Abstract
We undertook a cross‐sectional survey of a random sample of thoracic oncologists from the American Society of Clinical Oncology clinical directory to characterize whether prognostic uncertainty has increased and if tolerance of uncertainty is associated with prognostic discussion practices. We also assessed the Physicians’ Reactions to Uncertainty Scale and presented a vignette about an incurable patient with uncertain life expectancy. One hundred and ninety‐two of 438 surveys (43.8%) were received. Of the respondents, 52.1% agreed “there is more prognostic uncertainty in the management of lung cancer now than 10 years ago,” and 37.4% noted difficulty “staying up‐to‐date.” In multivariable analyses, physician‐reported anxiety about uncertainty (p = .05) and reluctance to disclose uncertainty (p = .04) were inversely associated with reporting having prognostic discussions with most patients. For the vignette, 92.1% reported they would discuss incurability, but only 76.3% said they would discuss the patient's life expectancy. Our data suggest prognostic uncertainty has increased in thoracic oncology and oncologists’ tolerance of uncertainty may affect discussion practices.
Short abstract
Limited tolerance of uncertainty may affect a clinician's communication practices and ability to assist patients with navigating such uncertainty. This article assesses the prevalence of prognostic uncertainty among thoracic oncologists, their personal tolerance of uncertainty, and whether their tolerance was associated with prognostic discussion practices.
Introduction
Clinician intolerance of uncertainty has been associated with the propensity to order more diagnostic tests, increased anxiety about malpractice, and the inclination to withhold ambiguous test results [1]. Indeed, although helping patients to process uncertainty is critical to effective patient‐centered communication [2], limited tolerance of uncertainty may affect a clinician's communication practices and his/her ability to assist patients in navigating such uncertainty [3]. In oncology, patients need to understand their prognosis to make decisions, even if it is uncertain; for example, patients with advanced cancer may opt for overly aggressive treatments because they do not understand their disease is incurable [4]. We hypothesized that intolerance of uncertainty is likely prevalent in lung cancer care, in which targeted treatments have made accurate prognostication difficult [5]. We aimed to assess the prevalence of prognostic uncertainty among thoracic oncologists, their personal tolerance of uncertainty, and whether their tolerance was associated with prognostic discussion practices.
Materials and Methods
We developed a 34‐item physician survey (Appendix) with questions about professional training, perceptions of prognostic uncertainty for advanced lung cancer, prognostic discussion practices, and uncertainty tolerance adapted from the Physicians’ Reactions to Uncertainty (PRU) scale [6, 7]. This scale has “Anxiety due to Uncertainty” and “Reluctance to Disclose Uncertainty” subscales with scores ranging from 5 (tolerates well) to 30 (tolerates poorly). We also presented a clinical vignette of a patient with newly diagnosed stage IV non‐small cell lung cancer and asked whether respondents would (a) tell the patient she was incurable and (b) discuss her life expectancy, specifically stated in the vignette as uncertain. We piloted the survey and conducted cognitive interviewing with five thoracic oncologists from the Dana‐Farber/Harvard Cancer Center.
The survey was sent to a random sample of thoracic oncologists (n = 438) from the American Society of Clinical Oncology online physician directory. Physicians had to be board‐certified in medical oncology, practice in the U.S., and longitudinally care for patients with lung cancer. As incentive, participants received a $25 gift card if they returned a completed survey. Nonrespondents were contacted three additional times, twice by mail and once by phone. Survey administration occurred from October 2018 to April 2019. Return of a completed survey implied consent to participate.
We first assessed the proportion of oncologists who reported “more prognostic uncertainty in the management of lung cancer now than 10 years ago,” agreed “the rapid pace of advances in lung‐cancer management makes it difficult for me to stay up‐to‐date on the most recent science,” reported explicitly discussing prognosis with “most” patients (>95%) at some point in the disease course, would tell the vignette patient she has incurable disease, and would disclose that her life expectancy is uncertain. We next scored the PRU “Anxiety” and “Reluctance” subscales and assessed whether scores in the highest quartile were associated with oncologists’ propensity to engage in prognostic discussions and/or their responses to the vignette. Odds ratios (ORs) and confidence intervals (CIs) were computed using Wald χ2 tests with a one‐sided p < .05 adjudged as statistically significant. All analyses were conducted using R and RStudio (R Group, Vienna). Methods were approved by the Dana‐Farber/Harvard Cancer Center Office for Human Subjects Research.
