“A Tool, Not a Crutch”: Patient Perspectives About IBM Watson for Oncology Trained by Memorial Sloan Kettering

Abstract

PURPOSE: IBM Watson for Oncology trained by Memorial Sloan Kettering (WFO) is a clinical decision support tool designed to assist physicians in choosing therapies for patients with cancer. Although substantial technical and clinical expertise has guided the development of WFO, patients’ perspectives of this technology have not been examined. To facilitate the optimal delivery and implementation of this tool, we solicited patients’ perceptions and preferences about WFO. METHODS: We conducted nine focus groups with 46 patients with breast, lung, or colorectal cancer with various treatment experiences: neoadjuvant/adjuvant chemotherapy, chemotherapy for metastatic disease, or systemic therapy through a clinical trial. In-depth qualitative and quantitative data were collected and analyzed to describe patients’ attitudes and perspectives concerning WFO and how it may be used in clinical care. RESULTS: Analysis of the qualitative data identified three main themes: patient acceptance of WFO, physician competence and the physician-patient relationship, and practical and logistic aspects of WFO. Overall, participant feedback suggested high levels of patient interest, perceived value, and acceptance of WFO, as long as it was used as a supplementary tool to inform their physicians’ decision making. Participants also described important concerns, including the need for strict processes to guarantee the integrity and completeness of the data presented and the possibility of physician overreliance on WFO. CONCLUSION: Participants generally reacted favorably to the prospect of WFO being integrated into the cancer treatment decision-making process, but with caveats regarding the comprehensiveness and accuracy of the data powering the system and the potential for giving WFO excessive emphasis in the decision-making process. Addressing patients’ perspectives will be critical to ensuring the smooth integration of WFO into cancer care.

Who is afraid of black box algorithms? On the epistemological and ethical basis of trust in medical AI

“…Indeed, the degree of reliability of Watson for Oncology at the Memorial Sloan Kettering Cancer Center is different from those used in the Rigshospitalet in Copenhagen, as has been extensively discussed. [36][37][38] As for additional reliability indicators, we believe that transparency, understood as a process that informs the inner workings of black box algorithms is the type of right indicator that will contribute to the overall reliability of algorithms. As suggested before, transparency by itself is necessary, although not sufficient for entrenching the reliability of black box algorithms and the overall trustworthiness of their results.…”

Artificial Intelligence in Decision-Making for Colorectal Cancer Treatment Strategy: An Observational Study of Implementing Watson for Oncology in a 250-Case Cohort

“…Many who were glad to accept WFO as a resource to provide oncologists with cutting-edge medical research and knowledge believed the ideal model of such tools in clinical practice is to be used as “a tool, not a crutch” ( 21 ). By addressing such perspectives, we wish to facilitate the use of WFO and other decision support tools, to help realize the promise of more effective clinical and precision healthcare.…”

<p>Progress of Artificial Intelligence in Gynecological Malignant Tumors</p>

“…Similar to the use of WFO, medical personnel should regard AI as “a tool, not a crutch”. 66 If AI is properly used and its applications in clinical practice are optimized, 67 it will be regarded as a valuable tool. Proper use requires AI to be only in the position of a complement to the doctor’s work, not a replacement.…”