "how do you, as a
practitioner who commonly works with geriatric patients requiring various psychotropic medications, work to prevent ADEs in your everyday work?"

ADEs is, I presume, adverse drug events, or what we'd call side effects or treatment emergent adverse events.

The short answer is that I don't prescribe medication unless there's clear, rational grounds to do so. Local audit showed I prescribe at substantially higher doses than peers, but prescribe a lot less often. This doesn't surprise me - either medication shouldn't be used (and homeopathic doses to placate folk is not just an easy option/a con, it's a potential cause of harm) or it should be used at a therapeutic dose and evaluated.

The commonest intervention is to reduce medication. Many patients were started on medication, appropriately, but as time's moved on the "need" for such medication diminishes, if considering a risk:benefit ratio or NNT (Number of patients Needed to be Treated to get one patient to respond beneficially to that treatment). If on a betablocker, say, at age 87, what is the NNT to avoid one serious adverse event over the next year? What's the NNH (Number Needed to Harm) and is it comparable to the NNT? If the chance of beneficial treatment is comparable to the chance of harm, should the drug be used? Sometimes yes, since the issues are so significant, but often the benefit is modest and the utility of the drug can be contested (which is a polite way of saying it can be reviewed and stopped).

As well as often seeing a list of medication that's grown over years and can be reduced, there's also the dose regimen. Taking medication twice a day is much better than taking it four times a day, in terms both of concordance with prescribed use and in delivering care (with carers supporting this through monitored dosage systems). Muddled folk taking extra meds or missing meds can then be optimised which matters for side effects profiles of many drugs but is vital (in the true sense of the word) in the safety of others (like digoxin, warfarin, lithium).

Having looked at rationalising the drug regimen then optimising appropriate concordance with the prescribed regimen and support the next consideration is non-drug strategies. Can the clinical issue be managed without medication? Or can it be managed with medication but with other whatnots too, reducing the prominence of medication (and thus hopefully reducing the polypharmacy/dose)? This is often the case. We don't have tablets to stop people shouting or stop them wandering; management of behavioural problems invariably profits from behavioural solutions. Medication is, at best, usually just an adjunct.

Really, this is a major issue. The DoH have published the National Dementia Strategy and Prof Sube Banerjee's report highlight the need to prescribe seldom, but prescribe in an evidenced manner, in dementia care. The NHS also touts this view. So do the Alzheimer's Society. The National Institute for Health and Clinical Excellence offer support in dementia care through the NICE Clinical Guideline, CG 42.

Politics, the NHS, voluntary/patient advocate groups and expert bodies all saying the same thing. Amazing! Continued routine use of antipsychotics is grabbing headlines and, validly, is seen as poor practice. The evidence that such drugs cause strokes and death mean we really need good evidence to countenance their use, in our patient's best interests.

What about antidepressants? In mainstream psychiatry we've known they're mostly used as snake oil linament, a trifle to offer folk to take whilst nature takes it's course and things get better. What's big news is that arguably the world's biggest mainstream (not mental health) medical journal published a paper this year, a meta-analysis of antidepressant use, that shows how antidepressants in clinical depression of mild, moderate and severe depression are as effective as placebo. Only in very severe depression was their statistical difference between antidepressants and placebo.
Antidepressants arguably have a tiny or no role in Primary Care, which is why non-directive counselling and IAPT and CBT blossomed.

Obviously, I do prescribe antipsychotics and antidepressants. But invariably I want to see evidence of what the problems are before prescribing, then start drugs at a decent dose and ramp them up to a very decent dose for a decent duration, then evidence if there's improvement. And if not, the drug's stopped.

Really, drugs often don't help. We know that. We see it in clinical practice. We know it from theoretical models (if you've structural brain changes from dementia, why should fiddling with brain chemistry fix anything?!). But although evidence on effectiveness in clinical practice (rather than drug companies' research on efficacy) shows effectiveness of drugs, generally, to be pretty much approaching zero. So statistically they don't often do a whole hill of beans. But for individual patients you can get stunningly good results.

