Physicians’ use of plain language during discussions of prostate cancer clinical trials with patients

Abstract

Objective:

This study described physicians’ use of plain language during patient-physician cancer clinical trial discussions.

Methods:

Video-recorded clinical interactions and accompanying transcripts were taken from a larger study of communication and clinical trials (PACCT). Interactions (n = 25) were selected if they included invitations to participate in a clinical trial. We used descriptive, qualitative discourse analysis, a method that identifies language patterns at or above the sentence level. We first excerpted discussions of clinical trials, then identified instances of plain language within those discussions. Finally, we inductively coded those instances to describe physicians’ plain language practices.

Results:

The analysis identified four plain language practices. Lexical simplification replaced medical terminology with simpler words. Patient-centered definition named, categorized, and explained complex medical terminology. Metaphor explained medical terminology by comparing it with known concepts. Finally, experience-focused description replaced medical terminology with descriptions of patients’ potential physical experiences.

Conclusion:

These plain language practices hold promise as part of effective information exchange in discussions of cancer clinical trials. Testing is needed to identify patient preferences and the extent to which these practices address patient health literacy needs.

Practice Implications:

Pending further testing, these plain language practices may be integrated into physician clinical trial and other communication training.

1. Introduction

The use of plain language has consistently been tied to higher-quality communication in healthcare settings [1]. Plain language is defined as communication where the “wording, structure, and design are so clear that the intended audience can easily find what they need, understand what they find, and use that information” [2]. By focusing on specific audiences’ ability to use communication, plain language can be a flexible range of approaches rather than simplified rules for “dumbing down” language [3]. This article presents an observational study of plain language in a context where it has not been extensively studied—patient-physician discussions of cancer clinical trials.

For over a decade, the Plain Writing Act of 2010 has required federal government documents to be clear, concise, and well-organized [4]. This act was operationalized through the Federal Plain Language Guidelines, which advise writers on word choice, sentence structure, visual design, and methods for testing communication [5]. Subsequently, both the Centers for Disease Control and Prevention (CDC) and the National Institutes of Health (NIH) have developed guidance for using plain language [6,7], and training in the use of plain language has been recommended as a part of healthcare education [8]. In the context of clinical trials, plain language guides have been developed to improve patients’ readiness to make decisions about participation [9,10].

Plain language is often framed as a means of improving healthcare outcomes for patients with limited health literacy [1]. Essentially, health literacy refers to patients’ abilities to obtain, understand, and use information to inform health-related decisions [11,12], though recent literature has acknowledged the ways that health literacy can fluctuate across time and context [13]. Limited health literacy has been regularly associated with worse health outcomes for patients [14–17], and previous research has identified the frequent use of medical terminology as an important barrier to patients with limited health literacy [14, 18]. Therefore, plain language has been identified as a strategy for providing culturally and linguistically appropriate services to diverse patient populations [19] and recommended as a component of physician training [8,20].

While there is an ongoing discussion of the written, oral, and listening aspects of health literacy [21], plain language literature has focused on written documents rather than spoken discourse. The limited literature on spoken plain language diverges from the focus on simple wording and sentence structure in written plain language. For example, a brief federal resource recommends using active voice to improve audience recall, using literary devices, such as metaphor and repetition, to engage audiences, and avoiding filler words during pauses [22]. In healthcare contexts, spoken plain language has been operationalized in the Plain Language Planner tool, which advises palliative care nurses to limit jargon, use second-person pronouns, and speak in short sentences [20,23]. It has also been studied in medical students’ speech during simulated patient interviews [24]. However, this literature has not included observational analyses of plain language use in actual patient-physician communication. By analyzing those practices more directly, we can potentially identify a broader and more nuanced range of practices for achieving the goals of plain language in spoken communication.

Spoken plain language is essential in sensitive or challenging communication contexts, including patient-physician discussions of cancer clinical trials. Clinical trials are the gold standard for research on the safety and effectiveness of new treatments in cancer medicine [25]. Meanwhile, the quality and content of verbal communication have been shown to significantly impact patients’ decisions about participating in clinical trials [26–28]. Additionally, complex presentations of clinical trials are associated with lower refusal rates for participation [29], suggesting that patients may make decisions about trials they do not fully understand. This article builds on prior observational research by using descriptive, qualitative discourse analysis to identify plain language practices used by physicians during medical encounters in which they invited a patient to participate in a clinical trial. As such, it seeks to contribute to ongoing research on improving patient understanding of cancer clinical trials.

