Respecting Patient’s Wishes—How an Electronic POLST Tool Can Drive Goal-aligned Care (FR435)

Abstract

•Describe the key principles to work effectively with your system’s Information Technology specialists to design an electronic POLST tool.•Discuss applying lessons from the experience of deploying an electronic POLST tool across multiple states.•Discuss how clinical decision support can help avoid administering unwanted treatments and aid delivery of goal-aligned care. Honoring a patient’s treatment preferences and priorities are key features of providing goal-aligned care. Having a clear record of a person’s choices related to life-sustaining treatments is particularly valuable. Forty-seven states have developed or are developing a Physician Orders for Life-Sustaining Treatments (POLST) form to unambiguously convey treatment plans that reflect a patient’s wishes to either receive or decline critical interventions. While conceptually straightforward, access to POLST forms and their utility are challenged by the information being static, a “snapshot” of a patient’s preferences, and the lack of interface to the rest of the EHR. In 2012, we created ePOLST, an integrated EHR-based tool for the electronic completion of POLST forms. In 2016, we improved upon the ePOLST tool leveraging EHR clinical decision support to alert acute care clinicians of potential discrepancies between orders on a patient’s POLST form and inpatient orders. Specifically, we developed a real-time clinical alert for providers of patients with POLST-prescribed ‘Do not attempt CPR’ (DNAR) status. In a Providence-affiliated ED or hospital, if a provider attempts to write a ‘Full Code’ order for a patient with an ePOLST order of DNAR, the provider is alerted to the discrepancy prior to signing the order. In the first 18 months, 16,570 ePOLST forms were generated across five states; 52% (8,548) included DNAR status, and 14% (2,311) also opted for comfort measures only. In patients with an ePOLST indicating DNAR, the alert was triggered approximately 200 times per month. Fifteen percent of the time, the ordering provider removed the apparently conflicting ‘Full Code’ status order and wrote an alternative code status order instead. This session will explore the principles and resources necessary to design and implement an ePOLST system. Updated data and detailed outcome analyses of the ePOLST clinical alert will be presented.