Thinking and Talking about Prognosis in Advanced Cancer

Abstract

This thesis aimed to improve how oncologists think and talk about prognosis in advanced cancer by evaluating the usefulness of oncologists’ estimates of expected survival time, studying recommendations about the timing of discussions of prognosis, and evaluating a method for formulating and explaining prognosis.

Data was collected from 8 clinical trials where oncologists' estimates of participant's expected survival times were compared with observed survival times. Oncologists’ estimates were well-calibrated, imprecise, and independently associated with observed survival times. Simple multiples (1/4, 1/2, 2, and 3) of each estimate provided well-calibrated ranges to describe worst-case, typical, and best-case scenarios for survival.

A systematic review summarised research on timepoints when discussions about prognosis should occur. These included: at first consultation; upon disease progression; and, when there were no further anti-cancer treatments. It is important for doctors to identify appropriate timepoints to elicit patients' preferences for the timing and content of discussions about prognosis.

A web-based tool was developed to help oncologists estimate and explain worst-case, typical, and best-case scenarios for survival time. The attitudes of patients, family members, and healthcare professionals to this information, were evaluated in a single-arm clinical trial. Over 84% found it useful to receive information about prognosis formulated as 3 scenarios for survival time. This trial introduced a simple, freely accessible, web-based tool that oncologists could use routinely in their clinical practice.

This thesis provides new information about the: usefulness of oncologists’ estimates of expected survival time; recommended timing of discussions about prognosis; and methods for communicating prognosis. Improving the way oncologists think and talk about prognosis should help people affected by advanced cancer and lead to better informed decision making.