2. Methods

2.1. Participants, settings, and IRB approval

The observational study described here is a secondary analysis of data from a parent study, Partnering Around Cancer Clinical Trials (PACCT) [30]. The parent study tested a communication intervention designed to increase clinical trial participation among Black and White men with prostate cancer. PACCT participants included physicians (medical oncologists, urologists, and radiation oncologists) from two NCI-designated comprehensive cancer centers: Wayne State University/Karmanos Cancer Institute in Detroit, Michigan, and Johns Hopkins Medicine/Kimmel Comprehensive Cancer Center in Baltimore, Maryland. All physician participants regularly treated patients with prostate cancer. PACCT patient participants had a confirmed diagnosis of prostate cancer (with a recruitment preference for patients with intermediate to advanced prostate cancer) and self-identified as either Black/African American or White.

All procedures were approved by the Institutional Review Boards (IRBs) at both data collection sites, and all participants signed written consent forms. Procedures for the parent study included obtaining participant self-report data and audio/video recordings of clinic visits with patients who were potentially eligible for a prostate cancer clinical trial (November 2016-January 2019). All recordings were transcribed verbatim.

Transcripts of patient-physician interactions (n = 25) were selected for the current study if they were identified as the patient participants’ first recording to include an explicit or pending invitation to participate in a clinical trial, using the method described in a previous study on the same sample [31]. All patient participants (n = 25) appeared in only one recording. The patients’ mean age was 67 (standard deviation, 6.8 years); 10 patients (40%) were Black, and 15 patients (60%) were White (Table 1). Among physician participants (n = 10), one appeared in six transcripts and another in five transcripts. The remaining eight physicians appeared in three or fewer transcripts.

Table 1.

Patient Participant Characteristics (n = 25).

Characteristic	Value
Mean age, years (SD)	67 (SD = 6.8)
Race
Black	10 (40)
White	15 (60)
Education
≤ High School	6(24)
Some or graduated college	13 (52)
Graduate or professional degree	6 (24)
Annual household income
< 20,000.00	1 (4)
20,000–39,000	1 (4)
40,000–59,000	4 (16)
60,000–79,000	5 (20)
< 80,000	12 (48)

NOTE. Data are No. (%) unless otherwise indicated. Abbreviation: SD, standard deviation.

2.2. Analysis

We conducted a descriptive, qualitative discourse analysis of transcripts of patient-physician discussions of clinical trials using NVivo qualitative analysis software [32]. Discourse analysis is a widely used method to examine patterns in naturally occurring language at or above the sentence level and seek to understand those patterns within social and cultural contexts [33–35]. As such, discourse analysis is contrasted with other forms of structural linguistic analysis, which look at smaller segments of language such as sounds (phonetics) or parts of words (morphology). Broadly speaking, discourse analyses can be descriptive, used to explain current practices, or critical, used to intervene in existing communication practices [34]. Our study used a descriptive, qualitative discourse analysis method to identify and interpret plain language communication practices employed by physicians during discussions of clinical trials with patients.

First, two authors (LT and EB) reviewed the transcripts and excerpted discussions of clinical trials. The lead author (LT) then identified instances of plain language within those excerpts. Based on previous literature, which aligned plain language in healthcare with practices that limit the use of medical terminology [20], plain language was operationally identified in this study as descriptions of clinical trials where physicians replaced or explained medical or technical terminology with nontechnical language. In the transcripts, discourse identified as plain language included discussions of treatment mechanisms, cancer prognosis, trial procedures, and molecular and cellular physiology. Then, the lead author (LT) used an iterative, inductive coding process typical of discourse analysis [36] to identify and categorize patterns in physicians’ use of plain language. Finally, three members of the research team (LT, LH, SE) condensed and refined the patterns into the plain language practices described below. As part of this final step, we reviewed relevant literature in health communication to clarify the definitions of the codes and to build connections with extant research.

3. Results

We identified four plain language practices used by physicians: lexical simplification, audience-centered definition, metaphor, and experience-focused description. While the latter two practices are not part of traditional guidelines for plain language, they are included here because physicians used them to replace or explain technical or medical terminology related to clinical trials with the apparent goal of making the information clearer and more usable for patients.

