Thursday 27 January 2011


We've had this and this paper look at antidepressant medication use in mild depression.

Yesterday we had updated guidance on the use of medication in the management of anxiety. NICE guidance now is that, "GPs should not offer benzodiazepines or antipsychotics to patients presenting with generalised anxiety disorder (GAD)," the one page summary is here.

Will that mean their drugs budget comes over to me, so I can treat them then? :P

Tuesday 25 January 2011


A patient has great understanding of other people. She's worked in the hospitality industry so has worked with large numbers of people all her life. She's often been a kindly ear to those with troubled needs.

Recently she's needed a bit of help.

On discussing how she's an incredibly resourceful soul with plenty of ideas for others' problems, so has the tools in her toolbox to manage with a helpful steer, she smirked at me and told me, "It's easier to look into a goldfish bowl than to be in one."

Insight and genius, all in one.

Thursday 13 January 2011

Depression's Not Common

Clinical depression's not common, to my mind.

Unhappiness is very common. Unhappiness is an emotion, not an illness.

A little bit of unhappiness, or a lot of it, is a varying intensity of one normal human emotion, so even intense unhappiness is but part of life's rich tapestry and unwelcome though it may be, it's not an illness or a disorder.

Although I'm not a slave to diagnostic checklists, they certainly have a place and shape my thinking. On discussing someone's experiences the details, the nuance, the context, they matter. On considering someone's mental state, eliciting psychopathology and ascribing significance to it/formulating it in other ways and discounting it matters. History taking and mental state examination are clinical activites, not check lists and tick boxes, so there's always room for inclusion of folk and diagnosis of depression for people who don't present with a perfect list for the diagnosis to be textbook.

With this caveat that diagnosis is clinical, contextual and will at times vary from diagnostic lists, I'd concede that most of the time diagnosis falls within accepted frameworks. My training and practice is using the World Health Organisation's International Classification of Diseases, 10th Revision (ICD-10).

It's very useful. It's not just determining what a diagnosis is, it also determines what a diagnosis isn't. For example, "alcoholic" and "alcoholism" aren't in there so aren't and can never be formal diagnostic labels I make. Using the ICD-10 framework engenders a more transparent, consistent, reasoned process to formulation and diagnosis which I find helpful.

Low mood is very very common. Endogenous functional clinical depression arising through abnormal serotin/neurochemistry balance, much less so.

Does this distinction matter?

People with clinical depression have abnormal brain chemistry, this causes illness which has physical and psychological symptoms and signs. By inference, the chemical pathology causing this upset can benefit from chemical solutions (antidepressants). Clinical practice illustrates this, clinically depressed patients do get better on antidepressants.

People may also have low mood because their situation's ghastly and gets them down. If you're down because you're lonely or hungry or in severe chronic pain or you're dying or your loved one's dying or you're abused then is your intense unhappiness clinical depression, caused by chemical illness, so will happiness come in a tablet? Unfortunately not.

People may have low mood through organic syndromes. Stroke damage within the base of the brain or poor blood supply causing ischaemic damage within the diencephalon damages the limbic system, the mood centre in the brain. I've patients who have had cheery dispositions, had stroke damage, are depressed and it's through structural organic brain damage caused by the stroke. Antidepressants have little place.

If unhappiness is sometimes situational and sometimes through structural brain changes, then how do we pick up those unhappy folk who have clinical depression? This matters since such folk with severe clinical depression usually profit from antidepressant drug therapy.

What are the ICD-10 diagnostic criteria?

You need to be depressed for 2 weeks. Being depressed for a few days, then being okay, isn't enough.

You need to be depressed in most situations, most of the time. If you're low for the whole day, but you're okay when your kids visit and take you out to the pub for Sunday lunch, that's not clinical depression. Neurochemistry doesn't rapidly shift as you change from one room to another, so shifting from alone in your house to company in a pub suggests a more reactive/situational cause for unhappiness, or loneliness, rather than being consistent with the chemical illness of clinical depression.