Results
Overall, 192 surveys (43.8%) were returned completed. The median respondent age was 54.5 years, duration of clinical practice was 29 years, and 37.5% reported a full‐time academic affiliation (Table 1). Approximately half of respondents (52.1%; 95% CI, 44.8%–59.4%) agreed “there is more prognostic uncertainty in the management of lung cancer now than there was 10 years ago,” whereas approximately a third (37.4%; 30.5%–44.7%) reported difficulty staying “up‐to‐date on the most recent science.” Respondents had a median PRU Anxiety score of 11.0 (interquartile range [IQR], 6.0) and median Reluctance score of 12.0 (IQR, 7.0). The majority (78.9%) reported discussing prognosis at some point during the disease course with “most” patients, whereas 21.5% reported discussing prognosis with 51% to 95% of patients, and 4.2% reported discussing prognosis with ≤50%.
Table 1.
Characteristics of respondents (n = 192)
| Characteristics | n (%) |
|---|---|
| Age, yr | |
| First quartile | 48.2 |
| Median | 54.5 |
| Third quartile | 62.0 |
| Gender | |
| Female | 39 (20.5) |
| Male | 151 (79.5) |
| Graduated medical school | |
| Before 1980 | 36 (19.4) |
| 1980–1999 | 120 (64.5) |
| After 2000 | 30 (16.1) |
| Board certification | |
| Medical oncology | 161 (83.9) |
| Hematology | 87 (45.3) |
| Palliative care | 7 (3.6) |
| Other | 2 (1.0) |
| Average weekly patients | |
| First quartile | 43 |
| Median | 70 |
| Third quartile | 100 |
| Trial enrollment | |
| Yes | 161 (83.9) |
| No | 31 (16.1) |
| Region | |
| Midwest | 50 (26.0) |
| Northeast | 52 (27.1) |
| South | 52 (27.1) |
| West | 38 (19.8) |
| Practice location | |
| Community | 101 (52.8) |
| Tertiary | 83 (43.5) |
| Both | 7 (4.7) |
| Academic affiliation | |
| No affiliation | 38 (19.8) |
| Partial affiliation | 82 (42.7) |
| Full‐time faculty | 72 (37.5) |
Because of incomplete survey responses, some columns do not add up to 192.
Those who were in the highest quartile of Anxiety (OR, 2.13; p = .05), in the highest quartile of Reluctance (OR, 2.31; p = .04), and male sex (OR, 3.82; p = .02) had higher odds of discussing prognosis with <95% of their patients (Table 2). For the vignette, 92.1% (95% CI, 87.4%–95.5%) reported that they would tell the patient she had incurable disease, but only 76.3% (69.6%–82.2%) would discuss her uncertain life expectancy. High reluctance to disclose uncertainty on the PRU was not associated with likelihood of discussing incurability but highly associated with unwillingness to discuss life expectancy (p < .001). Additional survey results and analyses will be reported in a separate publication.
Table 2.
Variables associated with frequency of discussing prognosis (n = 192)
| Variables | Reported discussing prognosis with ≤95% vs >95% of patients | |
|---|---|---|
| Odds ratio (95% CI) | p value | |
| PRU: Anxious due to uncertainty | 2.13 (0.99–4.58) | .05 |
| PRU: Reluctant to disclose uncertainty | 2.31 (1.03–5.19) | .04 |
| Male gender | 3.82 (1.11–13.15) | .02 |
| Graduated before 2000 | 0.50 (0.17–1.54) | .22 |
| Tertiary center | 0.84 (0.41–1.71) | .63 |
| Older than median | 1.86 (0.91‐3.80) | .09 |
| Weekly patients >70 | 1.28 (0.63–2.57) | .49 |
| Clinical trial enrollment | 0.41 (0.18–0.94) | .03 |
| Discuss incurability in vignette | 0.35 (0.12–1.06) | .05 |
| Discuss life expectancy in vignette | 0.26 (0.12–0.55) | <.001 |
Abbreviations: CI, confidence interval; PRU, Physicians’ Reactions to Uncertainty Scale.