3.1. Lexical simplification

Lexical simplification involves replacing medical terms with simpler words [37] and is closely tied to standard guidelines for plain language, such as selecting frequently used words instead of obscure ones [5]. To be effective, lexical simplification requires selecting terms that are likely familiar to the intended audience [38]. Example 1 illustrates how physicians used lexical simplification to replace discussions of research objectives.

Example 1.

Physician: So one of our studies is to use that same medication and see if we add a second pill to half the group whether you’d respond better. So I like that study because it’s all pills and no one gets a placebo when you’re thinking about the [hormone treatment] pills.

In this example, the physician replaced technical terminology related to clinical trials (“research questions,” “combination study”) with simplified phrases that conveyed similar concepts (“one of our studies is to … see if,” “we add a second pill”). Notably, the physician did not replace the term “placebo,” but mixing lexical simplification with medical terminology might be appropriate depending on the patient’s knowledge of specific concepts.

3.2. Audience-centered definition

An audience-centered definition names and categorizes a concept and then differentiates it from similar concepts by describing its qualities, while using audience knowledge to determine relevant categorizations and descriptions [39]. By naming medical terminology directly, audience-centered definitions can also play a key educational role [40, 41]. Example 2 illustrates how physicians used audience-centered definitions to describe the goals of clinical trials.

Example 2.

Physician: Now that is sort of an earlier phase trial, so it’s a phase one trial. Okay. But the advantage is that we’ve already been through a number of the dosing levels. So typically a phase one trial is where you’re getting something very new that’s worked amazingly well in the lab that now you’re testing in people. So the first few dose levels really are trying to see what kind of side effects and what dose level really has.

The physician initially named a concept through medical terminology (“phase one trial”) that may not be familiar to the patient. Then, in the third sentence, the physician used an audience-centered definition by naming the qualities of the treatment (“worked amazingly well in the lab”) and differentiating it (“trying to see what kind of side effects”) from other potential studies.

While Example 2 depicted a case where a physician offered an unprompted definition, Example 3 illustrates that, in some cases, physicians offered audience-centered definitions in response to prompting by patients or family members.

Example 3.

Family Member: I know it’s not cure. But you mentioned, you said remission a couple of times. What constitutes remission when you’re speaking of cancer? Physician: Remission means that the cancer disappears completely. Family Member: Uh hum. Physician: Or partially. So the complete remission would mean that the PSA level goes to zero.

Earlier in the conversation, the physician repeatedly stated that hormone therapy “causes remissions.” In response, the family member asked for clarification. The physician initially responded with a simple definition using concepts that were likely familiar to the patient (“disappears completely”) and then further clarified the definition (“PSA level goes to zero”). This kind of prompting from patients or family members suggests ways in which they can be active participants in plain language conversations.

While the first two plain language practices, lexical simplification and audience-centered definition, are consistent with guidelines for written plain language, the final two practices expand those guidelines in ways that reflect the unique qualities of spoken communication.

3.3. Metaphor

Metaphor connects two domains of knowledge by comparing abstract or unfamiliar concepts to concrete or known ones [42]. In healthcare settings, the use of metaphor has been recommended as a means to improve patient understanding of complex concepts [43,44] including in the context of clinical trial recruitment [27]. Some research frames plain language and metaphor as separate linguistic practices [44], though this differentiation is based on the narrow definition of plain language as the use of short sentences and simple words rather than as flexible approaches for replacing or explaining medical terminology. Other literature connects metaphor to plain language, including the federal primer on spoken plain language that recommends using metaphor to improve audience understanding [22].

In the current sample, metaphors acted as plain language when they were used to explain medical or technical concepts, including treatment actions, randomization, and trial benefits. Example 4 illustrates how physicians used metaphor to explain treatment mechanisms.

Example 4.

Physician: It starts making this version of uh the receptor for the hormone. So, I use this value of kind of a baseball and glove where testosterone is the baseball and the receptor is the glove. Patient: Um hum. Physician: And they have to be together. Patient: Um hum. Physician: So everything we’ve done so far is to try to keep the ball away from the glove.

The physician initially introduced the medical terminology of hormone receptors, but then shifted to a sports metaphor to explain how testosterone interacts with prostate cancer cells and clarify the goals of past treatments. Later in the conversation, the physician returned to this metaphor to describe how cancer could adapt to no longer need “the baseball” to function.

Physicians also used metaphor to replace clinical trial procedures, such as randomization, as illustrated in Example 5.

Example 5.

Physician: Now, the computer flips a coin and says whether you’re getting it or not. But you still would be coming in two times a week to get an infusion.

