Can Oncologists Prompt Patient Prognostic Awareness to Enhance Decision-Making? Data From the NEOetic Study

Abstract

Introduction: Anti-neoplastic therapy improves the prognosis for advanced cancer, albeit it is not curative. An ethical dilemma that often arises during patients' first appointment with the oncologist is to give them only the prognostic information they can tolerate, even at the cost of compromising preference-based decision-making, versus giving them full information to force prompt prognostic awareness, at the risk of causing psychological harm.

Methods: We recruited 550 participants with advanced cancer. After the appointment, patients and clinicians completed several questionnaires about preferences, expectations, prognostic awareness, hope, psychological symptoms, and other treatment-related aspects. The aim was to characterize the prevalence, explanatory factors, and consequences of inaccurate prognostic awareness and interest in therapy.

Results: Inaccurate prognostic awareness affected 74%, conditioned by the administration of vague information without alluding to death (odds ratio [OR] 2.54; 95% CI, 1.47-4.37, adjusted P = .006). A full 68% agreed to low-efficacy therapies. Ethical and psychological factors oriented first-line decision-making, in a trade-off in which some lose quality of life and mood, for others to gain autonomy. Imprecise prognostic awareness was associated with greater interest in low-efficacy treatments (OR 2.27; 95% CI, 1.31-3.84; adjusted P = .017), whereas realistic understanding increased anxiety (OR 1.63; 95% CI, 1.01-2.65; adjusted P = 0.038), depression (OR 1.96; 95% CI, 1.23-3.11; adjusted P = .020), and diminished quality of life (OR 0.47; 95% CI, 0.29-0.75; adjusted P = .011).

Conclusion: In the age of immunotherapy and targeted therapies, many appear not to understand that antineoplastic therapy is not curative. Within the mix of inputs that comprise inaccurate prognostic awareness, many psychosocial factors are as relevant as the physicians' disclosure of information. Thus, the desire for better decision-making can actually harm the patient.

Keywords: decision-making; depression; information; prognostic awareness; quality of life.

Figure 1.

(A) Relationship between belief in curability vs. interest in therapy. (B) Relationship between type of disclosure and belief in curability. Abbreviations: QoL, quality of life; Res, treatment that achieves tumor response; <6m, treatment that increases survival by at least 6 months; 6-12m, treatment that increases survival by between 6 and 12 months; >12m, treatment that prolongs survival by more than 12 months, >24m, treatment that increases survival by more than 24 months. Note: (A) Displays the percentage of individuals interested in antineoplastic therapy under different assumptions regarding its efficacy (symptom relief or improvement in quality of life only, antitumor response without prolonging survival, or survival increments <6, 6-12, >12, or >24 months). Each panel represents possible responses to the question: “Do you expect the treatment to help cure your cancer?.” Each panel in (B) represents a different communication style of the oncologist, with the bars showing whether the patient agreed with the belief that their cancer could be cured.

Figure 2.

Marginal effects for the model of belief in curability. For each variable, the plot shows the probability that the belief that cancer can be cured is scored as likely or very likely. All other covariates are held constant at the average or baseline level. The main assumptions of the model are shown. Wald tests for the most meaningful hypotheses in a design are shown. Abbreviations: Br, breast cancer; chemo, chemotherapty; co, colorectal; H&N, head and neck cancer; immu, immunotherapy; lu, lung cancer; oth, other tumors; Pr, primary eductation; qual− , qualitative, death not alluded; qual + , qualitative, death alluded; quant , quantitative; Se, secondary education; targ, targeted therapies; UG , upper gastrointestinal cancer; uni, university studies.

Figure 3.

Proportional odds model to predict interest in low-efficacy therapies. The graph displays the odds ratios resulting from this model. The complete model is shown in Table 2. Interquartile effects are presented in the case of continuous variables. Odds ratios >1 denote greater belief that cancer can be cured. Abbreviations: ECOG PS, Eastern Cooperative Group performance status; H&N, head & neck cancer; FACIT, Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being Scale; QLQ-C30, European Organization for Research, and Treatment of Cancer Quality of Life C30 Questionnaire; mini-MAC, Mini-Mental Adjustment to Cancer Scale..

Figure 4.

Kendall’s τ correlation coefficients between the belief in curability and different scores. Abbreviations: BSI, Brief Symptom Inventory-18; CWQ-FoR, Cancer worry (for health) questionnaire; HHS, Herth Hope Scale; Mini-MAC, Mini-Mental Adjustment to Cancer Scale; QLQ-C30, EORTC Core Quality of Life questionnaire; see questionnaires in Supplementary material. Interpretation of higher scores: QLQ-C30 global scale, higher level of QoL; for other QLQ-C30 scales, greater symptom burden or function impairment.