Earlier versions of this inexplicably EXCLUDED patients from the denominator who were ever diagnosed with depression or bipolar disorder. In attempting to address this issue, the measure steward changed the denominator to ALL patients, except those with a history of bipolar disorder. What they should have done was expand the denominator to ALL patients, except for those with an active problem of depression or bipolar disorder - as that would exactly meet the measure intent.                                                                                   

Change the measure logic to fit the measure intent.  
1. Exclude patients with an active problem of depression or bipolar disorder. Patients with a known condition don't need to be screened for a condition, they need to be followed for it.
2. Where a patient is newly diagnosed with depression and an appropriate follow-up is done and documented, have that follow-up count for the entirety of the measure year.  That is the approach that is currently being used for BMI and follow-up, and Tobacco status and cessation activity.  This website has published this logic for those modifications (paraphrasing) - if a patient has a BMI out of range and is counseled, educated, and referred (or if a current smoker is counseled, treated, or referred), one can reasonably assume that that follow-up applies not just to the day it was documented - but that it continues throughout the measure period.  It would be unreasonable to expect a referral to a nutritionist (for example) at every visit.