Discussion
Our diverse cohort of lung cancer oncologists reported increasing prognostic uncertainty within thoracic oncology and challenges staying current with the evolving science. Notably, only 79% reported discussing prognosis with most (>95%) of their patients at some point in the clinical course. Intriguingly, those with greater “reluctance” or “anxiety” due to uncertainty seemed significantly less likely to do so. This adds to our previous work in hematologic oncology suggesting that those endorsing prognostic uncertainty as a barrier to end‐of‐life care also tended to discuss prognosis with fewer patients [8]. Patients with cancer often desire their physicians to provide them with all information regarding diagnosis, treatment options, and prognosis [9]. Despite these preferences, oncologists may worry that frank communication about an uncertain prognosis will cause distress and loss of hope [10]. Prior literature directly links oncologists’ hesitance to discuss prognosis with patients’ inaccurate understanding of their disease severity [11]. Misunderstandings regarding disease severity and prospects for cure hinder treatment choices consistent with patients’ goals and lead to dissatisfaction with care [12]. Our findings suggest that an oncologists’ own feelings toward uncertainty may mediate his/her proclivity to discuss prognosis.
Our study has limitations. First, our response rate was not optimal; however, given the geographic diversity of respondents and the balanced percentages of community and tertiary physicians, we feel they were likely highly representative of lung cancer oncologists in the U.S. Second, although the results from our self‐reported practices regarding prognostic discussions and our vignette were largely consistent, neither can completely reveal what oncologists actually do when faced with a real patient. Finally, our relatively small sample size did not allow for correction for multiple hypothesis testing.
In summary, our analysis suggests the need to develop interventions to help manage oncologists’ uncertainty, as it may affect their likelihood to discuss prognosis with patients. Moreover, the recent shift from so‐called “therapeutic nihilism” [5] to hopefulness in the treatment of thoracic cancers presents a unique challenge, as patients may be less primed for bad news. In the complex clinical encounter, patient's expectations undoubtedly play a role and may conspire with the clinician's own intolerance of uncertainty to limit the flow of information that many patients so desperately want.
Disclosures
The authors indicated no financial relationships.
Acknowledgments
This work was supported by a Harvard Medical School Scholars in Medicine Award.
No part of this article may be reproduced, stored, or transmitted in any form or for any means without the prior permission in writing from the copyright holder. For information on purchasing reprints contact commercialreprints@wiley.com. For permission information contact permissions@wiley.com.
References
- 1. Allison JJ, Kiefe CI, Cook EF et al. The association of physician attitudes about uncertainty and risk taking with resource use in a Medicare HMO. Med Decis Making 1998;18:320–329. [DOI] [PubMed] [Google Scholar]
- 2. Mack JW, Joffe S. Communicating about prognosis: Ethical responsibilities of pediatricians and parents. Pediatrics 2014;133(suppl 1):S24–S30. [DOI] [PubMed] [Google Scholar]
- 3. Street RL, Epstein RM. Patient‐Centered Communication in Cancer Care: Promoting Healing and Reducing Suffering. Bethesda, MD: U.S. Department of Health and Human Services, National Institutes of Health, National Cancer Institute, 2007. [Google Scholar]
- 4. Weeks JC, Catalano PJ, Cronin A et al. Patients’ expectations about effects of chemotherapy for advanced cancer. N Engl J Med 2012;367:1616–1625. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Temel JS, Shaw AT, Greer JA. Challenge of prognostic uncertainty in the modern era of cancer therapeutics. J Clin Oncol 2016;34:3605–3608. [DOI] [PubMed] [Google Scholar]
- 6. Gerrity MS, White KP, DeVellis RF et al. Physicians’ reactions to uncertainty: Refining the constructs and scales. Motivation and Emotion 1995;19:175–191. [Google Scholar]
- 7. Politi MC, Clark MA, Ombao H et al. The impact of physicians’ reactions to uncertainty on patients' decision satisfaction. J Eval Clin Pract 2011;17:575–578. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Habib AR, Cronin AM, Earle CC et al. How do blood cancer doctors discuss prognosis? Findings from a national survey of hematologic oncologists. J Palliat Med 2019;22:677–684. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Degner LF, Kristjanson LJ, Bowman D et al. Information needs and decisional preferences in women with breast cancer. JAMA 1997;277:1485–1492. [PubMed] [Google Scholar]
- 10. Miyaji NT. The power of compassion: Truth‐telling among American doctors in the care of dying patients. Soc Sci Med 1993;36:249–264. [DOI] [PubMed] [Google Scholar]
- 11. Liu PH, Landrum MB, Weeks JC et al. Physicians’ propensity to discuss prognosis is associated with patients’ awareness of prognosis for metastatic cancers. J Palliat Med 2014;17:673–682. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Mack JW, Weeks JC, Wright AA et al. End‐of‐life discussions, goal attainment, and distress at the end of life: Predictors and outcomes of receipt of care consistent with preferences. J Clin Oncol 2010;28:1203–1208. [DOI] [PMC free article] [PubMed] [Google Scholar]