My title alludes to the tendency of nurses (and other non-medicine disciplines?) to prefer the more inclusive terms “evidence-based care” or “evidence-based practice.” Recently, my eye caught some rather disparate articles from the quality literature. (Drew blogged on one of them from JAMA last week, so I won’t dwell on it.)1)
The first, a NEJM Sounding Board article, emphasized the need for an evidence base prior to instituting changes intended to improve safety and/or quality. The title delivers much of the message: “The tension between needing to improve care and knowing how to do it.” The article posed 7 arguments (from the quality literature) for accelerating change, and then adroitly addressed each with a counter argument counseling caution. The missing evidence in proposals for change (including a well-known one—cutting the hours of house staff) fall into three obvious categories that are constantly overlooked, side-stepped, or blown off: efficacy, possible harms/side effects, strategies for implementation and post-change evaluation. Their closing statement: “Just as in the rest of medicine, we must pursue the solutions to quality and safety problems in a way that does not blind us to harms, squander scarce resources, or delude us about the effectiveness of our efforts.”
2)
Here’s a BMC journal I have not previously seen: Health and Quality of Life Outcomes. The specific article that led me there is “Quality of Life at the End of Life,” from a group at the University of Washington and Bastyr University. They followed over 100 hospice patients for several months to determine patient-perceived QOL over time. The statistics in such studies make my head spin, but basically they found that QOL was better than expected until the last 3 weeks of life, and that in those 3 weeks older patients did worse than younger patients. This is an open access journal, so those of you with the patience to wade through may find the methodology and discussion of some use.
3)
Bioethicist Eric Loewy, MD has an opinion piece in MedGenMed, a Medscape electronic journal, asking if evidence-based medicine and ethics are in conflict with each other. The quality of the writing is a bit disappointing, but the overall messages are pointed: EBM conceptually is great, but it is often misused. Specifically, generating EBM-based guidelines or protocols are excellent learning experiences, but then they become anti-intellectual when followed unquestioningly as shortcuts around thinking, or when applied blindly as if all patients are the same. The harshest criticism is aimed at HMO’s which use EBM-based protocols as bludgeons to control physician prescribing, going so far as to label such organizational practices “practicing medicine without a license.” The ethical issue here is that HMO-mandated guidelines are too often driven by financial considerations, the polar opposite of being patient-centered. There’s little new here in terms of criticism of EBM, except perhaps the emphasis on misuse of EBM being unethical.
4)
A companion article to one that Drew mentioned on last week (on the life-span of published systematic reviews) may be “How evidence-based are
the recommendations of evidence-based guidelines?” in the open access journal PloS Medicine. The authors selected prominent guidelines relating to cardiovascular treatments. They used multi-domain standardized instruments to rate the quality of the evidence used in the guidelines. They found that only 28% (range: 21 – 41% across guidelines) of the 369 recommendations were supported by high quality evidence. A major failing was the extrapolation from the “highly selected” research subject samples, to the larger, much more heterogeneous target population for the recommendations. 
3 comments:
There are an astonishing variety of BMC journals. I just came across one the other day: BMC Medical Informatics and Decision Making.
After reading thru the BMC article I see what you mean Tom. I have this rule that I'm highly suspect of articles which have seemingly simple premises (change in qol over time as people approach death) but then have lengthy methods with multiple data models etc etc. My sense from reading this paper is that these are qol researchers talking to other qol researchers and us curious clinicians aren't the target audience!
What I liked about the article is that they used, as one of their measures, simple qol measures (what is your quality of life?) as opposed to health-related qol tools which are difficult to interpret in dying patients. Dying patient's 'health related quality of life' (which, when you look at HRQOL instruments, seems to be a lot like functional status/measuring daily limitations due to one's disease) is by definition bad but it seems reasonable that many people may continue to rate their overall QOL, a more elusive and personal entity, as fine right up until the end (or, as this study suggests, until 3 weeks before death).
I'm not sure if we are witnessing an EBM backlash but we may be witnessing the realization from many quarters that the aspirations of EBM (to which I subscribe) are far from being realized and it's WAY too early to begin making & enforcing policy/standards based on shaky evidence. And people are getting really nervous as 'pay for performance' seems to be becoming a reality - that we are going to be asked to do things which aren't necessarily evidence based because The Man has decided we should and we're not going to get paid if we don't. Examples that are being thrown around as misguided: early antibiotics for community acquired pneumonia (quickly became a standard despite very weak evidence), pain as the 5th vital sign (that one was adopted with no evidence and has never been shown to have made a difference - which doesn't mean it's not a good idea - but demonstrates the inadequacy of simplistic policies to actually benefit patients). On the other hand we apparently live in a world in which many patients who are hospitalized with myocardial infarctions get discharged without being put on aspirin so there has to be some balance....
How all of this is going to affect palliative care & the hospice world is yet to be seen, but it's not reasonable to assume we'd be immune from the impulse to rationalize/standardize care based on evidence/best practice.
Thanks for the nice post, Tom.
Thomas,
Great post. I would be interested on how you see all of this affecting palliative care practice. We have so little knowledge about our current treatments when compared to more established fields of medicine, and we see that they are already having difficulties with EBM as you and Drew alluded.
Since we do not have a broad research base to begin with, do you think that palliative medicine should strike out to have higher standards for research or is that already creating a higher entry barrier when research on symptoms, QOL, and dying are scant to begin with?
Personally, I am having a bit of QOL backlash/burnout when it comes to reading research and applying it to clinical practice. Not that it should not be studied but I don't feel that we have the concept adequately understood in research terms a la Drew's comments.
Of course this questions is not just posted to Thomas but to other folks as well.
Christian & Drew,
In aggregate, you two have asked the $64,000 questions about EBP. The examples in the articles as well as Drew's highlighting of the "5th vital sign" point out the lack of sophistication that we have about "evidence." Many of us are little better than lay people. There is also the issue of the stick: some clinicians easily fall under the spell of the "evidence" argument by payers. 8 years later, many clinicians are still convinced that JCAHO 2001 standards (now, simply Joint Commission) mandate the use of 5th vital sign.
Bad use and misunderstanding of something as powerful as EMP may be inevitable, but the concept rings true for almost everyone.
As many have pointed out, the evidence base for much of what happens in EOL care is very weak, and really not getting dramatically better. I think researchers and clinicians need to be better attuned to each other. QOL research is important, but we need simple studies on basic symptom management and psychosocial support interventions. Yes, these studies are hard to power, and yes you get into issues of generalizability, and of course there are issues of time & cost. But almost any hospice or palliative care unit probably has the capacity to conduct prospective and retrospective case series. Long before Joanne Lynn and the quality folks were saying "what can you do by next Tuesday?" Bruera was conducting and publishing small studies on a myriad of clinical problems. We need to get beyond that micro approach eventually, but it is still a viable model, I believe.
Post a Comment