The specs for HH-02 indicate that any diabetes diagnosis before or during the encounter will qualify the patient for this measure. I'd like to suggest limiting it to just a diagnosis present during the encounter, or at least a year prior or something.  If diabetes was a true diagnosis and not an error, it will remain active on every encounter.  We are having issues with people pulling into this measure because there was incorrect coding or an errant diagnosis placed on a chart from a long ago encounter that identified them as incorrectly having diabetes.  It's extremely time consuming to try to track down where this errant code is coming from in the entire patient's history.  As our EHR is shared among several hospitals, once you find it, you may have to try to contact the coding department at another hospital to correct coding from 3 years ago or contact a pharmacist about an errant 'reason for visit' diagnosis code they entered once long ago when a patient came in for education. If these patients really had diabetes, it would be captured again during the current hospital admission the measure is reviewing.