Fig. 3: Comparison of MCTs to SYMPLIFY.

Shown is a subset of 40 hypothetical MCTs from the experiment randomly drawn from the full set of 2048 and the three tests in clinical practice: CA-125, FIT, and PSA. Each row shows an individual simulated MCT with its features listed on the y-axis. For each, the predicted probability of choosing that MCT over the SYMPLIFY MCT is presented. Purple bars are MCTs that are preferred to the SYMPLIFY MCT; blue bars denote that the SYMPLIFY MCT is preferred; pink and orange bars compare tests used in clinical practice to the SYMPLIFY MCT. The top row, for example, shows an MCT that: takes 1–2 weeks to return its results, does identify the cancer site, requires a faecal sample, does detect early stage cancer, tests for 25 different cancers, has a PPV of 80%, and a NPV of 99.9%. For GPs, there is around a 77% probability that this test would be chosen over the SYMPLIFY MCT; for patients, there is around a 52% chance that this test would be chosen over the SYMPLIFY MCT.