Oncologists’ Reluctance to Use the Terms Hope and Cure: A Bibliometric Analysis of Articles From Two High-Impact Oncology Journals

Abstract The words cure and hope are important terms in oncology, reflecting a balance of aspirations and realism for physicians and patients. Yet, some have suggested that oncologists are reluctant to use these terms. We tested this hypothesis by performing a bibliometric analysis of the frequency of use of these words in JAMA Oncology (JAMA Oncol) and the Journal of Clinical Oncology (JCO). The text of all articles in 3 categories—primary research, editorials, and narrative essays—appearing in JCO from 2000 to 2018 and in JAMA Oncol from 2015 to 2019 was analyzed. These analyses compared, across these categories, the proportion of articles containing the words cure and hope, as well as the proportion of total sentences containing these words. There were statistically significant differences in frequency of the use of the terms cure and hope as a function of the type of article published in the JCO and JAMA Oncol (2-sided P values ranging from .005 to <.001). Results were similar for both journals, with minor exceptions. Both hope and cure were used in a greater number of articles and sentences in the narrative and editorial categories than in primary research. Moreover, hope was used more often in narrative essays than in editorials. The relative reluctance to use these terms in more scientifically oriented original reports, despite concomitant improvements in oncologic outcomes, may reflect a bias worthy of future exploration.

During a 2018 keynote address (1) to the American Society of Clinical Oncology, Dr Norman Sharpless (director of the National Cancer Institute) observed, "An overarching worry of today's oncologist has been the management of expectations. We don't want to overpromise and give hope that is false . . . I think we have become scared to tell patients that we actually hope to cure them, and it may be time to examine how we communicate our efforts in this area." We realize that it is rationaland even ethical-for physicians to avoid exaggerating possible benefits of treatment. However, it was sobering to hear Sharpless further suggest that oncologists are simply uncomfortable with words like cure and hope even though patients crave them.
If oncologists use these terms infrequently, such caution may not be entirely warranted. With regard to cure, data from the American Cancer Society indicate that survival has steadily improved for almost 30 years, for nearly all common malignancies (2). For instance, the overall annual cancer death rate dropped continuously from 1991 to 2017 by 29%. In the United States, approximately 2.5 million fewer people have died of cancer during the past 3 decades than would have died if cure rates had remained unchanged. But, have oncologists been willing to acknowledge this reality to their patients and among themselves?
Hope, for its part, is no longer an amorphous concept, having been rigorously investigated with validated tools during the same 3-decade period (3,4). Specifically, an operationalized model known as Hope Theory has emerged that conceptualizes hope as a goal-directed construct. Given that goals consist of anything that an individual desires to do, be, get, or experience, hope matters in virtually every context and stage of life. In its absence, patients are often beset by emotional distress and decreased ability to cope with physical symptoms, and healthcare providers may be at risk for burnout (5)(6)(7). Given Sharpless' comments, however, it is important to note that hope need not be focused solely on goals related to cure or prolongation of life. In fact, it may be desirable to redirect patients diagnosed with malignant disease toward noncancer-related hopes that are important to them, even when cure is unlikely (8,9). But, are oncologists indeed reluctant to use the terms hope and cure? To answer this question in a data-driven manner, we performed quantitative bibliometric analyses of trends in the published literatures of 2 high-impact oncology journals.

Methods
The journals JAMA Oncology (JAMA Oncol) and Journal of Clinical Oncology (JCO) were selected for analysis not only because of their high impact but also because they both regularly publish articles in 3 categories: primary research (Original Investigations in JAMA Oncol; Original Reports in JCO), editorials (Viewpoints in JAMA Oncol; Editorials in JCO), and narrative essays (Cancer Care Chronicles in JAMA Oncol; Art of Oncology in JCO). In JAMA Oncol, narrative essays are defined (10) as "personal vignettes taken from wide ranging experiences in medicine," and in JCO, such essays are intended to (11) "explore the experience of suffering from cancer or caring for people with cancer." Editorials in JAMA Oncol are defined as "opinions that address any important topic in medicine, public health, research, discovery, prevention, ethics, health policy or health law." A strict definition of editorials is not provided in the "Information for Authors" section of JCO. To avoid confusion, we will use the terms primary research articles (PR), editorials (ED), and narrative essays (NE) to refer to the above categories, regardless of journal.
We analyzed usage patterns of the words hope and cure in all articles appearing since the inception of these journals' respective narrative sections. Accordingly, for JCO, we examined a 19year period from 2000 to 2018 (12 604 articles) and, for JAMA Oncol, a 5-year period from 2015 through 2019 (759 articles). We obtained raw text of all articles in electronic format and performed a series of 2-proportion z tests using the StatsModels module (12) in Python. These analyses yielded comparisons across all 3 categories (PR, ED, NE) of the proportion of articles that mentioned hope or cure at least once in their text as well as the total number of sentences containing these words. All P values are 2-sided, with initial critical values for statistical significance set at .05. However, given that we performed multiple comparisons, Bonferroni correction was used, resulting in a corrected critical value for statistical significance of .008 for each set of analyses.
For reasons we discuss in detail below, we expected that a greater proportion of NE articles than PR or ED articles would mention the words hope and/or cure, and a greater proportion of ED than PR pieces would mention these target words. Because it is possible that articles in any of these categories could mention a target word but not use it frequently, we also analyzed the proportion of total sentences containing these words, expecting a parallel pattern.

