Monday, May 11, 2009
JCO has a fascinating and frustrating study about physician-patient agreement about treatment goals predicting having a discussion about limiting life-prolonging treatments.
The study was a prospective study of ~70 hospitalized patients with cancer (mean age 58 years, the full spectrum of solid and hematologic malignancies was represented, 65% had KPS of 60 or less, and 25% had KPS less than 40) at a single German academic medical center. (They were eligible for enrollment if 'questions about forgoing treatment options were being discussed'; what that actually means, how they actually identified these patients, etc. is not entirely clear - one gets the sense that several research staff actively collaborated with hospital staff to identify them.) Patients were followed while hospitalized until death or discharge. One of the main thing they were looking at was whether patients described having a discussion about 'decisions to limit life-prolonging treatment' (DLT) with their physicians.
Most of the DLT that were discussed were about code status and not transferring to ICUs, with a smattering of others (e.g. stopping antibiotics). The persistent, yet always shocking, disconnect between what I'll generously describe as 'what we physicians think we are talking with our patients' about and what they actually perceive, recall, and identify what we are talking about is demonstrated by this paragraph:
Physicians informed the vast majority of patients about their diagnosis (99%), treatment plan (98%), and course of disease (93%), but only 38 (55%) about life-prolonging treatment [Drew here - they mean here that this is what physicians reported about their conversations with patients]. As measured by patient inquiry, 51% of the patients knew about their diagnosis and treatment plan, 43% about their course of disease, and 33% (23 patients) about life-prolonging treatment.... Of the 47 patients who did not state that they knew about life-prolonging treatment, we were aware of patient wishes for end-of-life care in 28 patients; 15 patients preferred comfort care, and 13 patients wished to prolong their lifetime.(They are vague about what exactly they mean by 'about life-prolonging treatment.' One assumes, as this is the only thing that makes sense in the overall context of the paper, is that they mean talking about LPT as LPT, framed that way, perhaps [but again not clear] as something that is 'optional'. Strictly speaking one could interpret this as a any conversation about antibiotics as being about LPT and I don't think that's what they mean.....)
They then looked at the 51 patients who clearly articulated treatment preferences as they die (about 2/3 saying they'd want no life-prolonging treatments). Those who expressed more life-prolonging goals were more likely to diverge from their physicians' opinions about what the treatment goals should be (90% of physicians supported treatment limitations in the 'comfort care' group whereas only about 50% of physicians supported life-prolonging goals in those patients who endorsed those goals). Patients who wanted life-prolonging treatments were also less likely to be involved in a discussion about DLT (38% vs 75% for those with 'palliative' goals). Age, performance status, and duration of disease did not impact whether someone had had a discussion about DLT.
The authors express some surprise (if I'm reading them right) about the divergence of opinion, suggesting they thought they'd find doctors more commonly wanting life-prolonging treatments for their dying patients than they actually found. What I'll say to that is that it looks like in those circumstances the physicians were actually talking with patients about things, and perhaps were more likely to be accepting of their treatment goals. That said, there are enough vagaries about how the data were presented here that I'm not really sure overall what it is they were actually describing.
When I read this paper's abstract the first thing that went through my mind was relating this to the Coping with Cancer Study, and my musing a while back (we've written enough about the CWCS that I can't easily find that post) about what was going on with those patients who reported not having an EOL discussion with their physician (and who were more likely to spend time in an ICU prior to death, etc.): did the 'not having a discussion' lead to 'more aggressive' care as they died, or did the whole phenomenon point to something else (e.g. one scenario being patients who were more likely to want 'aggressive' care as they died being more reluctant to discuss death, the docs somehow read that, and therefore are less likely to discuss that with their patients in the first place)? With this paper, I was wondering if it would lead some credence to that speculation, and while it doesn't undermine that theory, I can't say that it provides sufficient info to support it either.
What remains clear, however, is that we (clinicians) remain frequently on 'different pages' than our patients, and do not do a good job of ensuring that they even understand what we're talking about....
Winkler, E., Reiter-Theil, S., Lange-Riess, D., Schmahl-Menges, N., & Hiddemann, W. (2009). Patient Involvement in Decisions to Limit Treatment: The Crucial Role of Agreement Between Physician and Patient Journal of Clinical Oncology, 27 (13), 2225-2230 DOI: 10.1200/JCO.2008.17.9515