Sunday, June 7, 2009
Neurology has a fascinating study about the impact of delirium on cognitive decline in patients with dementia.
The patients, all with dementia, come from a single institution's large memory clinic, and were followed at regular intervals longitudinally. They essentially compared the rate/trajectory of cognitive decline in patients without incident episodes of delirium (~3oo patients) with those who with an episode of delirium (~70 patients).
(Note: they identified patients with delirium based on hospital chart review - it's unclear exactly how but it looks like it includes looking for documentation of acute alternations in mental status in the chart. This strikes me as a way of underestimating delirium as I often see it missed (diagnostically) and missed/ignored in charts. When I was a fellow I did a chart review for a project and remember seeing all these progress notes with physical exam documentations "A&O X3" pretty much every day. Concurrent with that the speech therapist would be leaving daily notes: 'Tried to do bedside swallow eval; patient lethargic and unable to cooperate.' This could go on for days. Anyway, one hopes they included speech path notes in their chart review.)
After adjusting for age, sex, educational level, dementia severity rating score, duration of dementia symptoms before diagnosis, family history of dementia, and number of comorbid medical diagnoses they found that those who had delirium had a more rapid rate of cognitive decline after delirium compared with those who didn't. In fact, the rate of decline for those without delirium was flat across the assessments; for those with delirium it sharpened (worsened) after the episode.
All this is based on a dementia severity rating score that I'm not familiar with and I can't really comment on the magnitude of the change, but it was statistically significant. The two groups' rates of decline (after controlling for the above factors) was similar prior to the interval in which the delirium occurred suggesting that in fact the episode of delirium was a disease modifying event, for the worst (the alternative explanation is that whatever caused the delirium was the disease modifying event). The authors give this ball-park estimation of the effect of an episode of delirium:
From a clinical standpoint, this study suggests that over 12 months, patients with AD who become delirious experience the equivalent of an 18-month decline compared to those who do not experience delirium.This so far has been the year of quantifying just how significantly delirium increases morbidity and mortality in older and particularly demented patients, and this paper adds to the mix.
A few thoughts & questions.... When I was taught about delirium in medical school and residency my sense was that it was this self-limited event/syndrome that occurred generally in the setting of acute medical illness/stressors (including drug toxicities) and then just kinda gets better once the problem causing the delirium is fixed/resolved/removed. I think I had some sense that it was a poor prognostic marker for those with dementia as well. However, it appears it can also be not just a prognostic marker but a disease modifying event as well (one that actively worsens prognosis and is not just an epiphenomenon of a poor prognosis) and can persist long term/indefinitely at times. Is this me just gaining some wisdom or is this an emerging concept that is coming out with stronger and stronger research backing the last several years?
In addition, if hospital delirium is so bad, why is its prevention and treatment so poorly studied? Prevention has some decent research base, treatment really has very little besides a few key articles. If the above is true one could argue that preventing a single episode of delirium is likely going to help someone as much as years of donezepil, yet why is delirium essentially being treated as an orphan syndrome? Cynical answer: drug companies aren't interested. Non-cynical answer: people are only just beginning to realize the long-term morbidity it causes.