Use of Persuasive Language in Communication of Risk during Prostate Cancer Treatment Consultations

Abstract

Background

Physician treatment preference may influence how risks are communicated in prostate cancer consultations. We identified persuasive language used when describing cancer prognosis, life expectancy, and side effects in relation to a physician’s recommendation for aggressive (surgery/radiation) or nonaggressive (active surveillance/watchful waiting) treatment.

Methods

A qualitative analysis was performed on transcribed treatment consultations of 40 men with low- and intermediate-risk prostate cancer across 10 multidisciplinary providers. Quotes pertaining to cancer prognosis, life expectancy, and side effects were randomized. Coders predicted physician treatment recommendations from isolated blinded quotes. Testing characteristics of consensus predictions against the physician’s treatment recommendation were reported. Coders then identified persuasive strategies favoring aggressive/nonaggressive treatment for each quote. Frequencies of persuasive strategies favoring aggressive/nonaggressive treatment were reported. Logistic regression quantified associations between persuasive strategies and physician treatment recommendations.

Results

A total of 496 quotes about cancer prognosis (n = 127), life expectancy (n = 51), and side effects (n = 318) were identified. The accuracy of predicting treatment recommendation based on individual quotes containing persuasive language (n = 256/496, 52%) was 91%. When favoring aggressive treatment, persuasive language downplayed side effect risks and amplified cancer risk (recurrence, progression, or mortality). Significant predictors (P < 0.05) of aggressive treatment recommendation included favorable side effect interpretation, downplaying side effects, and long time horizon for cancer risk due to longevity. When favoring nonaggressive treatment, persuasive language amplified side effect risks and downplayed cancer risk. Significant predictors of nonaggressive treatment recommendation included unfavorable side effect interpretation, favorable interpretation of cancer risk, and short time horizon for cancer risk due to longevity.

Conclusions

Physicians use persuasive language favoring their preferred treatment, regardless of whether their recommendation is appropriate.

Implications

Clinicians should quantify risk so patients can judge potential harm without solely relying on persuasive language.

Highlights

Physicians use persuasive language favoring their treatment recommendation when communicating risks of prostate cancer treatment, which may influence a patient’s treatment choice.

Coders predicted physician treatment recommendations based on isolated, randomized quotes about cancer prognosis, life expectancy, and side effects with 91% accuracy.

Qualitative analysis revealed that when favoring nonaggressive treatment, physicians used persuasive language that amplified side effect risks and downplayed cancer risk. When favoring aggressive treatment, physicians did the opposite.

Providers should be cognizant of using persuasive strategies and aim to provide quantified assessments of risk that are jointly interpreted with the patient so that patients can make evidence-based conclusions regarding risks without solely relying on persuasive language.

Keywords

shared decision making risk communication prostate cancer qualitative analysis

Shared decision making (SDM) is now regarded as standard of care for treatment decision making in numerous oncologic practice guidelines, including prostate cancer.¹ SDM has been described as “an approach where clinicians and patients share the best available evidence when faced with the task of making decisions, and where patients are supported to consider options, to achieve informed preferences.”^2,3 For the physician, high-quality SDM requires skillful integration of evidence-based medicine and patient-centered communication,⁴ accurate portrayal of evidence, solicitation of patient preferences, and minimization of inherent biases.⁵ There may be no disease state in which SDM is more relevant than prostate cancer, given its complex tradeoffs of cancer control and side effects.⁶ The mainstay treatment options for clinically localized prostate cancer—active surveillance, radical prostatectomy, and radiation therapy—are supported by level I evidence demonstrating no differences in survival at 15 y following initial diagnosis, allowing physician equipoise in recommending the various options.⁷ However, rates of metastasis on active surveillance are double that of surgery and radiation,⁷ while surgery or radiation incur the risk of side effects such as urinary incontinence and erectile dysfunction that can meaningfully impact quality of life (QOL).^8,9 Given the need to carefully weigh these tradeoffs, prostate cancer treatment guidelines universally endorse SDM as the standard of care.¹⁰

Success of prostate cancer SDM relies on accurate portrayal of risks and benefits of treatment options. In treatment consultations, physicians must convey 3 components of risk: life expectancy, cancer prognosis, and side effects. Life expectancy is critical for gauging whether the patient has sufficient longevity to benefit from treatment.^11,12 Cancer prognosis (both with and without treatment) is central to understanding risks of progression, metastasis, and disease-specific mortality and how treatment modulates these risks.^13–15 Risks of treatment-related side effects define negative consequences of treatment and long-term QOL.⁹ Together, information on life expectancy, cancer prognosis, and side effects allows the patient to weigh risks and rewards of treatment to determine whether treatment is preferred, and if so, which treatment maximizes their values and preferences.

Numerous studies have shown that physician characteristics—specialty, treatment setting, sociodemographics—strongly influence patient treatment choice, despite adjusting for patient- and tumor-related factors.^16,17 One possible explanation for this is that physicians may bias how patients perceive treatment tradeoffs, presumably through communication of risk during treatment consultations. Our previous work has shown that in prostate cancer consultations, the risks of cancer prognosis, life expectancy, and side effects are frequently described using either generalizations (“high/low”) or are omitted entirely,^18,19 which dampens patients’ risk perception^20,21 and allows physicians to control risk perception using persuasive language. Indeed, two previous meta-analyses showed that failure to quantify risk in consultations is associated with poorer patient risk perception.^20,21 We also recently found that patients overwhelmingly prefer quantitative assessments of longevity (88%) in prostate cancer consultations, because as one patient noted, generalizations of risk “don’t give patients enough credit” to be part of SDM. This sentiment rings true, since if physicians fail to quantify risk during consultations, the persuasive language used by physicians to describe these risks becomes the sole source of information for the patient to gauge their magnitude and acceptability.