Results
Results were nearly identical for the 2 journals, with only minor exceptions. Table 1 portrays article-level statistics. As expected, in both journals, hope was mentioned at least once in the text of a greater proportion of NE than ED or PR articles and in a greater proportion of ED than PR articles. Also in both journals, cure was mentioned at least once in the text of a greater proportion of both NE and ED articles than PR articles (the NE-PR comparison was approaching statistical significance). Only in JCO, however, was cure used in a greater proportion of NE than ED articles. Findings were similar at the sentence level (Table 2).
Of note, we also found evidence of diminishing use of the terms hope and cure over time (Tables 3 and 4) (2). Analyses evaluating this issue were only performed on articles from JCO, given the longer timespan of available data (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018). Specifically, we divided the articles into 4 time periods and performed a series of Cochran-Armitage trend analyses to test for historical change. Although most of these analyses yielded nonstatistically significant results , there were a number of exceptions to this rule. At the article level, the proportion of pieces containing the word cure statistically significantly decreased over time in both the NE (P ¼ .004) and PR categories (P < .001). At the sentence level, the proportion of sentences containing the word cure statistically significantly decreased over time in both the ED (P < .001) and PR (P < .001) categories, as did the proportion of sentences containing the word hope in the PR category (P < .001). Thus, despite the improving trends in oncologic outcomes over time, professionals seem less likely to mention both hope and cure in published work.

Discussion
Bibliometric analyses like those reported here are known and accepted methods for gauging trends in a field (13). Nonetheless, like any methodology, they are imperfect. For instance, it is difficult, if not impossible, to determine the exact meaning of every mention of hope and cure in the 13 363 articles represented in our analysis. For this, nearly prohibitively detailed qualitative analysis of every instance would be necessary. Moreover, alternative interpretations of our findings must be entertained. Does the relative absence of hope and cure from written communication indicate commensurate disinclination to speak these words with patients? Does the use of hope and cure differ between academic investigators-who author most articles-and practitioners who deliver care outside of academic settings? Do people have differing conceptions of the terms hope and cure (14)? As several authors have noted, despite attempts to define cure for patients with cancer, the term is used heterogeneously in the oncology literature (15,16). We see these as questions worthy of further investigation. Placing our 3 article categories on a continuum of descending scientific content, authors appear less likely to use cure and hope in more empirically rigorous primary research articles than  No. (%) of sentences with word hope editorial or narrative pieces. These results may not seem surprising, although the implications are potentially far-reaching. Divergent explanations exist. First, hope and cure might not be taken seriously by oncologists and therefore not incorporated into primary research articles. Conversely, heightened respect could be attached to these terms, such that authors do not feel they are appropriate in manuscripts where primary research is reported. Or, it may simply be that hope and cure are often omitted from primary research articles because they are not designated as outcome measures in the investigations typically reported in these 2 journals. Another viewpoint is that the narrative essay-a genre that, by design, enables authors to be reflective-is likely to be enriched by concepts like hope, which people may consider meaningful but are reluctant to mention in more professional contexts, given the perception that they are overly abstract or sentimental. Regardless, our findings are consistent with Dr Sharpless' commentary about reluctance to use these terms despite improving survival rates for patients and the growing understanding of hope as a construct that can be rigorously measured and is not merely synonymous with the possibility of cure (2,3). In the future, we propose further exploration of this hypothesis in the literatures of related specialties such as geriatrics or intensive care and of other healthcare professions such as nursing or social work.
In the meantime, physicians caring for patients with cancer may wish to reacquaint themselves with the accomplishments of their colleagues (2) to appreciate the instances where cure is achievable while simultaneously broadening their definition of hope when cure is elusive. Perhaps the time has come to judiciously embrace the words hope and cure.

Funding
This work was supported, in part, by the Pamm Gross Kahane Research Institute of Life's Door.

Notes
Role of the funder: The funder had no role in the writing of this commentary or the decision to submit it for publication.

Conflicts of interest:
The authors have no conflicts of interest to disclose.