This is my first “Ask the Expert” section – where I ask the smartest people I know about a topic that I find tricky.
My guest this post is Dr Stephen Ford, Consultant Psychiatrist, and keen worm-farmer. Steve gets very excited about neural pathways and lesions in old people. And don’t get him started on exotic neurotransmitters…. anyway I thought I would ask him a few garden-variety questions and see where I have been going wrong when it comes to assessing and managing the confused old person.
Collateral history about the timing of the onset of the confusion is the single most valuable piece of information in differentiating delirium from dementia. In individuals with pre- existing dementia, who are also at higher risk for delirium, establishing what constitutes a change from normal from an informant can be challenging.
At interview the presence of impaired attention is a useful sign. Attention has multiple components including attention to stimulus, maintenance of attention and shifting focus appropriately. At the bedside it is usefully tested by observation and the ability to name the months of the year / days of the week in reverse chronological order. Attentional deficits are usually mild in early dementia but much more prominent in delirium.
The presence of new onset visual hallucinations is more consistent with delirium though visual hallucinations do occur in dementia. Often these are accompanied by illusions and delusions and have a dream-like quality in delirious states.
Disorientation to place and the passage of time is prominent in both conditions. A trap is the patient who is reported to have had a disturbed night presenting in an oriented manner on the morning ward round (after reorientation by morning staff). Given the fluctuations in the condition people with a delirium are generally worse over the day than they appear first thing in the morning.
Something like a quarter of deliriums have their origin solely in pharmacological iatrogenesis. You are quite right to highlight anticholinergic medications (ie oxybutyinin, amitriptyline) as prime suspects given the importance of impaired acetyl choline neurotransmission in the pathway to delirium.
What’s less known is that many classes of medication have anticholinergic activity that may be significant in already compromised individuals (digoxin, warfarin, frusemide, H2 antagonists as per Tune et al Anticholinergic Effects of Drugs Commonly Prescribed for the Elderly. In general suspect any new medications that precede the onset of new confusion. Particular suspects include opioids, benzodiazepines, antihistamines and oral steroids. Dopaminergic medications in those with Parkinson’s disease are a common cause as is elevated lithium levels in those with chronic mood disorders. Drug interactions can frequently precipitate delirium. An under recognised interaction is synthetic opioids with serotonergic medications leading to a serotonergic syndrome (ie tramadol, pethidine [norpethidine is an anticholinergic metabolite], fentanyl or methadone with antidepressants). Buprenorphine and oxycodone are generally safe with antidepressants though may cause confusion in their own right. Beta blockers have been reported to cause delirium though I haven’t seen a clear case myself. Though not strictly medication side effects – alcohol and benzodiazepine withdrawal can be an easily missed cause of confusion in a hospital environment. A summary of the imperfect evidence for particular drug classes is here
High yield investigations would be a urea & electrolytes, full blood picture and urinalysis, though a broader screen is generally performed. Delirium is ultimately a clinical diagnosis based on the findings of the classic features. Generally a cause is evident but not always so trust your clinical findings over your potentially normal investigation results. Simple things to re-check include the drug chart, history of constipation and the possibility of withdrawal. Repeat the physical looking for focal neurological features. Be careful not to mistake a receptive aphasia for confusion. Its worth bearing in mind that in those with pre- existing cognitive impairment, recovery from a delirium may be prolonged even after the cause has resolved. If all else fails then further investigation of the rare will be needed and sometimes our only option is “the use of time as a diagnostic tool” ie waiting for the process to declare itself.
Management must be directed at the underlying cause of the delirium. This should include gentle reorientation, asking family to accompany the patient and use of a night light. The cycle of nocturnal agitation followed by sedation and then daytime somnolence leading to night time agitation is a troublesome one and can only be managed by avoiding sedation and keeping the patient awake by day. Sedation only has a role in preventing harm by virtue of a patient’s agitation (falls, aggression or absconding).
Antipsychotics are the least harmful but do carry risks of oversedation, parkinsonism and falls. The 2007 Cochrane review failed to show a difference in efficacy between haloperidol (doses of < 3mg) and atypical antipsychotics in managing delirium. There is some support for antipsychotics shortening the duration of postop delirium but a study with olanzapine and an orthopaedic population found the opposite so the jury is still out. In patients with marked persecutory delusions from their delirium then an antipsychotic may relieve distress. Some preclinical studies suggest atypical antipsychotics may be better at increasing cortical acetyl-choline than haloperidol but this is not well established clinically. I tend to use risperidone as it is cheap, comes in multiple forms and has a reasonable onset and duration of action. The main problem is the alpha blockade and postural hypotension. Doses should be much lower in the elderly than those used for patients with schizophrenia. If injection is required haloperidol is suitable with the main risk that of Qtc prolongation. There is an FDA black box warning about the use of injectable olanzapine and injections of benzodiazapines that dissuades me from using them together. Those with Parkinson’s disease should have dopamine agonists (pramipexole, cabergoline) stopped before entacapone stopped before L-Dopa. Obviously antipsychotics are relatively contraindicated – quetiapine is the least likely to worsen the Parkinson’s disease with doses generally in the 12.5-25 mg range (mean daily dose for psychosis in Parkinson’s disease is 75mg). Haloperidol absolutely should not be given in this population. If aggression is extreme benzodiazepines are the acute sedation of choice in this group.