In this study, we sought to identify persuasive language used when discussing life expectancy, cancer prognosis, and side effects in relation to a physician’s recommendation for aggressive or nonaggressive treatment. We aimed to better understand how physicians use language to shape risk communication to convince patients to treat or not treat their cancer, regardless of whether their recommendation would be regarded as appropriate. In the context of this study, we defined persuasive language broadly, inclusive of rhetorical constructs (e.g., use of emotive language, comparing and contrasting, repetition, using anecdotal evidence), as well as intentional inclusion or omission of concepts and quantitative data. We hypothesized that physicians employ such language in the form of generalizable persuasive strategies that reflect their treatment recommendation. Identifying persuasive strategies favoring aggressive or nonaggressive treatment will offer insight into how physician treatment preference may influence patient treatment choice. In addition, awareness and mitigation strategies to prevent overreliance on persuasive language can also present an opportunity for quality improvement in SDM.

Materials and Methods

Study Cohort

We recruited 40 men undergoing initial treatment consultation for Gleason ≤ 7, clinical stage ≤ T2c prostate cancer. We specifically chose to focus on men with a Gleason ≤ 7, clinical stage ≤ T2c prostate cancer in our study because these cancers have a wide spectrum of guidelines-endorsed treatment options ranging from conservative management to aggressive local therapy^22–24 and exhibit substantial heterogeneity of treatment in practice. These cancers, including very low, low, favorable intermediate, and unfavorable intermediate American Urological Association tumor risk categories,²⁴ have a relatively narrow range of cancer mortality observed among patients with similar disease, as evidenced by the PIVOT trial, with 6.6% to 8.6% cancer mortality at 19.5 y for watchful waiting.¹⁵ In addition, this range of disease stage represents prostate cancers for which the American Urological Association/American Society of Clinical Oncology/American Society for Radiation Oncology guidelines recommend SDM in determining a patient’s treatment plan.^5,24 We excluded patients < 18 y of age and non-English speakers. To obtain a maximum variation sample capturing a range of specialties, we recruited 4 medical oncologists, 4 urologists, and 2 radiation oncologists from our institution, a tertiary referral center. The study was approved by the Institutional Review Board (Pro#00053972).

Consultation Coding: Persuasive Strategies

Treatment consultations were digitally recorded and transcribed verbatim. We extracted physician quotes related to life expectancy, cancer prognosis, and side effects. Quotes were segmented to provide sufficient context to understand discrete concepts being described within the body of text. Quotes were randomized and reviewers were blinded to any context, including quotes from the same consultation and the physician’s treatment recommendation. Three coders (R.G., T.J.D., J.M.) were asked to predict whether the physician recommended aggressive (surgery, radiotherapy, high-intensity focused ultrasound, cryotherapy) or nonaggressive (active surveillance, watchful waiting) treatment based on each isolated, randomized quote. Quotes deemed purely informational by reviewer consensus were excluded.

Five coders (R.G., T.J.D., A.N.-T., M.L., A.V.) independently used an open coding approach to identify persuasive strategies used to favor aggressive or nonaggressive treatment for each quote. Upon completion of individual open coding, coders then met to create a consensus codebook of persuasive strategies, which was done to improve data reliability. The entire library of quotes was then recoded using the consensus codebook. Coders then met to adjudicate any discrepancies in consensus coding. Persuasive strategy codes were then aggregated within content categories (life expectancy, cancer prognosis, and side effects). Throughout the article, cancer risk refers to the risk of cancer recurrence, progression, or mortality, while tumor risk refers to the tumor risk categorizations as noted in treatment guidelines.¹⁰

We estimated that at least 20 patient consultations would allow for thematic saturation based on experience with our prior studies.^18,19 Patients were recruited until thematic saturation was reached, under which analysis of additional patients yielded no new codes (i.e., persuasive strategies). Of the total cumulative codes, 50%, 75%, and 90% of codes (persuasive strategies) were captured by subjects 3, 12, and 18, respectively (Supplementary Figure 1). Physician recruitment and patient enrollment then continued until a higher level of balance in terms of tumor risk and physician specialty was achieved.

Consultation Coding: Treatment Recommendations

We extracted physician quotes related to physician treatment recommendations from transcripts. Quotes from each consultation were grouped together. Two coders (R.G., T.J.D.) independently analyzed quotes to identify whether the physician was recommending for or against aggressive treatment or was inconclusive. Coders met to establish consensus for each quote. If more than two-thirds of quotes from a given consultation favored aggressive or nonaggressive treatment, then this was identified as the physician’s treatment recommendation. If this threshold was not met, the recommendation was labeled as equivocal. A sensitivity analysis comparing varying thresholds (>66% v. >75% v. >85%) found little change for designation of treatment recommendations.

Statistical Analysis

Testing characteristics (accuracy, sensitivity, specificity, positive/negative predictive value) between coders’ consensus predictions from blinded quotes and the physician’s treatment recommendation were calculated among quotes from consultations in which aggressive or nonaggressive treatment was recommended.