Okay, you've someone with low mood, they've had low mood for 2 weeks, their mood's persistent and pervasive, isn't that just them being unhappy? Yes, it is. To be clinical depression you need to have a number of features :
· Depressed mood that is definitely abnormal for the person, present most of the day, almost every day, largely uninfluenced by circumstances, sustained for at least 2 weeks
· Loss of interest or pleasure in activities
· Decreased energy or increased fatiguability
· Loss of confidence or self esteem
· Unreasonable feelings of self-reproach or excessive and inappropriate guilt
· Recurrent thoughts of death or suicide, or suicidal behaviour
· Diminished ability to think or concentrate
· Changes in psychomotor activity (with agitation or retardation)
· Sleep disturbance
· Change in appetite (decrease or increase) with corresponding weight change

How many of the features correlates with severity :

Mild depressive episode :
"Two or three symptoms are usually present. The patient is usually distressed by these but will probably be able to continue most activities."

Moderate depressive episode :
"Four or more symptoms are usually present. The patient is likely to have great difficulty in continuing with ordinary activities."

Severe depressive episode :
"Eight or more symptoms must be present. Symptoms are marked and distressing, suicidal thoughts and acts are common and a number of ‘somatic’ symptoms are usually present."

The additional somatic symptoms that usually are present would consist of :
· Loss of interest or pleasure in activities
· Reduced emotional response
· Waking in the morning 2 or more hours before the usual time
· Depressed mood is worse in the morning
· Objective evidence of psychomotor retardation or agitation (reported/remarked on by another person)
· Marked loss of appetite
· Weight loss (5% or more of body weight in the last month)
Marked loss of libido

This is important, I believe, because if we diagnose clinical depression when it isn't then people get the wrong treatment. We know from last year's JAMA paper that antidepressants work no better than placebo in mild, moderate or severe depression and only are shown to work better than placebo in very severe depression. We know from a paper in this month's British Journal of Psychiatry that, still, antidepressants don't work in mild depression. Okay okay, studies show us averages and trends, individual patients may respond brilliantly to antidepressants despite have mild or moderate depression, but on average response is the same with antidepressant or placebo.

If we need 2, 4 or 8 core features of depression, and typically there're also some of the 7 somatic features of depression, to my mind that's getting to a much more specific (and smaller) group than all people who have intense low mood.

It's also a harder way to work. If we equate intense low mood with depression, so refer for psychological therapy and start an antidepressant, then everything's done and is easy. If we're sleuthing out who has a depression that's reactive/situational, who has a depression that's organic/structural and who has a depression that's endogenous/chemical, that's a more involved assessment. Worse, if only the last group generally profit from an antidepressant, we're then having to help people with ghastly low mood, suicidality and feelings of not coping through support that doesn't typically include antidepressants. We have to do more than just offer a prescription and refer to a psytchologist.

Harder work, both in assessment and interventions, but to my mind increasingly it's looking untenable to work in any other way.

Wednesday 12 January 2011

Being Left Alone

I work with a number of senior, experienced specialist practitioners who, in addition to the perk of getting to pay the NMC a little more each year, also are non-medical prescribers.

Like every Trust, we have a large number of policies, strategies, frameworks, pathways and protocols that are seen as telling staff what to do. Who likes this? I know of no clinical colleague who embraces such policy documentation with joy and enthusiasm. Some colleagues vociferously oppose it as curtailing clinical freedom and, in maveric piques, go against it all.

People don't generally like being told what to do, how to do it, when to do it and what format they have to use.

There's no policy documentation on how non-medical prescribers have to document their prescribing activity. As part of medicines management within our corner, I reviewed it. All of it. There was rather a lot of it. They'd extensive documentation of their Continuing Professional Development, tidily presented in a folder, with the activity and supervision and time taken. They'd evidenced letters to the GP and referrer for all prescriptions, with copies of all of these. They'd detailed the discussions before prescribing, with every case I picked up specifically detailing what had been discussed between them, the patient/carer and a Consultant Psychiatrist. Letters had details of all prescribing, including the specific prescription number of the script. They'd a photocopy of all prescriptions.

