RESUMEN
BACKGROUND: Patients' prognostic beliefs are known to influence treatment decisions. However, the evolution of these beliefs over an extended period in patients with metastatic cancer is understudied. We assessed longitudinal changes in prognostic beliefs and investigated their association with patients' changing health status. METHODS: We surveyed a cohort of 600 patients with solid metastatic cancer every 9 months, up to 54 months. At each time point, we assessed whether patients believed their current treatments would cure them (responses classified as accurate, inaccurate, or uncertain belief) and tested the association of their response with symptom burden and recent unplanned hospital admission. RESULTS: Only 29% of patients had accurate prognostic belief at baseline, and 24% of patients changed from having accurate to uncertain/inaccurate belief at some point during follow-up. Patients who experienced greater symptom burden were less likely to report inaccurate (relative risk ratio [RRR], 0.87; 95% CI, 0.84-0.90) or uncertain prognostic belief (RRR, 0.90; 95% CI, 0.87-0.92), whereas those with a recent unplanned hospital admission were more likely to report inaccurate (RRR, 2.71; 95% CI, 1.48-4.94) or uncertain belief (RRR, 2.34; 95% CI, 1.34-4.07) compared with accurate belief. An increase in symptom burden was associated with change toward accurate belief (RRR, 1.75; 95% CI, 1.33-2.31) as opposed to no change. CONCLUSIONS: In our study of long-term changes in prognostic beliefs among patients with metastatic cancer, reported prognostic beliefs were unstable, changed from accurate to inaccurate/uncertain and vice versa, and were associated with their changing health status. Our findings imply that conversations about goals of care must occur regularly to factor in these changes.