What is a quality-adjusted year of life for a dialyzor?

RenalWEB links to an article from the Wharton School of the Univ. of Pennsylvania and a related bloq post from The Huffington Post that comments on and analyzes a new paper that looks to update the renal dialysis cost effectiveness standard. RenalWEB even offers a rare, blog-like commentary:

Dialysis patients whose treatment costs exceed $240,000 per QALY [quality-adjusted year of life] fall in the top 10%. The study points out that 50% of the Medicare's  ESRD costs are used to treat the "sickest" 10% of dialysis patients.

The question is if and/or when this kind of cost data will be used to determine a patient's access to dialysis treatments. The study will appear in an upcoming issue of Value in Health.

I just can't explain these numbers. The coverage is tantalizing but it's clear I don't understand the key points. From the Knowledge@Wharton article:

The study results show that the incremental cost effectiveness ratio of dialysis in current practice relative to the next least costly alternative is on average $61,294 per year, or $129,090 for a quality-adjusted year of life (QALY) -- a measure that combines the length of time that life is extended and the quality of that life.

However, the distribution of cost effectiveness across the entire population is wide. For the lowest percentile, it costs only $65,496 to provide an additional quality-adjusted life year. For the top percentile, the figure is $488,360. The higher costs per quality adjusted life year were strongly associated with old age and additional chronic illnesses in addition to end-stage renal disease, the researchers found.

I am afraid I can't make sense of this I am familiar with the USRDS numbers. As I would normally describe it, total Medicare spending for someone on Dialysis (not in a SNF) is about $55,000/year, since this is 80%, that puts total average total cost around $69,000. However the lowest percentile must be quite a bit less.

Taking myself as an example, my "cost" in 2007 it was almost entirely the composite rate, along with below average use of injectable medications (10% of average EPO use) and low use of medications that typically fall under Part D. I haven't added it all up but it must be in the $15,000 to $20,000 range when including physician fees. So, at least for 2007, I reckon I'm a bargain and would be in the lowest percentile. Where do I fit into the data?

If commenters want to try and explain where the numbers come from I'd be obliged - I've emailed the author too, I'll share anything I learn.