Frequencies of consensus persuasive strategies used to promote aggressive and nonaggressive treatment were reported across subgroups of content categories (life expectancy, cancer prognosis, and side effects). The numerator and denominator for each content category were the frequency a given persuasive strategy was used on a per-quote level and the total frequency of quotes for each category, respectively. Univariate logistic regression identified which persuasive strategies were associated with aggressive or nonaggressive treatment recommendations at the quote level. Quantitative analyses were performed in R (version 3.5.1; R-Foundation) using 2-sided tests with a significance level of P < .05.

Results

Sample characteristics were typical of a US low- and intermediate-risk prostate cancer population (Table 1). We identified 496 unique quotes related to cancer prognosis (n = 127), side effects (n = 318), and life expectancy (n = 51). On average, quotes were composed of approximately 80 words and 5 sentences. The majority of quotes (n = 284) were associated with consultations recommending aggressive treatment (n = 18) and the minority of quotes (n = 73) with consultations recommending nonaggressive treatment (n = 10). The remaining quotes (n = 139) were associated with consultations with an equivocal treatment recommendation (n = 12). For aggressive or nonaggressive treatment recommendations, there was 100% concordance between patient treatment choice and physician treatment recommendation.

Table 1

Population Characteristics

Characteristic	All Patients (N = 40)
Age, y, median (IQR)	67 (61, 73)
Race, n (%)
Black	7 (17)
White	33 (83)
Clinical T-stage, n (%)
T1a-b	1 (2)
T1c	35 (88)
T2a-b	2 (5)
T2c	2 (5)
PSA, $\bar x$ (s)	8.77 (5.57)
Diagnostic Gleason score, n (%)
3+3	10 (25)
3+4	19 (48)
4+3	11 (27)
Prostate Cancer Comorbidity Index score, median (IQR)	2 (0, 4)
Prostate Cancer Comorbidity Index score, n (%)
0	12 (30)
1–2	14 (35)
3–4	5 (13)
5–6	4 (10)
7–9	4 (10)
10+	1 (2)
D’Amico tumor risk, n (%)
Very low risk	5 (12)
Low risk	4 (10)
Favorable intermediate risk	13 (33)
Unfavorable intermediate risk	16 (40)
High risk	2 (5)
Treatment type, n (%)
Watchful waiting	4 (10)
Active surveillance	15 (38)
Radical prostatectomy	4 (10)
Radiation therapy	6 (15)
HIFU	2 (5)
Undecided/not recorded	9 (22)

HIFU, high-intensity focused ultrasound; IQR, interquartile range; PSA, prostate-specific antigen.

Coders predicted whether the physician was recommending aggressive or nonaggressive treatment based on an analysis of blinded, randomized, individual quotes with 91% accuracy. When coders predicted aggressive treatment, the physician recommendation was aggressive in 96% (186/194) of predictions (positive predictive value). Among quotes that were associated with an aggressive treatment recommendation, coders correctly identified 92% (186/203) of quotes (sensitivity). When coders predicted nonaggressive treatment, the physician recommendation was nonaggressive in 73% (45/62) of predictions (negative predictive value). Among quotes that were associated with a nonaggressive treatment recommendation, coders correctly identified 71% (45/63) of quotes (specificity). Fleiss’s Kappa test for interrater reliability was 0.9 (95% confidence interval [CI] 0.84–0.94) for coder predictions.

Frequencies of persuasive strategies used to promote an aggressive or nonaggressive treatment recommendation among quotes related to cancer prognosis, life expectancy, and side effects are shown in Table 2. Quotes illustrating common persuasive strategies are shown in Table 3 and Supplementary Table 1.

Table 2

Frequency of Persuasive Strategy Use Based on Discussion Content and Physician Treatment Recommendation for Prostate Cancer (Strategies Used in at Least 5% of Quotes per Category)

Physician Recommendation
Content Category	Nonaggressive Treatment	Frequency of Use	Aggressive Treatment	Frequency of Use
Cancer prognosis	Favorable interpretation of cancer risk	(23/32) 72%	Direct recommendation for aggressive treatment	(14/48) 29%
	Superlative adjectives describing low cancer risk	(14/32) 44%	Quoting literature/guidelines	(13/48) 27%
	Reassurance	(9/32) 28%	Long time horizon for risk due to longevity	(11/48) 23%
	Quoting literature/guidelines	(8/32) 25%	Opportunity for cure	(9/48) 19%
	Repetition	(6/32) 19%	Doubt apprehension concern about nonaggressive treatment	(9/48) 19%
	Direct recommendation for nonaggressive treatment	(6/32) 19%	Unfavorable interpretation of cancer risk	(8/48) 17%
			Express urgency with regard to timing	(7/48) 15%
			Bellicose/fear terms	(6/48) 13%
Life expectancy	Leveling with the patient	(2/10) 20%	Long time horizon for risk due to longevity	(18/25) 72%
	Life expectancy as a complicating factor	(2/10) 20%	Superlative adjectives describing long life expectancy	(8/25) 32%
	Exceptional case	(2/10) 20%	Exceptional case	(7/25) 28%
	Direct recommendation for nonaggressive treatment	(2/10) 20%	Repetition	(5/25) 20%
	Balancing quality of life and longevity	(2/10) 20%	Leveling with the patient	(4/25) 16%
			Life expectancy as a complicating factor	(3/25) 12%
			Direct recommendation for aggressive treatment	(3/25) 12%
	Unfavorable interpretation of side effects	(11/31) 36%	Favorable interpretation of side effects	(64/211) 30%
	Short longevity and baseline risk predict worse side effects	(10/31) 32%	Downplaying side effects	(55/211) 26%
Side effects	Emphasis on quality of life	(8/31) 26%	Superlative adjectives describing favorable side effects	(31/211) 15%
	Downplaying side effects	(7/31) 23%	Youth as indicator for low side effect profile	(23/211) 11%
	Large scope of procedure/recovery	(4/31) 13%
	Guarantee	(4/31) 13%