In this example, the physician replaced the technical terminology (“randomization”) with a common metaphor (“a coin flip”). Several physicians used similar metaphors for randomization in the data set.

3.4. Experience-focused description

Experienced-focused descriptions depict feelings or physical effects associated with treatments [45] and have been tied to improved patient evaluations of decision-making processes [46]. As a plain language practice, they replace or explain medical terminology with descriptions of potential patient experiences. In the current sample, experience-focused descriptions often explained potential side effects of a trial, as illustrated in Example 6.

Example 6.

Patient: I’ve got a couple questions for ya now. If I get the drug, what are the most common side effects? Physician: I would say, fatigue. Patient: Fatigue? Physician: Yeah. Yeah. It’s kind of, it’s a little bit of a weird fatigue. You know? Like sometimes your brain’s okay, but your body doesn’t want to get off the couch.

In this example, the physician explained medical terminology (“fatigue”) by describing the patient’s potential experience (“your body doesn’t want to get off the couch”). Notably, the physician used second-person pronouns (“your brain,” “your body”), which is recommended in existing plain language literature [23].

As illustrated in Example 7, other experience-focused descriptions used literary devices, such as repetition and dramatic language, to emphasize specific points.

Example 7.

Physician: I’m not so sure we should be pounding on your chest. Uh, you should be thinking about how far will we go even if it looks bad. ((Pause)) Does that make sense? I’m not tryin’ to give up on you ((laughs)), but I mean – you know, if you’re gonna die from this cancer, I don’t want you dyin’ on a machine in a hospital, people pounding on your chest unless that’s how you wanna leave this world.

Earlier in the conversation, the physician explained both the unproven nature of the clinical trial and its potential side effects, including pneumonia. The patient largely disregarded these concerns, saying that participating in the trial was “better than sittin’ at home and worrying.” At that point, the physician began discussing hospice. Using medical terminology, they might have discussed “quality of life,” “palliative care,” or “end-of-life decision making.” Instead, they described cardiac arrest in a hospital with dramatic language (“pounding on your chest”) to make the potential experience clearer to the intended audience. Additionally, the repetition of this language at the beginning and end of the description might contribute to the persuasive effect as well as the audience’s ability to recall the concept [47].

Finally, other experience-focused descriptions used first-person pronouns and constructed speech to convey a patient’s potential experience. Constructed speech occurs when a person describes statements that someone else might say [48]. In the sample, physicians repeatedly used constructed speech from the patient’s perspective to clarify aspects of clinical trials, as illustrated in Example 8.

Example 8.

Physician: You know say, “I don’t like any of those trials that [the physician] gave me; they don’t make sense; it’s too cumbersome. Let’s just go with what’s tried and true” and if that doesn’t work well enough long enough, then we can sort of see what trials might be available.

In this example, the physician used words the patient might say to replace concepts of voluntary participation (“I don’t like any of those trials”) and standard of care (“let’s just go with the tried and true”). Several other physicians used similar constructed speech (e.g., “I’m not interested,” “I not ready”) to convey the idea that patients can decline a trial or discontinue participation.

Previous research has identified constructed speech of past patients as a linguistic practice in recruitment for biobanking [49] and discussion of pain during prenatal counseling [50]. We are aware of no work that explores physicians’ use of constructed speech from the current patient’s potential perspective.

4. Discussion and conclusion

4.1. Discussion

This observational study identifies plain language practices that physicians used during clinical trial discussions with patients and family members. It advances prior work on plain language in several important ways. First, this study contributes to the sparse literature examining spoken rather than written plain language in healthcare contexts [20,23,24]. The plain language practices identified here expand on and in some cases conflict with existing guidelines, such as avoiding metaphors [38], but they also adapt plain language to the widely recognized structural and procedural differences between spoken and written communication [51]. Second, while each of the plain language practices has been discussed individually in previous healthcare communication contexts, we are aware of no prior work that discusses them together as a flexible means to address health literacy needs. Patients face numerous challenges regarding health literacy that can change across time and context [13]. To respond effectively to the health literacy needs of diverse patient populations, physicians require a range of communication practices, particularly in complex contexts such as cancer clinical trials. Therefore, this study identifies varied practices that have been used to communicate healthcare information in a way that minimizes the use of medical terminology and provides explanations of complex concepts based on patient prior knowledge and experience. As such, this study lays the groundwork for further observational work to build guidelines for spoken plain language that reflect clinical reality in a variety of settings.

