Feasibility of implementing a decision aid for advanced cancer patients in a comprehensive cancer centers' outpatient...

Decision aids (DAs) have been proposed as tools to empower patients in decision-making. However, implementing DA for decisions in advanced cancer is challenging. This study focuses on perspectives of oncologists and other healthcare providers on the hindering and facilitating factors for implementing a customizable DA. This DA is designed to support discussions about whether patients with advanced, incurable cancers should pursue or forgo further anticancer treatment.

This qualitative study utilized content analysis following Sandelowski's approach on two different datasets derived from research projects on decision-making in advanced cancer. To inform the development of a DA and understand challenges related to its implementation, we conducted semi-structured interviews with 24 oncologists (Dataset 1) and three focus groups with various healthcare professionals (n = 19) and in addition post-implementation semi-structured interviews with oncologists who used the DA in conversations with their cancer patients (n = 5) (Dataset 2). Furthermore, the insights from process evaluation of developing and implementing the DA in an outpatients' settings was incorporated, enriching the analytical framework.

Two overall themes emerged: (1) Tension between standardizing the use of a DA and responding to individual patient needs. To address this conflict, a two-part DA was developed, combining structured elements with flexibility. (2) Prerequisites for the use of the DA in outpatients' settings: Senior physicians' support, education in palliative care options and in the use of the DA, and organizational conditions.

Oncologists identified both structural and content-related aspects for a successful implementation of a DA for patients with advanced cancer. While the content-related aspects were factored into the development of the DA, structural (organizational) issues need to be especially focused on during implementation.