ABSTRACT
OBJECTIVE: to examine the relationship between carphology (aimlessly picking at bedclothes), floccillation (plucking at the air) and delirium in older patients admitted to specialist elderly care wards. DESIGN: daily observation for behaviours of carphology and floccillation embedded within a 'before' and 'after' study. SETTING: three specialist elderly care wards in a general hospital. SUBJECTS: older people admitted to hospital as emergencies. METHODS: patients recruited into a delirium prevention study were observed daily for delirium using the confusion assessment method (CAM). Occurrences of carphology and/or floccillation were also recorded. Sensitivity, specificity and positive and negative predictive values for carphology/floccillation for the diagnosis of delirium were calculated. Inpatient mortality rates were compared for patients who did, and did not exhibit features of carphology and/or floccillation. RESULTS: four hundred and thirty-seven patients were recruited into the study. One hundred and ten participants experienced an episode of delirium, 21 exhibited behaviours of carphology and/or floccillation. The sensitivity and specificity of carphology and/or floccillation for the diagnosis of delirium were 14 and 98%, respectively; positive likelihood ratio 6.8. Carphology and floccillation were associated with both hyper- and hypo-active delirium subtypes, and occurred early during incident delirium (67% within 2 days of delirium onset). The inpatient mortality rate in patients with carphology/floccillation was double the rate in patients without the behaviours (23.8 versus 11.2%, Fisher's exact test P = 0.16). CONCLUSIONS: carphology and floccillation are uncommon physical signs, but their presence is highly suggestive of delirium. The behaviours are unrelated to delirium subtype making their presence particularly useful in the diagnosis of hypo-active delirium.