I thought it'd be a lot of work for them. It looked fantastic. It wasn't a lot of work, or time, they'd been incredibly efficiant. They simply detailed a one line record of CPD in their folder as it happened. They succinctly summarised prescribing in a couple paragraphs to the GP for each script. They used the copier in the office to keep a copy of all prescriptions, so it was just a few moments to do that. There was little time invested in to it, yet their audit trail of non-medical prescribing activity wasn't just okay, it was flawless.

It's good to see.

It supports what I believe, so is something that stuck in my mind. If folk are told what to do, it's oft times seen as a nuisance and hassle. Leave a nurse (all the non-medical prescribers were nurses) to just crack on and do it herself, without any direction, without any manager, without any Modern Matron, and they're utterly brilliant at sorting stuff themselves.

Now, if I can just persuade folk to leave clinicians alone more of the time . . .

Tuesday 11 January 2011


I took my children to see Megamind recently. Great fun it was, too. One line from it made me laugh out loud at the time then resonated again with me when I saw someone today.

After making a mistake, but both wanting to save face and keep a positive sense of self, Megamind 'apologised' by saying, "You were right. I was . . . less right."

Brilliant :P

Saturday 8 January 2011

Liaison Work

Acute hospital wards often stuggle with psychiatric problems. Medical and surgical colleagues frequently refer patients with depression who are miserable but not clincially depressed and are not ill. I used to think it odd, doctors (who all trained in some psychiatry as undergraduates) can't usually suss out if someone's ill or not.

I used to think that colleagues should be able to undertake a decent history, examination, ask about informant/contextual history and put together some sort of formulation. Partly through training, partly through prejudice, partly through not being their core business/not routine, and mostly because working on acute wards is now described as beyond grim, even competent colleagues with a will to do so struggle to make sense of psychiatric dimensions to their medical/surgical patients.

If history, assessment and formulation aren't undertaken in a systematic, thoughtful, involved manner then how can care planning be person centred and be effective? Largely it can't.

This generates more difficulties because the culture forced on clinicians through current pressures precludes better ways of working. Nurses know that is someone is wandering around or shouting or pulling drips out or pushing buttons on other patients' syringe drivers that usually drugs makes it worse.

Of course they do. We have no "anti wandering" or "anti shouting" or "anti fiddling" drugs, those are behaviours they're not illnesses. What's sought is abolishing symptoms, without attending to the primary cause of these presentations.

It's ingrained, though, so nursing staff are disempowered/under resourced in managing one to one patient care, call medical staff since it's unsafe and problematic, medics feel they can't do anything but prescribe. What do they prescribe? Usually they prescribe the drugs that don't work/make it worse. If used correctly, as well as usually doing little beneficial and lots that's harmful, the appropriate use of the drug also needs extensive ECG and blood monitoring that ties up even more nursing time, which could otherwise have been used in direct patient care to manage the behavioural disturbance.

There's a bit of an inconsistency, though. On the one hand the acute hospitals generally want a chemical solution, and nothing else, to manage behavioural disturbance. On the other hand, their care pathways for this are very different from most any other care pathway they'd generate.

The Dobbing Doctor explained his hospital has a "Rapid Tranquillisation Protocol" and asked what drug would be better.

A protocol on the administration of a drug to treat a behaviour. Is there an "Inhaled Breathing Protocol" giving the procedure for nebulised n-saline in obstructive airways disease or RSV/croup and the like? Is there a "Bolus Steroid Protocol" for injection of steroids in COPD? No, there's no "Bolus Steroid" protocol but there is a "Rapid Tranquillisation" protocol. Why? Why have a protocol for one class of drugs, major tranquillisers, but for most every other protocol it's about the disease state/care pathway? They have Stroke Pathways, COPD pathways, MI pathways but no delirium/acute confusional state pathway. Instead there's a drug protocol for "rapid tranquillisation."

