Fatal haemopericardium and gastrointestinal haemorrhage due to possible interaction of cranberry juice with warfarin.

We report a case of fatal internal haemorrhage in an elderly man who consumed only cranberry juice for two weeks while maintaining his usual dosage of warfarin. We propose that naturally occurring compounds such as flavonoids, which are present in fruit juices, may increase the potency of warfarin by competing for the enzymes that normally inactivate warfarin. While traditionally regarded as foodstuffs, consumption of fruit juices should be considered when patients develop adverse drug reactions.

Validating a quality of life rating scale for idiopathic parkinsonism: Parkinson's Impact Scale (PIMS).

The signs and symptoms of idiopathic parkinsonism (IP, Parkinson's disease) are most commonly documented using one or more rating scales that assess physical limitations due to illness and drug side-effects with some attention being paid to depression. Scant attention has been paid to the impact of these limitations on a patient's life. Nurses in the Parkinson Foundation of Canada Clinical Assistance/ Outreach Programmes have designed a Quality of Life Rating Scale (Parkinson's Impact Scale, PIMS) to measure the impact of IP on 10 aspects of a patient's emotional, social and economic life. The scale had to fit onto one side of an 8 x 11 in. piece of paper, take a patient less than 10 min to complete, take fluctuations in symptoms ('on/off' attacks) into account, have unambiguous guidelines for the definition of each item, and a simple scoring system. A study was carried out to assess the reliability and validity of the scale. Nurses in nine Movement Disorder Clinics and one Outreach Programme participated. A total of 167 patients were asked to use the scale on three separate occasions, 1 month apart, without referring to the scores they had assigned to the scale in the previous month. Factor analysis identified four factors among the items in the scale: psychological, social, physical and financial. Internal consistency was 89.8% and the test-retest reliability was 72%. Construct validity was assessed by comparing factor scores and a global score between non-fluctuating patients and fluctuating patients in their 'off' state. The scores were significantly higher in the more severe 'off' state (p < 0.0001).