Wednesday, December 3, 2008
Dirty secrets, revealed only in letters?
I'm recovering from both Thanksgiving (ha ha Mom just kidding it was wonderful) and a terrible 'head cold' (I'd never really had a proper notion of what that involved until the last few days - it's like my entire head is a bucket of increasingly tenacious mucous) and the whole world seems a bit blurry still. So perhaps you'll forgive the logorrhea below....
I wanted to highlight a couple letters to the editor recently about palliative care and patient survival. The letters are in recent Journal of the American Geriatrics Society (here with authors' reply here) and Journal of Palliative Medicine (here - you should be able to read the full text of this one freely online). The letters are from Drs. Andy Billings & Craig Blinderman from the Massachusetts General palliative care program & focus on a trial published in JAGS (here, and briefly blogged about here) last year of a home based palliative care program.
The letters point out that a notable and potentially controversial finding in the study (a randomized controlled of a home-based palliative care service intervention vs usual care for homebound patients with prognoses less than 1 year) was not discussed prominently in the article: that patients randomized to in-home palliative care had worse survival than the usual care patients (196 vs. 242 days, statistically significant). The authors respond that after adjustment for age and primary medical condition there was no significant difference in survival and so they didn't make much of that point. (This was a randomized trial so I don't quite understand if/how it's appropriate to control for baseline demographics after the fact - that's what randomization is for - and looking at the differences between groups at baseline they seemed pretty minimal to me.)
I'm not going to argue the point about whether the survival difference should have been addressed more prominently in the article or not: it wasn't the focus of the trial, and there may (??) be statistical reasons to question if it was a real finding, and perhaps the authors didn't want to gnash their teeth about it.
But that doesn't mean we can't or that it's not a topic worth discussing, because findings like this in some ways go to the heart of what we're doing in palliative care. I'm going to assume for this that, in fact, that outcome was valid: randomize a bunch of home-bound patients with poor prognoses to a palliative care service, meticulously discuss and document their treatment preferences, create a system which is more likely than it was prior to honor those preferences including ones for treatment limitations, and, yup, a sufficient number of those patients whose wish is to not go back to the hospital as long as they can be comfortable at home, or not go back on the vent, or whatever it is - a sufficient number of them get what they want - and die a couple months earlier than if no one had asked those questions/put that system of care in place - that this difference in mortality can be appreciated in a modestly sized trial.
It is possible that the JAGS trial demonstrated just that. The question to ask ourselves, then, is Is it time to panic?
I'd love to take a poll of readers (leave anonymous comments if you'd like) as to how many think that the above finding is 1) anything more than a little bit surprising, and 2) disturbing to them from a professional sense. My answers would be No and No. No to the second because I believe what I was taught and dutifully learned as a HPM fellow - in some cases goals determine prognosis. (Goals probably don't make much difference in advancing solid tumor malignancies, but for other diseases - dementia, COPD, CHF, etc. - people sometimes do get better with antibiotics, vents, etc. - if someone's survived the last 5 ICU stays there's no particular reason to think they're not going to survive the 6th - and if they make the decision for no more vents, antibiotics, ICUs, whatever then a good number of them will likely die a little earlier than otherwise - days, weeks, few months [but not likely longer than a few months for most patients with advanced illnesses]. Enough, though, that it could be measured in a trial....) Thus, sometimes the highest quality medical care, focusing on patient wishes and goals, will be life-shortening care. And that's an OK thing - in fact it's good as it's providing patients with the sorts of deaths they want - and these findings should be met with equanimity.
The question, for me, is that just because we're all comfy with this, if subsequent trials/studies reproduce these findings are patients/families, other clinicians, and policy makers going to be groovy with this as well? And if we're in a position in which we're very solidly comfortable as a professional community with the whole quality>quantity balance in advanced illness and others aren't then, well, there may be some issues. This is what I'm gnashing my teeth about a little, and am young enough not to have much perspective on this, and am curious as to what people who've been doing this much longer than me think. There has been a push by some within our community to deliberately de-link palliative care with death/dying/end of life care/hospice etc. I don't deny this is the right thing to do (not to get us out of the care of the dying - which, I suspect, motivates most of us clinically - but to remove the barriers to palliative care for 'non-dying' patients with chronic illnesses etc.) - so what does it do to our message if it's shown that palliative care (even if it's high quality, patient-centered medical care, and even if it's only in certain populations like the homebound chronically ill) 'hastens' death? 'Hastens' seems so bloody wrong there, doesn't it? But in a medical (and research?) culture in which mortality and survival are the premiere outcomes of merit, and in which death is medicalized and not seen as natural or ok/right, if you give 50 people X and 50 people Usual Care and the X people die a little quicker then you've 'hastened death,' even if all you've done is allowed some patients to die in the manner of their choosing so to speak.
I'm trying to think of what my patients would have to say about all of this, and I think most of them wouldn't have any problems with this: they get the whole quality>quantity thing (and, arguably, 'patients' have always been way ahead of the curve on this than docs have anyway....). I find that reassuring. However, there's a minority of my patients who would not 'get' or accept that, who are unwilling to contemplate any treatment limiting steps (outside of the period of imminent death where pretty much everyone says they want to 'be comfortable'), and for whom anything associating palliative care with treatment limitations or God forbid 'hastened death' would be a major barrier to them accepting our services.