This is less than ideal since we know medication has a tiny role to play, so if it's the only role your hospital offers in such patients' care then, frankly, patients aren't going to get the right care. Last September's update on National Dementia Strategy outcomes and Banerjee's Time for Action report reiterate the best practice guidance that non-pharmacological approaches must be tried first.

The reality is, they work.

I work with older adults. All my in-patients are ill, with major mental health problems. Almost all of them have physical comorbidity too, such as infections, contributing to their acute deterioration necessitating hospital in-patient care. Most are pretty elderly and pretty frail. Despite being confused, ill, unsettled and presenting with challenging behaviour, use of medication is modest. My last patient with hypomanic features arising through stroke damage was managed and discharged on no psychiatric medication. I've never used our equivallent of the "Rapid Tranquillisation" policy on any of my in-patients patients, ever.

If the frail, elderly, ill, most mentally and behaviourally challening in-patients can be managed with little/no psychiatric medication, it suggests that less mentally ill patients within acute hospitals also should be manegeable without drugs. Which is what best practice guidance, Royal Colleges, patient advocate groups and DoH reports say.

The crunch, of course, is that the environment and nursing ratios need to be fit for purpose. Nurse Anne reckons, ". . . that years of intense research have shown that the maximum number even the best nurse can take and not make mistakes is 4-6 on a general floor." She described how when things went well, "I worked a late shift on a 25 bed ward and there was 4 of us staff nurses and 2 care assistants for the ward." Our wards do have such staffing levels, with one qualified nurse per 4 to 6 patients, plus support workers, plus ward physio time and OT time and ward social worker time.

Rationally, if they're not psychotic, or having abnormal dopamine excess needing pharmacological treatment, why is an antipsychotic indicated?

The right environment and the right staffing obviates the need for antipsychotic medication, most of the time. I reckon that's one for the Ghost of Christmas Future to sort out . . .

Friday 7 January 2011


A post from my liaison psychiatry work, with musings on haloperidol. I'm not a huge fan of haloperidol.

It has been used by different countries as a drug of torture. Even in both low and moderate therapeutic doses it can cause akathisia.

But quite apart from being a drug with rather unpleasant side effects, more than unpleasant symptoms it can be quite harmful. In older adults with dementia, use of haloperidol is associated with increased risk of death, ramping up risk over the 10 weeks or so on average it was used for by about 60% to 70%. That's less than ideal.
It also explains how it triples your risks of strokes. Nice.

It's a drug with unpleasant side effect profile, it's dangerous, but what about practicalities of its use? Unfortunately it's a tricky drug to use correctly.

The Summary of Product Characteristics (SPC) reminds us that, "Cases of sudden death have been reported . . ." and that it has significant risks if you've heart problems. By significant risks, it's not that the drug has cautions and careful considerations to weigh up, no, it's that it's contra-indicated and should not be used. When is it explicitly contra-indicated, in black and white? When you have :
- clinically significant cardiac disorders
- recent acute myocardial infarction (heart attack)
- uncompensated heart failure
- arrhythmias treated with class IA and III antiarrhythmic medicinal products
- QTc interval prolongation
- history of ventricular arrhythmia
- history of torsades de pointes
- clinically significant bradycardia
- second or third degree heart block
- uncorrected hypokalaemia

Quite a list with pretty common problems, like heart attacks and heart failure, precluding the use of haloperidol.

If a physician is going to prescribe, what do they need to evidence in their records? The SPC directs that, "The risk-benefit of haloperidol treatment should be fully assessed before treatment is commenced," detailing how, including heart monitoring with 12 lead ECGs.

ECGs "particularly during the initial phase of treatment."
"Baseline ECG is recommended prior to treatment in all patients, especially in the elderly . . . during therapy, the need for ECG monitoring (e.g. at dose escalation) should be assessed on an individual basis."

"Periodic electrolyte monitoring is recommended, especially for patients taking diuretics, or during intercurrent illness," so older adults on water tablets or who are ill should have electrolyte monitoring, which by definition is pretty much all patients in acute general hospital beds (because they'll have intercurrent illness).