Table 3

(A) Illustrative Examples of Persuasive Strategies That Were Significantly Associated with Aggressive Treatment Recommendations (Surgery or Radiation)

Aggressive Treatment
Persuasive Strategy	Definition	Illustrative Quotes
Favorable interpretation of side effects	Interpreting or contextualizing risk of treatment-related side effects data in a favorable manner	“The likelihood that you are going to regain your baseline potency is—is pretty high. So, being 62 [y old] and having perfect erections now, I’d say you’re 80% likely of getting to that at a year. Now, there’s going to be a recovery period, but you’ll likely regain those erections.” “I think you’re highly likely to have good continence long term. At a year, you would regain your continence. There’s about a 10% likelihood you’d need one pad daily going forward, but it’s a low risk and 90% of people recover their continence fully . . . so I would say the odds are good for that.”
Downplaying side effects	Minimizing the impact on quality of life or risks of side effects associated with prostate cancer treatment	“But those are fairly infrequent events . . . on the order of, you know, a handful percent of the time that happens, but it’s possible. I’ll tell you that the vast majority of people tolerate the surgery just fine, even though it is a big surgery.” “I look at [incontinence] as a temporary inconvenience because beyond a year about 10% of men need a pad and the rest are totally continent.”
Superlative adjectives describing favorable side effects	Use of superlative adjectives or adverbs that emphasize extremely low likelihood of an unfavorable side effect outcome	“The likelihood that you’re completely wet, just can’t control anything, really, really, really low. It’s not zero, but it’s really, really low.” “Radiation can potentially damage the rectum, [but] very, very low risk.”
Youth as indicator for low side effect profile	Emphasis on the young age of a patient to suggest superior outcomes	“The complications are minimized in young men. You’re sort of the setup for who would do well with surgery.” “But again, being young and healthy, you’re going to be on the lower end of those numbers.” “So, you are super young and your erectile function is excellent. So the likelihood you’re going to regain your baseline erectile function is very high.”
Long time horizon for risk due to longevity	Indicating that the patient is more susceptible to harm from cancer risk given a relatively longer life expectancy	“. . . when we diagnose cancers very early in life, we like to be very aggressive with them because I’m planning for you for the next 45 y until you’re 90.” “. . . you’ve got 20+ y ahead of you. I hope you live to 100, but you know, on average, guys who are 62 are going to live like 20, 25 y, right? So, I wanna plan for that to make sure that the cancer doesn’t harm you.” “The likelihood that it threatens your life over like a 20-y period is going to be around 10%, so not high—10–15% over 20 y. Not high, but high enough to make me worry because you’ve got 40–50 y to go.”
Positive anecdote	Sharing favorable experiences of previous patients	“In fact, all the patients that I’ve operated on who have been in their 50s . . . all uniformly did well. They’re continent right out of the gate and they have erectile function pretty soon.” “So every 50-y-old I’ve operated on has had recovery of continence very quickly compared to older patients like in their 60s or 70s.” “In the series of patients that we’ve done . . . almost all the men have had excellent urinary health. Almost all of them have remained potent if they were potent before the surgery, and most importantly to me, all of them have had excellent cancer control.”

Table 3

(B) Illustrative Examples of Persuasive Strategies That Were Significantly Associated with Nonaggressive Treatment Recommendations (Active Surveillance or Watchful Waiting)

