Monday, May 11, 2009

Pall-Pourri: TAL & a couple on prognosis

This American Life (a public radio show) last week included a moving monologue by Dan Savage about his mother's death (you can stream or podcast TAL for free). We previously mentioned Dan Savage's column about his mother's death last year shortly after she first died. This piece expands some of the thoughts from that column and focuses on how her death effected his relationship with his former faith (he was raised Catholic, and his mother died devoutly Catholic), and is also just a deeply moving account of her death, his love for her, and his ongoing grappling with his loss. For those of you who don't know, Dan Savage is a sex-advice columnist, and basically nothing he writes could be considered 'Safe For Work' (even this - I think - I heard the unexpurgated version that was broadcast live to movie theaters last month and they may have edited it a bit for the radio broadcast) so don't go blasting this from your computer in insecure locations.

Mayo Clinic Proceedings has a case-control study looking at the prognostic importance of acute kidney injury in non-(yet)-critically ill hospitalized patients. ('AKI' is a more general, and less threatening, term for what used to frequently be called 'acute renal failure' or 'azotemia' and is defined as a rise of .3mg/dl in serum creatinine within 48 hours.) This study takes a bunch of patients from one hospital who had AKI, matches them with age-matched controls, and compares outcomes. The study generates some good data, but is probably mostly only of interest for prognosis-wonks.

Basically: yes AKI is a profoundly important prognostic indicator for in-hospital mortality (those with AKI do much worse than those without); however taken alone it is probably not anything one can bring to the bedside. AKI was actually the most powerful prognostic indicator in the study - beat out age and number of comorbidities, for instance. Relative risk of in-hospital death was 10 for those with AKI (meaning those with AKI has 10x the risk of death than those without; again these were only age matched controls not co-morbidity matched); in a multivariate model those with AKI had an odds ratio of 8 for death compared to those without along with much longer hospital lengths of stay and transfer to ICUs. Overall hospital mortality was ~15% in those with AKI compared to 1.5% in those without; over 40% of survivors were discharged to other care facilities (not home) compared to 20% of those without AKI. So, bad, yes, but sort of in the category of hypoalbuminemia, hyponatremia, elevated uric acid, etc.: we know that those who have those findings do worse as a group than those who don't in multiple disease states, but doesn't help us much more, as individual findings, than adding to a gestalt impression....


American Journal of Medicine
has a paper about mid-long term prognostication in patients with non-cardiac vascular disease. It's a multicenter Dutch study which looks at ~700 patients undergoing vascular surgery (e.g. endovascular grafting, arterial stenting, carotid endarterectomies, etc.), all of whom underwent preoperative cardiac evaluation and were risk stratified by the Lee Risk Index. The LRI stratifies patients' operative cardiac risk based on the presence of: high-risk surgery, ischemic heart disease, history of congestive heart failure, history of cerebrovascular disease, insulin therapy for diabetes mellitus, and renal insufficiency.
This study follows these patients out for 3 years and looks at whether the LRI predicted survival over that time-frame.

Those with 3 or more risk factors on the LRI had a substantially worse survival at 3 years (hazard ratio 3.3; 3-year mortality of ~40%; most of those who died did so within a year of the index surgery). This is compared to a ~10% mortality at 3 years for those with no risk factors.

While this does not give us helpful short-term prognostic data, this is on par with the sort of data we get from the BODE index for COPD, so I thought I'd mention it. I'm curious as to whether palliative clinicians are seeing PVD/vascular surgery patients in their practice outside of those who are imminently dying: pain consults, goals/care planning, etc.? I see a few from time to time; it's a patient population which by all rights is 'perfect' for palliative care: severe PVD is a chronic, debilitating, highly morbid/symptomatic life-limiting illness.

Pallimed | Blogger Template adapted from Mash2 by Bloggermint