Delirium and C-reactive protein
SIR—The paper by Macdonald et al. [1] suggests that C-reactive protein (C-RP) independently predicts incident delirium and recovery. This finding is based on a sample of 86 acutely ill patients aged 70 years or more, of whom 26 had prevalent delirium but only six developed incident delirium. The authors comment on the limitations of results based on small samples and the need to replicate their findings. We tried to do so using data from a larger prospective observational study of 283 unselected medical emergency patients aged 75 years or more, recruited within 24 h of admission in a district general hospital. Of these, 76 patients had prevalent delirium [identified by Confusion Assessment Method (CAM) and fulfilling DSM-IV criteria] and 29 developed incident delirium. C-RP levels on admission were available for 81% of patients.As previously reported [2], we found significantly higher levels of C-RP on admission in those with prevalent (67.1 mg/l ± 88.8) (mean ± SD) or incident delirium (80.1 mg/l ± 101.5) compared to those who never developed delirium (35.8 mg/l ± 51.3), P = 0.001. However, there was no difference between the prevalent and incident delirium groups (P = 0.83). In eight patients with incident delirium, C-RP levels measured at the time of onset of delirium did not differ significantly from those on admission (P = 0.091). We found no association between admission C-RP level and delirium duration (r = 0.041, P = 0.674) or delirium severity measured using the Delirium Assessment Scale [3] (r = 0.111, P = 0.286). Using a similar definition of recovery from delirium as Macdonald et al. (a CAM-positive assessment followed by CAM negatives until discharge) all but a single case of delirium recovered.
Our data show an association between high C-RP level on admission and both prevalent and incident delirium, but it remains a non-specific marker for a condition that is notoriously heterogeneous in its aetiology and presentation. We were unable to replicate the suggestion by Macdonald et al. that low initial C-RP level predicts recovery nor did it predict delirium duration or severity. Greater patient heterogeneity in our larger study of unselected patients may account for C-RP being less discriminatory in this population.
Multiple pathologies are implicated in the causation of delirium. The focus on a single pathway or marker risks over-simplifying the condition, which is potentially misleading clinically. Nevertheless, quantification of the delirium syndrome remains a challenge and, as suggested by George and Mukaetova-Ladinska [4], even non-specific markers can be helpful in raising awareness of this most frequent complication of hospitalisation for older people.
None
Department of Geriatric Medicine, Cardiff University, Llandough Hospital, Penarth CF64 2XX, UK
* To whom correspondence should be addressed E-mail: susan.white{at}cardiffandvale.wales.nhs.uk
References
- Macdonald A, Adamis D, Treolar A, et al. C-reactive protein levels predict incidence of and recovery from delirium. Age Ageing (2007) 36:222–5.
[Free Full Text] - White S, Calver BL, Newsway V, et al. Enzymes of drug metabolism during delirium. Age Ageing (2005) 34:603–8.
[Abstract/Free Full Text] - O'Keefe ST. Rating the severity of delirium: the delirium assessment scale. Int J Geriatr Psychiatry (1994) 9:551–6.[CrossRef][Web of Science]
- George J, Mukaetova-Ladinska EB. Delirium and C-reactive protein. Age Ageing (2007) 36:115–6.
[Free Full Text]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||