Nonaggressive Treatment
Persuasive Strategy	Definition	Illustrative Quotes
Unfavorable interpretation of side effects	Interpreting or contextualizing data of treatment-related side effects risk in an unfavorable manner	“The likelihood that you’ll return to your baseline erectile function I would say is around 40–50%, so it’s not good odds with this. Urinary continence I can give you good odds on that but for return of potency, being 71, I would say the odds are worse.” “Well, radiation causes side effects. It’ll worsen your urgency symptoms, frequency of going. Waking up at night will get worse. And it can cause other issues like blood in the urine and in the stool. So there are real costs to radiation. You don’t do it just to do it.”
Old age/poor health/lower baseline as an indicator of worse side effects	Referencing the patient’s age, comorbidities, and poor functional baseline as indicators of a worse side effect profile following aggressive treatment	“So the impact of surgery on you function-wise would be worse than if I did it on a 50-y-old. So most 50-y -olds I operate on are continent and potent, by you know, a shorter time period, every one. You, if I operate on you, you’ll have a longer trajectory curve for recovery of those functions and potentially you’ll never recover your potency.” “We don’t generally do surgery in patients who are over the age of 70 because the surgery has risks, and we don’t feel that the risk is generally warranted in patients who are older because they tend to derive less benefit from the operation.”
Emphasis on quality of life	Focusing on details involving quality of life including anecdotal scenarios	“Anything we do, surgery, radiation, chemotherapy, all of that, is likely to negatively affect your quality of life and very unlikely to improve your quantity of life.” “But it’s a special thing to be able to enjoy that intimacy with someone you love and to have that taken away from you. I mean, your efforts to fight cancer, while it can be an acceptable loss, if it can be avoided without compromising your health, it’s important.”
Life expectancy as a complicating factor	Questioning whether the patient will live long enough to benefit from the delayed survival benefits of aggressive treatment	“Metastasis within 10 y is 1%, and the adverse pathology is only 14% if we were to take it out, but the other thing to consider here is how long are you going to live? What is your life expectancy in the context of all this?” “And with radiation therapy, it doesn’t necessarily cause a lot of leakage of urine, but it causes overactivity of the bladder: urgency, frequency, nocturia. It can cause bleeding in the urine and in the stool, bowel problems. There are side effects that come with these therapies that are worth considering, especially when you’re 88 [y old] and this disease might not, is highly unlikely to harm you during your lifetime.”
Balancing quality of life and longevity	Suggesting that a patient with short life expectancy should prioritize quality of life over dubious longevity gains given short life expectancy	“And the likelihood it would threaten your life is like less than 5%, something like that. So there is some risk there, but the risk is really low. And I think the key is maximizing your quality of life.” “We know that the older you get the longer you are likely to live so part of what I have to think about doing is not making matters worse for you when it comes to prostate cancer.”
Favorable interpretation of cancer risk	Interpreting or contextualizing data on cancer risk/prognosis in a favorable manner	“14% [cancer mortality risk] seems like a number that’s high compared to the <1% or 1% of risk in terms of metastasis and prostate cancer death that this readout gives you, but it’s a very low number. It’s a very low number.” “. . . And based on their calculator, at 10 y from now, only 1 person—if you didn’t treat the cancer with anything—would be killed by the cancer, 1 out of 100. So those are pretty good odds.”
Superlative adjectives describing low cancer risk	Use of superlative adjectives or adverbs that emphasize extremely low likelihood of an unfavorable cancer outcome	“It’s not zero, so we don’t want to ignore it. We do want to monitor it and then try to intervene if it starts to behave aggressively in the future, because there is some chance that this could kill you. It’s not zero, but it’s also very low.” “I would put this in the very-low-risk category. And so what that means is the likelihood this tumor is going to go, and spread, and cause you harm, at least in your lifetime, is really, really low.”
Reassurance	Addressing or assuaging patient anxiety regarding their disease, cancer risk, life expectancy, and side effect profile	“I tell you, I get all these guys who come in here and they think that they’re about to die. That’s really not where you are.” “Medically speaking, you’re not about to die from all this.” “So hopefully that’s a little reassuring. [If] it was me in your seat, I would be reassured that my chance of dying of prostate cancer is extraordinarily low for at least 10 y and probably way more.”
Short time horizon for risk due to longevity	Indicating that the patient is less susceptible to harm from cancer risk given a relatively short life expectancy	“Well, you’re much more likely to die of something else other than the prostate cancer over this time. But there is a risk of the prostate cancer harming you. So what’s the risk of the prostate cancer threatening your life? Well, it’s about 6% to 7% over 10–15 y if we did nothing about it. Now, if we did something about it, we could cut that risk down in, let’s say, in ½ or in ¾, down to a handful percentage” “You are so much more likely to live with your cancer your whole life and not have it bother you.”
Direct recommendation for nonaggressive treatment	Direct or explicit physician recommendation in favor of nonaggressive treatment	“And in patients . . . who are on active surveillance, meaning monitoring with serial PSA checks and biopsies every so often, the likelihood that it would kill you approaches zero over like a 15-, 20-y period, which is the range of years that we’re talking about here. So I think that it’s, in my mind, it’s the right choice.” “And so, in general, with these types of very favorable intermediate early-stage risks, watching it alone and just kind of keeping an eye on it, surveillance, has actually proven to be a really, really reasonable and safe outcome.”
Direct recommendation against aggressive treatment	Direct or explicit physician recommendation against aggressive treatment.	“So I’m going to come back and we’re going to talk briefly about the risks of surgery, not because I’m recommending surgery, but so that it will be obvious to you when you listen to the risks of treatment that the risk reduction, what you’re paying and what you’re getting in return, isn’t worth it.” “So, again, this is a fairly low number, and that’s why at this point, you know, treating this with surgery and radiation, which can have lifelong consequences and side effects is too aggressive.”

Bolded excerpts represent specific phrases within the quotes that correspond with the consensus persuasive strategy. PSA, prostate-specific antigen.

When recommending aggressive treatment, physicians frequently used persuasive strategies that emphasized cancer risk (i.e., risk of cancer mortality, progression, and/or recurrence), highlighted long expected life expectancy, and downplayed risks of side effects (Tables 2 and 3A). For example, when describing cancer prognosis, common persuasive strategies centered around amplification of cancer risk (long time horizon of cancer risk due to longevity [23%], time-sensitive window of opportunity for cure [19%], doubt/apprehension/concern about safety of nonaggressive treatment [19%], unfavorable interpretation of cancer risk [17%]). When describing life expectancy, physicians used strategies that emphasized the patient’s long life expectancy and time horizon to incur cancer risk (time horizon for cancer risk related to the patient’s expected longevity [72%], superlatives describing the patient’s long life expectancy [32%], exceptional case [in terms of long life expectancy; 28%]). When describing side effects, physicians frequently used persuasive language that downplayed treatment-related side effects (favorable interpretation of side effects [30%], rationalizing side effects [28%], superlative adjectives describing favorable side effect outcomes [15%], youth as an indicator for low side effect profile [11%]).

