Are Big Punitive Awards in HRT Cases Justified?

Law.com is reporting that Philadelphia juries have awarded a total of $103 million in punitive damages alone to two women in separate breast cancer product liability trials. The women claimed that hormone replacement therapy (HRT) was responsible for their subsequent development of breast cancer.

In light of the recent controversy over the use of Bayesian decision-making approaches to mammography and Pap testing in which probabilities of outcomes are estimated and benefits are then weighed against costs (including other bad outcomes) I thought it might be of interest to see if such an approach had been applied to HRT. Sure enough, "Bayesian Meta-analysis of Hormone Therapy and Mortality in Younger Postmenopausal Women" was just published in The American Journal of Medicine.

So what does it show? It shows that across a number of randomized controlled trials of HRT in postmenopausal women under 60 those women had a reduced overall mortality compared to those postmenopausal women under 60 who weren't on HRT.

As is often the case in these modern times science does not yield a cure but does allow one to pick one's poison as it were; not to avoid death but to influence the odds of whether you die of stroke instead of breast cancer.

HRT Litigation: An Alternate Cause or Perhaps an Explanation

In the hormone replacement therapy litigation there's obviously a fight over whether estrogen caused the plaintiff's breast cancer. Is the demarcation simply a line between those breast cancers fueled by estrogen and those that aren't? If so, are there alternate causes of estrogen-fueled breast cancers?

What about human papillomavirus? It's been identified as the cause of cervical as well as head and neck cancers. Now HPV has been suggested as a cause of breast cancer.  http://www.nature.com/bjc/journal/vaop/ncurrent/full/6605282a.html

But is HPV a cause independent of estrogen or does estrogen facilitate whatever HPV does that leads to breast cancer? And if it's the latter, how do you sort out causation?