Patient Preferences for Communication of Life Expectancy in Prostate Cancer Treatment Consultations

Abstract

This qualitative study examines patient preferences among men with prostate cancer for communications from clinicians about life expectancy.

Life expectancy (LE) is a critical factor in treatment decision-making for men with prostate cancer, since limited LE is associated with lower likelihood of sufficient longevity to benefit from treatment,¹ higher treatment morbidity,² and decreased treatment effectiveness.³ Despite a prominent role of LE in guidelines, men with limited LE are often overtreated for indolent cancers and undertreated for high-risk cancers.⁴ Previous studies suggest that such mismanagement may be because of ineffective communication of LE by clinicians.⁵ Yet, patient perspectives on how LE should be ideally communicated are unknown.

Methods

We conducted semistructured interviews with men after treatment consultations for prostate cancer (Gleason score ≤7) between March 2019 and December 2020. Views on importance of LE, if or how it had been expressed, barriers to discussing LE, confidence in LE estimates, and ideal modes of communication were assessed using open-ended questions (eMethods in the Supplement). We used an inductive qualitative coding approach with ATLAS.ti version 8.4.5 (ATLAS.ti Scientific Software Development GmbH) to identify emerging themes. The study was approved by the Cedars-Sinai Institutional Review Board. Informed consent (written or verbal) was obtained by the study team.

Results

A total of 26 men participated. The median age in the sample was 67 (IQR, 60-72) years, and 5 of 26 had LEs less than 10 years (19%). Of participants who recalled discussing LE (19 of 26 [73%]), three-quarters recalled LE as being communicated as a generalization (eg, “long”; 7 of 19 [37%]) or a survival probability at a point (eg, “20% probability of living 15 years”; 7 of 19 [37%]), while one-quarter recalled it as a number of years (“you’ll live to 90”; 5 of 19 [26%]). Six of 7 (86%) who did not recall discussing LE thought it would have been helpful.

In a hypothetical scenario addressing the ideal mode of LE communication, most preferred a number of years (12 of 24 [50%]) or survival probability at a point (9 of 24 [38%]) rather than a generalization (3 of 24 [13%]). Those who preferred a number of years noted it was easily understandable (6 of 12 [50%]) and concrete (5 of 12 [38%]). Those who preferred survival probability indicated it offered hope, even if it was low (5 of 9 [56%]). Those opposed to generalization felt that “doctors [were] not giving patients credit” as part of shared decision-making. Preferences did not differ among those with LEs less than 10 years compared with those with LEs greater than 10 years.

The predominant barrier to discussing LE was anxiety (15 of 26 [58%]), which participants noted could be reduced by using a range of years (eg, “10-15 years”; 3 of 15 [20%]), depersonalizing language (eg, “for patients like you”; 2 of 15 [13%]), and other methods (Table). Most had low (6 of 24 [25%]) or moderate (13 of 24 [54%]) confidence in LE estimates, which participants noted could be improved by explaining the calculation method (11 of 25 [44%]), mentioning that health status (7 of 25 [28%]) or family history or genetics (5 of 25 [20%]) was considered, and other methods (Table). Twenty-two of 26 (87%) felt that LE should always be provided in consultations.

Table. Patient-Identified Methods to Reduce Anxiety and Improve Confidence in Life Expectancy (LE) Estimates.

Methods to reduce anxiety	Participants endorsing, No./total No. (%)	Methods to improve confidence	Participants endorsing, No./total No. (%)
Provide a more generalized range of years rather than 1 number (eg, “12-15 y” rather than “13 y”)	3/15 (20)	Explanation of how method was calculated	11/25 (44)
Use language that depersonalizes the information (eg, “LE for patients like you…”)	2/15 (13)	Assurance that medical conditions or history was included in LE estimate	7/25 (28)
Communicate data in a compassionate, down-to-earth way	2/15 (13)	Assurance that family history or genetics was included in LE estimate	5/25 (20)
Explanation of how method was calculated or scientific data	1/15 (7)	Experience of physician	5/25 (20)
Ask patient how much they wish to know and tailor the vagueness or specificity of the information given	1/15 (7)	Second opinion	4/25 (16)

Discussion

We sought to understand patient perspectives on how LE should be optimally communicated to create patient-centered strategies that address the emotional, rhetorical, and attitudinal challenges preventing effective LE communication. Most men in the sample (1) endorsed LE as a key component of treatment counseling that should not be omitted (87%); (2) preferred LE to be communicated as specific, quantitative information rather than a generalization (88%); (3) had low to moderate confidence in LE estimates (79%) but offered language-based methods for improving confidence; and (4) endorsed anxiety (58%) as the main barrier to discussing LE, which could be alleviated by language-based approaches (Table). Interestingly, the preference for detailed LE information in men with prostate cancer contrasts with LE preferences in screening populations,⁶ which may be attributable to their need to weigh competing risks of mortality against substantial treatment-associated morbidity. Future studies will need to confirm the generalizability of these findings to men with low health literacy or numeracy (since our sample was drawn from a well-educated, health-literate population), non–English-speaking populations, and individuals in other clinical settings (eg, advanced disease).

Based on our findings, we created an annotated example for how LE should be optimally communicated (Figure). Using patient-centered language to describe this computationally challenging and emotionally evocative concept may improve understanding and acceptance, leading to more informed shared decision-making.

Figure. Annotated Example of Patient-Centered Communication of Life Expectancy (LE) and Competing Risks of Mortality.

Language-based strategies to improve communication are coded by methods to reduce anxiety, communication of LE and cancer prognosis, and methods to improve confidence.

Supplement.

eMethods.