When recommending nonaggressive treatment, physicians frequently used persuasive strategies that downplayed cancer risk and emphasized QOL and the short time horizon for risk due to limited life expectancy while negatively characterizing the risks of side effects (Tables 2 and 3B). When describing cancer prognosis, in contrast to aggressive treatment, the most common strategies centered around minimizing cancer risk (favorable interpretation of cancer risk [72%], superlative adjective describing low cancer risk [44%], reassurance about the low risk of cancer [28%]). When describing life expectancy, common strategies included highlighting the short time horizon for risk due to limited life expectancy (leveling with patient [20%], life expectancy as a complicating factor [20%], exceptional case due to short life expectancy [20%]) or maximizing QOL [balancing QOL and longevity; 20%]). When describing side effects, physicians frequently negatively characterized treatment-related side effects (unfavorable interpretation of side effects [36%], limited life expectancy as an indicator for worse side effect profile [32%], large scope of procedure/recovery [13%]), and focused on maximizing overall QOL (26%).

In univariate logistic regression (Figure 1), significant predictors of an aggressive treatment recommendation at the quote level included 1) youth as an indicator for low side effect profile, 2) positive anecdote, 3) favorable interpretation of side effects, 4) superlative adjectives describing favorable side effect profile, 5) downplaying side effects, and 6) long time horizon for cancer risk due to longevity (all P < 0.05). Significant predictors of a nonaggressive treatment recommendation included 1) direct recommendation for nonaggressive treatment, 2) direct recommendation against aggressive treatment, 3) superlative adjectives describing favorable cancer risk, 4) favorable interpretation of cancer risk, 5) short time horizon for cancer risk due to longevity, 6) emphasis on QOL, 7) life expectancy as a complicating factor, 8) balancing QOL and longevity, 9) unfavorable interpretation of side effects, 10) old age/poor health/lower baseline QOL as an indicator of worse side effects, and 11) reassurance (all P < 0.05).

Figure 1

Univariate associations between persuasive strategy use and physician treatment recommendation during prostate cancer treatment consultations. (A) Aggressive treatment recommendation. (B) Nonaggressive treatment recommendation.

Discussion

There has been a paradigm shift away from a paternalistic model toward SDM in prostate²⁵ and other cancers,^26–29 as evidenced by SDM being increasingly codified as a requisite for appropriate care in clinical guidelines.¹⁰ Despite this, the ideal balance of the physician in their multiple roles in SDM—as educator, expert opinion stakeholder, and agent of persuasion—is ill-defined. While it is expected that physicians communicate their expert opinions on treatment, the fundamental role of the physician as an educator on disease prognosis, treatment options, and treatment side effects should not be compromised to achieve the physician’s desired treatment choice, whether it is an appropriate treatment recommendation or not. Previous work has shown that physician characteristics are strongly associated with treatment choice despite correcting for patient and tumor characteristics,^16,17 suggesting that physicians play a significant role in patient perception of risks and rewards of therapy. Furthermore, prior literature has shown that evaluative labels (similar to generalizations) can have a wide range of interpretation by patients and skew risk perception.^20,21,30 In the absence of providing numerical risk estimates and jointly interpreting their significance with the patient, persuasive language used by physicians may have a substantial impact on how risks and rewards of therapy are perceived by patients.

In this study, we found that persuasive strategies favoring aggressive or nonaggressive treatment were consistently embedded in communication of factual content, such that the physician’s treatment preference could be perceived at the level of the individual quote. Even when analyzing a quote taken completely out of context, coders were able to correctly predict an aggressive or nonaggressive treatment recommendation with 91% accuracy. When favoring aggressive treatment, physicians tended to use persuasive strategies that downplayed the risks of side effects, emphasized the long time horizon for cancer risk due to long life expectancy, and amplified the risks of cancer. When favoring nonaggressive treatment, physicians tended to use persuasive strategies that amplified the risks of side effects, emphasized QOL and the short time horizon due to limited life expectancy, and downplayed cancer risk. These persuasive strategies were found to be significant predictors of an aggressive or nonaggressive treatment recommendation (odds ratio >1), although confidence intervals overlap and limit our ability to suggest any strategy as more predictive than another.

While risks of life expectancy, cancer prognosis, and side effects are static numbers, the interpretation of whether these numbers are favorable or unfavorable is key to whether the patient feels the risks are acceptable or unacceptable. Although the literature suggests that certain tumor risk factors may favor nonquantitative communication of risks,³¹ our previous work on variation in risk communication during prostate cancer treatment consultations (in a nested population within the current study) demonstrated that 88% of patients prefer patient-specific, quantitative assessments of longevity (an example of 1 key tradeoff in SDM) instead of generalities.³² These findings are in line with other studies championing quantification of risk to improve risk perception in cancer patients.^20,21,33,34 Despite this strong preference for quantification of risk, we found that life expectancy, cancer prognosis, and side effects are often not quantified and are described using either generalizations (“high/low”) or omitted entirely 34%, 31%, and 62% to 93% of the time, respectively, on a per consultation level.^18,19 In the absence of risk quantification, the persuasive language used to describe these risks becomes the sole source of information for the patient to gauge whether the risks are acceptable or not. While the patient’s health numeracy and literacy are important in determining how risk information should be optimally communicated (e.g., with visual depictions such as icon arrays better than probabilistic descriptions in low numerate individuals),³⁵ in these patients, persuasive language alone should not be considered an acceptable substitute for communication of risk. In fact, lower numerate individuals appear to be least capable of differentiating between generic verbal risk labels (similar to generalizations), which would make them extremely susceptible to persuasive language in shaping risk perception.³⁰ Even when risks are quantified, persuasive language influences how these risks are interpreted by the patient as part of the patients “measured vulnerability,” or self-projection of current and future health.³⁶

