Outcomes of Prognostic Disclosure: Associations With Prognostic Understanding, Distress, and Relationship With Physician Among Patients With Advanced Cancer

Abstract

Purpose: To determine how prognostic conversations influence perceptions of life expectancy (LE), distress, and the patient-physician relationship among patients with advanced cancer.

Patients and methods: This was a multicenter observational study of 590 patients with metastatic solid malignancies with progressive disease after ≥ one line of palliative chemotherapy, undergoing follow-up to death. At baseline, patients were asked whether their oncologist had disclosed an estimate of prognosis. Patients also estimated their own LE and completed assessments of the patient-physician relationship, distress, advance directives, and end-of-life care preferences.

Results: Among this cohort of 590 patients with advanced cancer (median survival, 5.4 months), 71% wanted to be told their LE, but only 17.6% recalled a prognostic disclosure by their physician. Among the 299 (51%) of 590 patients willing to estimate their LE, those who recalled prognostic disclosure offered more realistic estimates as compared with patients who did not (median, 12 months; interquartile range, 6 to 36 months v 48 months; interquartile range, 12 to 180 months; P < .001), and their estimates were less likely to differ from their actual survival by > 2 (30.2% v 49.2%; odds ratio [OR], 0.45; 95% CI, 0.14 to 0.82) or 5 years (9.5% v 35.5%; OR, 0.19; 95% CI, 0.08 to 0.47). In adjusted analyses, recall of prognostic disclosure was associated with a 17.2-month decrease (95% CI, 6.2 to 28.2 months) in patients' LE self-estimates. Longer LE self-estimates were associated with lower likelihood of do-not-resuscitate order (adjusted OR, 0.439; 95% CI, 0.296 to 0.630 per 12-month increase in estimate) and preference for life-prolonging over comfort-oriented care (adjusted OR, 1.493; 95% CI, 1.091 to 1.939). Prognostic disclosure was not associated with worse patient-physician relationship ratings, sadness, or anxiety in adjusted analyses.

Conclusion: Prognostic disclosures are associated with more realistic patient expectations of LE, without decrements to their emotional well-being or the patient-physician relationship.

Fig 1.

Patients' self-estimates of life expectancy versus actual patient survival according to patients' recall of whether their physician disclosed prognosis (n = 299). Data are presented with Tukey box-whisper plots, where boxes represent medians and interquartile ranges (IQRs), and whispers represent data within 1.5× IQR; X indicates mean.