Probably, digging up old discourse here, but I really really really do not trust any QALY measurement that doesn’t allow for negative values. If you’re rating a year of depression at 0.6 times the value of a non-depressed year, something in your model has probably gone off the rails. I mean seriously, even if a person spends an entire year in a nursing home in constant pain, barely able to sit up, and isolated from all the friends and family they used to see every day, the people claiming to be making rigorous utilitarian calculations will STILL insist on expressing that as a non-negative percentage value relative to a year of health, no matter how much the patient insists that they would rather be dead.

I wish this was just an isolated methodological error, but sadly this seems to be the norm for everything when we talk about happiness economics, life satisfaction, and healthcare.