​Response 1: The measure looks to see if a PCI procedure was performed using the Procedure, Performed: PCI data element (PCI Grouping Value Set 2.16.840.1.113762.1.4.1045.67). As long as there is an end dateTime (high value) which demonstrates the procedure was actually performed, then it looks at the start dateTime (low value) for timing. The start dateTime equals the time the procedure begins which should be one of the following (whichever is first): (1) time of the first balloon inflation, or (2) time of the first stent deployment, or (3) time of the first treatment of lesion with another device (Time thrombectomy device used, Time of aspiration, Time of suction, Time of device pass, Excimer time, Laser time, Time Rotablator used). Response 2: "Procedure, Performed not done: Patient Refusal" and "Procedure, Performed not done: Medical Reason" were removed in version 4 because a PCI must be performed for the case to fall into the denominator. If a PCI was not done, the case would not pass the denominator so the execution of the logic would stop, thereby never getting to the exception logic. The exception element guidance has also been removed in version 5. ​

​The intent of the AMI-8a measure is to capture STEMI AMI patients that received a primary PCI within 90 minutes of hospital arrival time. We are assessing the feasibility of creating a denominator exception for initial patient/family refusal of PCI/reperfusion/cath/transfer to cath lab for those cases where the patient initially refused but changed their mind and the PCI occurred within the first 1440 minutes of arrival time. With the current logic, cases that had a PCI procedure within the first 1440 minutes, but greater than the 90 minutes from arrival time, would fail the numerator, negatively impacting the provider’s outcome rate. By creating a denominator exception within the logic, the provider would not experience a negative outcome for delays outside their control.   While the number of patients who initially refuse, later change their mind, and receive a PCI within the first 24 hours is anticipated to be low, this exception is not currently captured within the logic.   The paper-based specifications from the Data Dictionary v4.4a, page 282 (I-290), data element criteria for Reason for Delay in PCI states:   Documentation must be made clear somewhere in the medical record that (1) a “hold,” “delay,” “deferral,” or “wait” in doing PCI/reperfusion/cath/transfer to cath lab actually occurred, AND (2) that the underlying reason for that delay was non-system in nature.   Examples of ACCEPTABLE documentation for the paper-based measure: · “Hold on PCI. Will do TEE to rule out aortic dissection.” · “Patient waiting for family and clergy to arrive - wishes to consult with them before PCI.” · “Thrombectomy catheter did not cross lesion. Balloon catheter successfully crossed the stenosis. Flow reestablished after 30 min. delay.” · “PCI delayed due to intermittent hypotensive episodes when crossing lesion.” · "Hold PCI. Need to consult with neurology regarding bleeding risk." · “Cath initially deferred due to shock.”   Electronic information is documented in discrete data fields or otherwise codified to be retrieved within the EHR. To align with guidance provided in the paper-based measure specifications, this would be achieved through a value set that uses codes to capture the above concepts. Therefore, we suggest revising the AMI-8a logic to: 1) Add Denominator Exception logic 2) Create a ‘Reason for Delay in PCI’ value set   We would appreciate vendor/implementer feedback on these proposed changes. https://ecqi.healthit.gov/eh/ecqms-2016-reporting-period/primary-pci-received-within-90-minutes-hospital-arrival