In our study, when describing cancer prognosis, physicians frequently used persuasive strategies to downplay the risk of cancer mortality when they were advocating for nonaggressive treatment and, conversely, amplified mortality risks when advocating for aggressive treatment. For example, when a physician was recommending nonaggressive treatment, they often promoted this recommendation by “favorable interpretation of cancer risk” in describing cancer prognosis: “As long as we don’t totally ignore this, the risk of death from this prostate cancer is very low.” Without quantitative data, this language provides both interpretation and judgment of the numbers without actually providing them. This tends to lessen the patient’s role in judging tradeoffs, since there is uncertainty around what “very low” means, presumably to support why observation is justified. A separate example of this same persuasive strategy actually quantifies cancer mortality risk but goes on to interpret the data for the patient: “14% [cancer mortality risk] seems like a number that’s high . . . but it’s a very low number. It’s a very low number.” While this approach clearly avoids the perils of omitted quantification in the previous example, it is unclear if the physician should interpret the risks as low or high, since individual patients may vary with regard to how much risk is acceptable to them. Perhaps an ideal approach would include jointly interpreting whether 14% is high or low in conjunction with the patient.

It is worth noting that quantification of risk in and of itself does not eliminate the possibility of bias. For example, physicians may present statistics that tend to favor their argument, even if presented objectively. In addition, the incidence of an event can be quantified in either a positive or negative fashion, which may affect patient interpretation. For instance, a “14% chance of having a side effect” versus “86% chance of not having a side effect” may be perceived differently. Despite this, inviting patients to jointly interpret these risks from their perspective may help to minimize (but not eliminate) the impact of physician bias on reporting of risk.

While it is intuitive that some cancers have higher risk than others in our sample, we do not believe that this explains the observed range of persuasive strategies used to describe cancer prognosis. In situations in which cancer prognosis was nearly identical, physicians used persuasive language to interpret data in opposite ways, thereby matching their treatment recommendation. For example, in contrast with the preceding quote noting 14% cancer mortality risk as “a very low number,” a quote from a different consultation recommending nonaggressive treatment communicated unfavorable interpretation of nearly identical cancer risk: “The likelihood that your prostate cancer would kill you at, you know, 20 y, is like 15%. I mean these are scary statistics, right?” Moreover, we restricted recruitment to focus tumor risk (including very low, low, favorable intermediate, and unfavorable intermediate risk) to a narrow range of cancer mortality, 6.6% to 8.6% at 19.5 y based on the watchful waiting arm of the PIVOT¹⁵ trial. Furthermore, frequencies of persuasive strategies used in the context of aggressive or nonaggressive treatment recommendations did not differ substantially when analyzed in tumor risk subgroups.

Physicians also frequently used language to downplay or amplify of risks of side effects based on their treatment recommendation. Similar to cancer prognosis, persuasive language was occasionally substituted for quantification of risk, as in the following example of “favorable interpretation of side effects” favoring aggressive treatment: “I’m not going to improve your life with this surgery, but there’s a high likelihood that I’m not going to worsen it. [Urinary incontinence is] just a transient inconvenience. The leakage is—I look at that as kind of—a bother for some months.” Conversely, when recommending nonaggressive treatment, physicians used imagery to negatively portray risks of side effects: “You go off to help your son move some things around and you pick up the box and you get a little wet in the underpants. You get on that bike, bend forward and the urine starts dripping down your leg. That’s not a good feeling.” Similarly to cancer prognosis, even when the side effects risks are quantified, they can be interpreted as high or low using persuasive language. For example, in describing a 20% risk of erectile dysfunction, a physician recommending aggressive treatment says, “So, being 62 [y old] and having perfect erections now, I’d say you’re 80% likely of getting to that at a year. Now, there’s going to be a recovery period, but you’ll likely regain those erections.” Conversely, a physician recommending nonaggressive treatment says, “After 2 to 2½ y, if they haven’t recovered, the erectile dysfunction is permanent. The permanent erectile dysfunction occurs in 10% to 20% of patients. So these are the 2 really big life altering, QOL altering complications.” Again, an ideal approach might include jointly interpreting whether these risks are high or low in conjunction with the patient.

It is critical to note that the use of persuasive language and its resultant bias can occur even in the presence of an appropriate treatment recommendation. For instance, a radical prostatectomy can be appropriately recommended to an unfavorable intermediate-risk patient, although persuasive language was used to amplify the risks of cancer (e.g., “it’s a scary cancer”). Similarly, active surveillance can be appropriately recommended to a low-risk patient while persuasive language was used to amplify side effects of aggressive treatment (e.g., “you’ll end up incontinent and impotent with radical prostatectomy”). To address this, physicians should aim to root risk communication in data that allow patients to make conclusions about risks on their own without overly relying (or solely relying) on persuasive language. In the absence of awareness of this phenomenon, physicians may solely rely on persuasive language to characterize key risks, which prohibits the patient from making their own conclusions on whether risks are significant. These results do not suggest that physicians should not have a treatment preference or offer their opinion on treatment. However, since patients have autonomy to choose alternative treatments going against a physician’s recommendation, it is imperative that physicians should not substitute persuasive language for evidence-based estimates to sway a patient in favor of their treatment preference.

