Purpose: Knowing when patients are too ill to benefit from intensive care is essential for clinicians to recommend aggressive or palliative care as appropriate. To explore prognostic ability among critical care fellows, the authors compared fellows' and attendings' assessments of futile critical care and evaluated factors associated with assessments.
Method: Thirty-six attendings and 14 fellows in intensive care units at the University of California, Los Angeles, were surveyed daily for three months (December 2011-March 2012) to identify patients perceived as receiving futile treatment. Frequency of futile treatment assessments and reasons listed by attendings versus fellows were compared. Predictors of futile treatment assessments by provider type were assessed using multivariate probit models.
Results: Attendings made 6,897 assessments on 1,125 patients; fellows made 4,407 assessments on 773 patients. Fellows assessed 161 (20.8%) patients as receiving futile treatment, compared with attendings (123 [10.9%] patients, P<.001), and listed fewer reasons that treatment was futile (P<.001). Fellows were more likely to assess a patient as receiving futile treatment by the second day, whereas attendings took four days. Patients assessed as receiving futile treatment by fellows were less likely than patients so assessed by attendings to die in the hospital (51% versus 68%, P=.003) and within six months (62% versus 85%, P<.001).
Conclusions: Fellows made earlier assessments and judged more patients to be receiving futile treatment than attendings, and their assessments were less predictive of mortality, suggesting that assessment of treatment appropriateness develops with experience.