Five simple factors: a 'no' answer to the surprise question, older age, decreased serum albumin, presence of dementia, and presence of peripheral vascular disease (blockage of an artery that leads to an arm or a leg), could be mathematically combined to accurately predict that a patient is unlikely to survive past six months. When comparing a patient who died within six months with one who remained alive, 87% of the time the model accurately predicted that the former patient had a higher risk of dying within that timeframe than the latter. The researchers validated their model by testing its accuracy in another 514 kidney disease patients on dialysis, where the model's predictive accuracy was only slightly lower (80%).

Discussing a kidney disease patient's likelihood of dying can help seriously ill patients and their families make informed clinical decisions: some will decide to stop dialysis and start hospice care, while others may prefer continuing vigorous treatments to prolong life as long as possible. "Terminal care is complicated and it is always preferable if decisions can be discussed in advance, goals established, and decisions reached collaboratively between patient and physician," said Dr. Germain.

Source: American Society of Nephrology