What does this mean? It means you've a drug with significant side effects, significant risks, marked increase in strokes and death, that shouldn't be used in patients with many common cardiac problems, it needs frequent blood monitoring for electrolytes, an ECG before starting and repeating an ECG on changing dose.

Odd, then, that on medical and surgical wards in the acute Trust it's widely used. Used on acute hospital wards, I can't say it's an intervention I'd wish for myself, or for my nearest and dearest . . .

Thursday 6 January 2011

Decision making

Towards the end of last year I generated a number of letters purely explaining decision making.

One was to a patient's wife, one to a GP and one to a clinical team on a general hospital's orthopaedic ward.

In all 3 cases the patient had been jointly seen by a CPN and me, investigations done, a diagnosis made, explanations given and appropriate management put in to place. But in all 3 cases there was a desire by another party, not the patient, to have more information. The theme was "why" with a desire to know why management was suggested as it was.

Now in all 3 cases the request for more information was polite and inquistive, rather than a forthright demand for explanations, so I'm more than happy to oblige. Sharing reasoning can only be a good thing. The original assessments and letters were extensive; typed up they were a minimum of 5 pages each. The management plans at the end were explicit and numbered, with from 7 to 11 action points in each case. With so much information, it's no surprise that folk want to know a little more about details of specific bits of this.

The wife sought to know why her husband's dementia medication was being withdrawn. The severity of his dementia, the side effects he was experiencing, the risks he'd shown (with ECG changes and collapses), the changes over time on the medication with no benefit seen by us or his wife, the diagnosis (he came to me on medication for Alzheimer's dementia but clearly has vascular dementia, which kind of accounts for why the drugs haven't worked) made decisions on the use of the medication pretty clear cut. Since he's off it he's fantastically better, his mood's settled/undistraught, his thinking's better with less preoccupation and his presentation's better with more successful acceptance of interventions, support and reassurance from his wife. But on stopping his medication his wife, not unreasonably, worried about this and sought more information. She explained how I'd spent an hour reviewing her husband with her, but only 10 minutes talking about medication. She's right, that's pretty much how time was spent. "Only" 10 minutes on why medication's not helpful is to my mind a reasonable amount of time to spend on sharing reasons and understanding, but a carer's mind is a whirl when a Consultant's questioning and investigating and formulating and changing stuff, so it's no surprise she mulled it over then wanted more information. It was no trouble to write to her, my secretary copying and pasting from the GP letter on file, detailing the 11 factors we'd considered that concluded in withdrawl of medication being appropriate. On subsequent review she's been very happy with it all.

The GP and surgical team both sought to know why his patient wasn't getting antidementia medication. It's a GP I know well and with her patient this had been explained but the patient's husband had turned up to the GP with lots of stuff clipped from papers and printed from the internet. The GP wasn't clear on how it all fitted with the current best practice and whether we'd considered relevant factors. On going through the notes it was clear we had, with over half a dozen specific statements about prescribing decisions in one letter alone and almost a dozen factors specified that influenced decisions on the patient's needs and consequent care planning. I liked this one, it was a chance to share with a colleague the reasoning behind prescribing decisions, the evidence base supporting this and the patient factors that impacted upon this. The GP gained understanding around how developing abnormal heart rhythm impacts on risk of medication causing serious harm. Similarly with the ward team.

I guess that like most clinicians, I favour spending most of my time speaking with patients and improving things with them. But sometimes it's also satisfying to have gentle challenge, stirring thoughts on decision making, to organise a response that articulates the contexts and evidence and findings and formulations and risks and degree of benefits and consequences, then share how these have been weighed up and how conclusions have been made.

All 3 times it was joint work with a CPN and others so the responses had effectively a case conference before responses were generated and this open discussion of how we worked and what we did is healthy for our team, reflecting on how we've worked.

So although I favour clinican contact and wouldn't want to navel gaze all the time, having gentle challenge/inquisitiveness to stir reflection of decision making is welcome and, to my mind, keeps us sharp in looking at how we work.