The plain language practices discussed here may contribute to effective information exchange in the context of cancer clinical trials. However, clinicians should use these and other communication practices with care and attention to the needs of individual patients and situations. Clinicians’ perceptions of patients’ knowledge are subject to stereotypes and biases [52,53], and some linguistic practices, such as metaphor, may not communicate effectively across social backgrounds. For example, research has shown that rural, low-income women in Appalachia had negative perceptions of “coin flip” metaphors due to the association with risky gambling behaviors [54]; and some metaphors, such as describing depression as “feeling down,” have been shown to contribute to miscommunication across cultures [55]. Therefore, physicians should use plain language alongside other patient-centered communication practices such as ask-tell-ask [56,57] or teach-back [58,59] to assess and verify patient understanding.

Our findings should be considered within the study’s limitations. We conducted a secondary analysis of data from an intervention study designed to improve patient-physician communication and increase clinical trial participation. First, it is possible that the intervention influenced patients’ communication and prompted the physicians to adopt new communication practices. However, the eight examples of plain language presented here were taken from different interactions, and only three included patients who received the intervention. The presence of plain language practices in interactions from both groups suggests that the intervention was not a primary contributor to the practices. Second, the parent study did not collect data about physicians’ previous communication training. Therefore, it is not possible to determine the extent to which their prior training influenced the plain language practices identified here. Third, the descriptive, qualitative methodology and small sample did not allow us to examine the frequency of the practices or to compare them across patient or physician characteristics or clinic sites. However, the results as presented reflect best practices in qualitative research [60] and discourse analysis [35], which allow for thick description and theory generation that can lay the groundwork for larger studies. Fourth, we studied physician language rather than focusing on additional members of the health care team because physicians generally make treatment recommendations and initiate clinical trial discussions and invitations during interactions with patients. Fifth, one physician appeared in six transcripts (24% of the sample). Future research should expand on this work by considering factors such as patient and physician characteristics, other types of cancer, and the role of additional members of health care teams [27]. Finally, an important limitation is that the patient participants in this secondary analysis were relatively highly educated (75% had some college education) and had high incomes (48% had income above 80k). The sample was drawn from a larger, more diverse sample [61], but included only those patients who received an explicit invitation to participate in an available trial. This sample, therefore, reflects the unfortunate fact that clinical trial enrollment is not equitable across patient populations [62]. To further address patients’ varied health literacy needs, future research on spoken plain language in patient-physician discussions of cancer clinical trials should expand on this sample to improve generalizability.

Future research could also determine the extent to which these plain language practices support effective information exchange and informed consent. For example, future studies could compare plain language practices to determine which are the most effective for improving patient understanding of medical concepts, as has been done in prior research on metaphor and plain language [44]. Alternative studies could test practices across patient and physician demographics to determine their effectiveness within various contexts, similar to prior research on metaphors for randomization [54]. Finally, future research could also expand on the setting and sample for this observational work to examine plain language in a wider range of clinical contexts.

4.2. Conclusion

The current study identified plain language practices that physicians used to replace or explain medical terminology in discussions with patients about cancer clinical trials: lexical simplification, audience-centered definition, metaphor, and experience-focused description. Research shows the use of plain language improves patient-physician communication and responds to patient health literacy needs, but most of this research analyzes written communication and rarely in the context of clinical trials. By expanding this focus, our findings may contribute to patients’ ability to understand and make informed decisions about cancer clinical trials.

We advance the understanding of how clinical trials might be clearly explained to patients by examining physicians’ spoken plain language. We do this through an observational analysis of naturally occurring cancer clinical trial discussions. We demonstrate that a range of plain language practices are used to present information about clinical trials to patients. These data are useful as we begin assessing the effectiveness of these practices and move toward intervention development and testing.

4.3. Practice implications

Spoken plain language is critical to effective communication about clinical trials because it holds promise as a way to meet the informational needs of patients with varying levels of health literacy. The practices presented here represent a range of approaches that physicians have used to clarify medical terminology and concepts. Pending further outcomes-based testing, these practices might also be integrated into future physician training on the communication of information related to cancer clinical trials. By offering multiple practices for replacing or explaining medical or technical terminology, this training can offer physicians more effective means to communicate in ways that meet the needs of patients with diverse backgrounds and facing difficult decisions associated with complex medical information.