Part of the approach is offering something. It's not as if you get drugs or you get nothing. Our team supports staff/carers working with our patients, offering other solutions. This may be seeing a patient in day hospital every week over a month to see how they interact with others and respond when being encouraged to complete a task, to better inform their spouse how to manage with tasks in company like when they go around a supermarket or journey on a bus. It may be our staff going in to a care home at 7.00am to show the carers how to get someone out of bed and washed and dressed, successfully. It's sitting down with someone and showing them how to facilitate engagement with meals, so the patient won't wander off after 5 minutes or start throwing food or fiddle with other peoples' plates. Our team does teaching at 9.00pm to support night staff in optimal care. These interventions when successful often save staff time. Not using drugs can only viably be part of a care plan if there're other successful, meaningful options being offered instead.

The subsequent use of medication needs to be considered over time. Although it may be that psychotropic medication's used reasonably, with evidenced benefit, it then needs review at a later date to ensure it's continued use remains appropriate. whether it's a mental health team or community pharmacist or GP/Primary Care that do it, there needs to be a system for medication review. Being obsessive, I go through every MARS sheet in every care home in my patch, doing medication clinics, so review all medication that anyone's prescribed (GPs or out of hours or people moving into the area) of everyone, to minimise medication use. I can't see how we can undertake effective medicines management without regular review.

So there we have it, the themes I'd suggest to minimising medication use and thus reducing adverse drug events are :
- Stopping medication that's no longer necessary/of significant benefit
- Ensuring concordance is supported through reduction of number of doses
- Ensuring concordance is supported through Monitored Dosage Systems, if relevant, reducing medication errors/accidental overdose
- Use non-pharmacological strategies with practical support of staff showing people what to do
- Documented objective evidence of problems to address, then introduction of medication, then documented objective evidence of effectiveness (and withdrawl of the drug of it's not proven to be useful)
- Medication review


DeeDee said...

Hmm well I remember that conclusion on anti-depressants, but I was only ever dxed with "moderate" depression (because I could get up and walk around and go to M&S for chocolate and stuff - actually I was way too good at the getting up to no good part which was why I was so often in hospital) but I can definitely tell you that anti-depressants, when the right one came along, worked like night and day, and it was DEFINITELY not a placebo effect. (Nor was the mania caused by another anti-depressant). Or the improvement due to lithium...

Or perhaps you mean severe as in "severely batshit" not the strict medical term meaning "cannot get out of bed, won't wash or eat" or whatever...

I get very twitchy at anything that suggests my pills might be yanked from me. I like my pills, and I want to keep them. (The irony of a patient insisting that she wants to keep taking her medication does not escape me).

XE said...

What a truly excellent explanation, merci beaucoup Shrink!

The Shrink said...

DeeDee, where there's evidenced beenfit, we're all happy and the drugs should be continued. To "yank" them would arguably be unethical. An appropriate intervention's been initiated to help manage a health need, the intervention's shown to be successful, so there's no good reason to stop it. You're happy (you've got good support from your drugs) and as a shrink I'm happy (the drugs have been shown to make a substantial, meaningful difference for the better) so all's good!

Xavier, de rien!

moviedoc said...

I agree to an extent with your skepticism of medication efficacy, but does the patient not have any say at all? Most patients come to me because they want medication, and either psychotherapy has failed already or they don't want it. I offer them choices. Some drugs, like benzos and often stimulants I frequently refuse to prescribe, but there is usually a chance that something will be tolerable and effective. I also tell the patient that if they get better while taking meds they should not assume it was the med, in fact why not give yourself credit, then you have to decide whether to stop just to see if you relapse.

Again, the patient should have a say in this. It their mind, their body.

The Shrink said...

Moviedoc, of course. The patient has a say. But as medics we're obliged to offer what's appropriate for clinical need, not offer what's requested. We try and marry the two, but if a medication's not clinically indicated we can't offer it.

Obstetricians have been successfully sued for offering elective c-sections to women who requested them, because they weren't clinically needed, thus shouldn't have been offered/undertaken, even though the women really wanted them.

As such, if a drug's shown to be useful, I'm fine with it's use. What I'm not fine with is continuing a drug that's not shown to be useful, simply to be seen to be "doing something" since that's unhelpful all round.

moviedoc said...


Seems I've spent most of my career helping patients stop meds they assumed might have been doing something good. Most of them do fine (or better!) with less, different or none.

The Girl said...

Thank-you for this excellent explanation. It was a very helpful post.