Tuesday, January 27, 2009

Care coordination and quality...

1)
Health Affairs has an article from the Dartmouth Atlas people looking at inpatient care intensity and patients' rating of their hospital experiences. Like much of the DA research it looks at administrative Medicare data (specifically regional variations on spending, intensity of care such as hospitalizations, physician utilization, etc.), and using a mortality-follow back design (patients selected for the study had already died). This paper looks at how patient satisfaction ratings compare with that and finds, consistent with other DA research, that more is not better. Patients in the highest utilization regions (for some reason there is a huge spike in the Kansas City region, and given how Christian is talking lately on the blog about how he loves to reverse DNR orders I figure it's got to be due to him**) had the lowest satisfaction (regarding aspects of their hospital care like physician and nurse communication, pain control, etc.).

Hospice use at the end of life did not differ regionally (30%) however no data is presented about length of hospice enrollment. By itself these sort of data don't mean much - maybe there are regional variations in how whiny and dissatisfied patients are and how demanding they are of medical care (or other regional confounders). However this is part of a larger pattern that has emerged (from the work of the DA people and others) that more (more expensive, more intense, more hospital-focused) care for the sickest patients is plainly worse care, however you measure it: mortality outcomes, quality measures, and patient satisfaction. And hell if 'more care' doesn't do any of those things then what are we doing? Interestingly the authors speculate at the end that the 'causal pathway' in these high-utilization regions leading to worse outcomes is due to lack of 'care coordination'...

2)
...Which brings us to the next paper suggesting this phenomenon in a specific circumstance. It's from Archives of Internal Medicine and looks at continuity of care in cancer patients and how it affects ICU use at the end of life. It's a retrospective study using administrative data (the SEER database and Medicare data) to look at associations with care continuity and ICU usage in ~28000 advanced lung cancer patients who died in the US between 1992 and 2002, and who were hospitalized in the last 6 months of life. The key aspect they looked at was if their 'usual care provider' (essentially a physician who saw the patient at least 3 times in the year prior to the index hospitalization) actually saw the patient in the hospital (they looked at the patients' final hospital stay prior to death), and if that was associated with differences in ICU use.

Across the 10 years of the study out- to in-patient continuity declined (the usual care provider seeing the patient in the hospital) from ~60% to ~50%. Many of the usual suspects in care discontinuity were found to be associated with an ICU stay: academic hospital, large metro areas, non-White race. Specifically looking at care continuity they found that it was, in fact, 'protective' against an ICU stay in the final hospitalization (19% vs. 23% of patients without continuity - in the multivariate analysis odds of ICU stay were reduced by 25% with continuity). Really interesting was the very strong association of being cared for by a hospitalist (defined in the study as a doc with over 50% of their billings being inpatient based) with having an ICU stay. The authors speculate that this association was explained by lack of care continuity.

All of this has me speculating about the role of palliative care teams in this. While this is somewhat of a free-association based on my own observations of what I frequently do on the inpatient consultation service, it seems that what we often are doing (or trying to do) is essentially a 'work-around' for this lack of continuity by providing lots of care coordination (how many of you frequently do consults in which your major intervention was getting everyone to talk to everyone at the same time?) as well as trying to provide some of the more ephemeral aspects of care that continuity is supposed to provide (trust, patient-centered care based in a patient's values and goals, etc.).


**This is a joke. Christian, I'm sure, does wantonly waste medical resources but probably not in this way. Kansas City is in the middle; Newark, LA, Miami, Manhattan, and Chicago are at the top.

Thanks to Drs. David Weissman for the Health Affairs piece.

Pallimed | Blogger Template adapted from Mash2 by Bloggermint