Paroxetine for hot flashes; cancer caregiver mental health burden

This week's J of Clinical Oncology has a few articles to note. First is a randomized & placebo controlled trial of 10 or 20mg of paroxetine to treat hot flashes (many, but not all, of whom had had breast cancer). It employed what seems to be a needlessly complicated double cross over design which lead to phenomenally complicated figures such as the one below. Edward Tufte would not approve. If anyone understands why they chose this study design let me know.

Nevertheless hot flashes did decrease in the paroxetine groups by ~40-50% & the difference was apparent within the first couple of weeks. Interestingly there were no real differences on quality of life (using several scales) between groups--people did sleep a little better with paroxetine though. 10mg was overall better tolerated (markedly less nausea than the 20mg) & preferred by the patients.
I'll also point out that besides the good folks at GlaxoSmithKline, a grant from the "Fashion Footwear Foundation/QVC Presents Shoes on Sale" people funded this study.

There's also piece about mental health problems amongst caregivers of patients with cancer. Briefly--13% of those studied probably met DSM-IV criteria for a psychiatric condition & 25% had sought some help for mental health reasons. However only 65% of those caregivers contacted participated in the study & there's no real comparison group. Also the cancer patients had to be able to give consent so presumably the caregivers of many of the sickest patients were excluded. What was most disturbing in some ways was that of those with psychiatric disorders only half of those who had sought care for their problem prior to their loved-one's cancer diagnosis had sought care after the diagnosis was made suggesting perhaps this is an easy time for these people to fall through the cracks.

Finally there was a review of acupuncture for chemo-related nausea, and a fascinating piece about presenting quality of life data to people in different graphical formats (lines vs. graphs etc.) that's a fine read.