An Actionable Expert-System Algorithm to Support Nurse-Led Cancer Survivorship Care: Algorithm Development Study

Abstract

Background Comprehensive models of survivorship care are necessary to improve access to and coordination of care. New models of care provide the opportunity to address the complexity of physical and psychosocial problems and long-term health needs experienced by patients following cancer treatment. Objective This paper presents our expert-informed, rules-based survivorship algorithm to build a nurse-led model of survivorship care to support men living with prostate cancer (PCa). The algorithm is called No Evidence of Disease (Ned) and supports timelier decision-making, enhanced safety, and continuity of care. Methods An initial rule set was developed and refined through working groups with clinical experts across Canada (eg, nurse experts, physician experts, and scientists; n=20), and patient partners (n=3). Algorithm priorities were defined through a multidisciplinary consensus meeting with clinical nurse specialists, nurse scientists, nurse practitioners, urologic oncologists, urologists, and radiation oncologists (n=17). The system was refined and validated using the nominal group technique. Results Four levels of alert classification were established, initiated by responses on the Expanded Prostate Cancer Index Composite for Clinical Practice survey, and mediated by changes in minimal clinically important different alert thresholds, alert history, and clinical urgency with patient autonomy influencing clinical acuity. Patient autonomy was supported through tailored education as a first line of response, and alert escalation depending on a patient-initiated request for a nurse consultation. Conclusions The Ned algorithm is positioned to facilitate PCa nurse-led care models with a high nurse-to-patient ratio. This novel expert-informed PCa survivorship care algorithm contains a defined escalation pathway for clinically urgent symptoms while honoring patient preference. Though further validation is required through a pragmatic trial, we anticipate the Ned algorithm will support timelier decision-making and enhance continuity of care through the automation of more frequent automated checkpoints, while empowering patients to self-manage their symptoms more effectively than standard care. International Registered Report Identifier (IRRID) RR2-10.1136/bmjopen-2020-045806

User-Centered Design and Usability of a Culturally Adapted Virtual Survivorship Care App for Chinese Canadian Prostate Cancer Survivors: Qualitative Descriptive Study

“…User inputs to the questionnaire were triaged via a decision-tree algorithm [ 45 ]. The algorithm was designed to return in-app self-management resources within a progress note (“Nurse’s Note”) automatically available to the user after input submission.…”

A Decision Support Application for Patients with Cancer: A Feasibility Study

“…Furthermore, a rule-based survivor algorithm based on expert information as a decision-support tool was developed as a prostate cancer (PCa) survivor care model. This is expected to enable patients to selfmanage their symptoms while setting automatic checkpoints more effectively for timely decision support and enhancing the continuity of care [4]. Furthermore, 73% of breast cancer survivors responded that a decision aid (DA) for navigating breast cancer survivorship care was useful for decision making, indicating that the developed DA information structure can be expanded and applied to other care models and that arti cial intelligence can be utilized to optimize information delivery [5].…”