Use of Machine Learning and Lay Care Coaches to Increase Advance Care Planning Conversations for Patients With Metastatic Cancer

Abstract

PURPOSE: Patients with metastatic cancer benefit from advance care planning (ACP) conversations. We aimed to improve ACP using a computer model to select high-risk patients, with shorter predicted survival, for conversations with providers and lay care coaches. Outcomes included ACP documentation frequency and end-of-life quality measures. METHODS: In this study of a quality improvement initiative, providers in four medical oncology clinics received Serious Illness Care Program training. Two clinics (thoracic/genitourinary) participated in an intervention, and two (cutaneous/sarcoma) served as controls. ACP conversations were documented in a centralized form in the electronic medical record. In the intervention, providers and care coaches received weekly e-mails highlighting upcoming clinic patients with < 2 year computer-predicted survival and no prior prognosis documentation. Care coaches contacted these patients for an ACP conversation (excluding prognosis). Providers were asked to discuss and document prognosis. RESULTS: In the four clinics, 4,968 clinic visits by 1,251 patients met inclusion criteria (metastatic cancer with no prognosis previously documented). In their first visit, 28% of patients were high-risk (< 2 year predicted survival). Preintervention, 3% of both intervention and control clinic patients had ACP documentation during a visit. By intervention end (February 2021), 35% of intervention clinic patients had ACP documentation compared with 3% of control clinic patients. Providers' prognosis documentation rate also increased in intervention clinics after the intervention (2%-27% in intervention clinics, P < .0001; 0%-1% in control clinics). End-of-life care intensity was similar in intervention versus control clinics, but patients with ≥ 1 provider ACP edit met fewer high-intensity care measures ( P = .04). CONCLUSION: Combining a computer prognosis model with care coaches increased ACP documentation.

Machine Learning to Allocate Palliative Care Consultations During Cancer Treatment

“…Related studies demonstrate that ML systems can be paired with interventions to improve outcomes for patients with cancer. 18,19 For example, Manz et al 19 conducted a stepped-wedge randomized clinical trial of an ML system that predicted 6-month mortality providing behavioral nudges to clinicians, leading to an increase in serious illness conversations and a decrease in systemic therapy near the end of life relative to controls. Of note, these studies demonstrate the feasibility of implementing complex ML-based prognostic systems within the electronic medical record and clinical care.…”

“…[13][14][15][16] ML predictions of prognosis outperform clinicians 17 and prompts from prognostic ML systems can increase the frequency of serious illness conversations. 18,19 To our knowledge, no study to date has evaluated whether ML could improve the allocation of PC resources.…”

Precision Palliative Care as a Pragmatic Solution for a Care Delivery Problem

“…6,43 Successful work in behavioral science has already incentivized earlier referral to specialty PC, 11 hospice use, 41 and better conversations for seriously ill patients. 29,30,44 In conclusion, specialty PC has a major role in cancer care delivery, although it must be deployed in a staged and sustainable fashion, targeted to individual and caregiver needs. As it now stands, current clinical guidelines for specialty PC referral apply a one-size-fits-all framework to often nuanced levels of PC need.…”

“…29 An analogous mortality prediction model–based intervention at Stanford demonstrated improvement in advance care planning and documentation of prognosis. 30 At Princess Margaret, in the STEP trial, targeted PC referral was implemented according to an algorithm based on symptom severity. 9 Forty percent of patients never screened positive and QOL, symptom control, and mood remained stable over time demonstrating early PC was not necessary for those with mild symptoms.…”

“…6,43 Successful work in behavioral science has already incentivized earlier referral to specialty PC, 11 hospice use, 41 and better conversations for seriously ill patients. 29,30,44…”