Dear Editor:The goal of advance care planning (ACP) is for patients to communicate their end-of-life (EoL) treatment preferences to a selected proxy. However, it is not clear what information must be shared to adequately inform a proxy. Given the practical challenges of measuring ACP conversations, previous studies have focused on measuring ACP surrogates such as code status documentation or completion of advance directives (ADs) and/ or physician orders for life-sustaining treatment (POLSTs). 1,2 However, it has been demonstrated that AD completion alone does not necessarily promote high-quality EoL communication or understanding between patients and proxies. 3 The aim of this pilot study was to determine the feasibility of completing a focused ACP conversation identifying an informed proxy in a single clinic visit. This study was completed in adult cancer patients with a prognosis of less than one year. A clinical social worker led the ACP intervention between study patients and proxies focusing on three EoL preferences: (1) the patient's personal definition of quality of life, (2) his or her specific plan if he or she cannot achieve this quality of life, and (3) desired location of death. The proxy was deemed ''informed'' if he or she understood these three EoL preferences. Patients were encouraged but not required to complete an AD/POLST and followed until death.Thirty-five patients were screened and 34 patients were available for the analytic sample (Table 1). Eighty-two percent (n = 28) of proxies were ''informed'' following the intervention, and 65% (n = 22) completed the intervention in a single visit. Following the intervention, 54% completed a new AD (n = 15) and 9% (n = 3) completed a POLST. There was a statistically significant increase in AD/POLST completion ( p < 0.001). For those patients that died (n = 31), there was 81% (n = 25) and 61% (n = 19) concordance of desired and actual code status and location of death, respectively. Neither concordance was significantly different based on the completion of an AD/POLST ( p = 0.34 and p = 0.27). These rates are higher than those demonstrated in prior studies, 1 including our institutional historical rate. 4

PURPOSE: Advance care planning (ACP) is a clinical skill that can be taught. An opportunity exists to teach how to conduct ACP to clinicians not typically engaged in these conversations to increase the likelihood that patients and caregivers engage in ACP. We conducted a prospective study exploring the feasibility of a pharmacist-led ACP intervention. METHODS: We completed a prospective, single-center study from July 2015 to July 2017. We included patients of age ≥ 18 years with incurable cancer referred to the palliative care clinic. A trained pharmacist led an ACP discussion with the patient and selected proxy. We defined feasibility as completion of ≥ 30 pharmacist-led ACP discussions over the study period. Additionally, we defined an informed healthcare proxy as someone who understood three key end-of-life (EOL) treatment preferences: the patient's personal definition of quality of life, desired resuscitation status, and preferred location of death (in or out of the hospital). Patients were followed until the end of the study or death. For those patients who died, the pharmacist contacted the proxy for follow-up and explored satisfaction with the ACP intervention. RESULTS: Thirty-four patients completed the study. All selected proxies completed the intervention and were able to understand the three EOL preferences. At the time of the patient’s death (n = 20), proxies reported that 66.6% received their preferred resuscitation status and 72.2% died in their preferred location. Proxy satisfaction with the ACP process was 7.6 ± 2.5 (mean ± SD) on a 11-point Likert scale. CONCLUSION: These findings indicate the potential for pharmacists to lead and engage in ACP in the outpatient setting.

136 Background: To increase the probability that cancer patients receive care consistent with their wishes, medical providers require simple, practical, and effective communication tools to initiate and complete meaningful advance care planning (ACP) discussions. Ideally, ACP discussions occur in the non-emergent setting with a selected health care proxy and trusted medical provider and may lead to completion of an advance directive (AD). The primary aim of this study is to determine the proportion of advanced cancer patients who identify an informed health care proxy implementing a novel ACP tool. Methods: In an American Cancer Society funded pilot study, advanced cancer patients were evaluated in an academic oncology palliative care clinic (n=35). Subjects engaged in a 1-hour ACP intervention completed by a licensed clinical social worker utilizing a novel ACP tool developed by the investigators. The ACP tool identifies a health care proxy and defines three key elements an informed health care proxy must know. Details of this ACP conversation were documented in the electronic medical record with the goal of completing an AD. After subject’s death, health care proxies were contacted to determine if end-of-life wishes were honored. Results: 35 subjects (51% woman, 71% Caucasian, 54% married) were enrolled with a mean age ± SD of 57.4 ± 14.1 years with gastrointestinal as the most common primary cancer (13/35). The ACP intervention was completed in 80% in 1-2 clinic visits. After the ACP intervention, 94% of subjects (33/35) identified an informed health care proxy. The most common identified health care proxies were either a spouse (17/35) or child (11/35). 43% of subjects (15/35) completed an AD after the ACP intervention. As of July 2014, 69% (24/35) of subjects had died and 71% of these (17/24) died in a setting consistent with their end-of-life wishes. Mean time to death from ACP intervention was 4.5 ± 3.2 months. 11 subjects remain in surveillance. Conclusions: Results suggest that this novel ACP tool facilitates successful identification of an informed health care proxy. Despite low AD completion, death consistent with end-of-life wishes was achieved in the majority of patients.

No abstract