While we strongly feel that verbal articulation of risk is critical to appropriate risk perception by patients, the volume of verbally communicated risk information has the potential to overwhelm patients. Visual decision aids are a well-established adjunct to patient counseling that can reinforce or supplement information discussed during the consultation.^37–39 Best practices reported in the literature for communication of risk in visual decision aids parallel the best practices for verbal articulation of risk that we describe in this article. In a review of best practices for risk communication in decision aids, Bonner et al.⁴⁰ recommended “presenting numbers over a set time period with a clear denominator, using consistent formats between outcomes and interventions to enable unbiased comparisons, and interpreting the numbers for the reader to meet the needs of varying numeracy.” It is likely that a combination of verbal and visual information will provide the best opportunity for optimizing patient risk perception.

Prostate cancer provides an ideal paradigm for study of persuasive language in risk communication given its many treatment options, each with tradeoffs in cancer control and side effect profile.⁶ The mainstay treatment options for low- and intermediate-risk prostate cancer—active surveillance, radical prostatectomy, and radiation therapy—are supported by level I evidence demonstrating no differences in survival at 10 y following initial diagnosis.¹³ Yet despite this, the observed treatment patterns by specialty indicate a clear preference for aggressive or non-aggressive treatment based on provider specialty, hospital, and even provider identity within the specialty and same range of tumor risk.^16,17 Persuasive language in describing risks is highly likely to be a mediator of patient choice, since this is presumably a way providers motivate patients to choose one treatment option over another. We recommend that providers seek to provide evidence-based risk estimates and avoid using persuasive language alone to characterize risk, which allows patients the opportunity to judge risks and benefits for themselves. Further prospective study of these strategies within more focused populations of tumor risk and across multiple centers are needed to validate these findings. In addition, although our study focused on prostate cancer, the importance of transparent communication and quantitative risk portrayal is generalizable to other disease states, including those for which guidelines recommend SDM.

This study is subject to limitations. First, practices within our academic, tertiary referral setting may not be generalizable to other physicians of broader demographics, including age, gender, and practice setting (i.e., private practice, county hospitals, or Veteran Affairs Medical Centers). It is possible that differences in payment model, efficiency concerns, and/or patient demographics may affect physicians’ use of persuasive language. Second, the sample size of quotes recommending nonaggressive treatment was not as robust as compared with quotes recommending aggressive treatment, limiting our power for quantitative comparisons in this group. Third, variation in quantitative tumor risk across our patients may affect the persuasive strategies used when describing cancer prognosis (i.e., our patients did not all have uniform tumor risk). However, the frequencies of persuasive strategies used to promote aggressive or nonaggressive treatment were consistent within tumor risk subgroups. Fourth, our study was underpowered to identify appreciable differences in persuasive language across provider specialties.

Conclusions

A physician’s treatment preference appears to influence the way that risks of prostate cancer are described. When favoring nonaggressive treatment, physicians use persuasive language to amplify side effect risks and downplay cancer risk. When favoring aggressive treatment, physicians downplay side effect risks and amplify cancer risk. When cancer prognosis, life expectancy, and side effects are not quantified, this persuasive language acts as a substitute for the risk itself, since it becomes the sole source of information by which a patient judges that risk. This is true of situations in which the treatment recommendation is appropriate and those in which it is not appropriate. Since our previous work suggests that risks are often not quantified in consultations, clinicians should be cognizant of their use of persuasive language when communicating risk. Physicians should strive to provide quantified, precision estimates that are jointly interpreted with the patient as equal partners to fully engage in informed SDM.

Supplemental Material

sj-docx-1-mdm-10.1177_0272989X241228612 – Supplemental material for Use of Persuasive Language in Communication of Risk during Prostate Cancer Treatment Consultations

Supplemental material, sj-docx-1-mdm-10.1177_0272989X241228612 for Use of Persuasive Language in Communication of Risk during Prostate Cancer Treatment Consultations by Aurash Naser-Tavakolian, Rebecca Gale, Michael Luu, John M. Masterson, Abhishek Venkataramana, Dmitry Khodyakov, Jennifer T. Anger, Edwin Posadas, Howard Sandler, Stephen J. Freedland, Brennan Spiegel and Timothy J. Daskivich in Medical Decision Making

Supplemental Material

sj-pdf-2-mdm-10.1177_0272989X241228612 – Supplemental material for Use of Persuasive Language in Communication of Risk during Prostate Cancer Treatment Consultations

Supplemental material, sj-pdf-2-mdm-10.1177_0272989X241228612 for Use of Persuasive Language in Communication of Risk during Prostate Cancer Treatment Consultations by Aurash Naser-Tavakolian, Rebecca Gale, Michael Luu, John M. Masterson, Abhishek Venkataramana, Dmitry Khodyakov, Jennifer T. Anger, Edwin Posadas, Howard Sandler, Stephen J. Freedland, Brennan Spiegel and Timothy J. Daskivich in Medical Decision Making

Supplemental Material

sj-xlsx-3-mdm-10.1177_0272989X241228612 – Supplemental material for Use of Persuasive Language in Communication of Risk during Prostate Cancer Treatment Consultations

Supplemental material, sj-xlsx-3-mdm-10.1177_0272989X241228612 for Use of Persuasive Language in Communication of Risk during Prostate Cancer Treatment Consultations by Aurash Naser-Tavakolian, Rebecca Gale, Michael Luu, John M. Masterson, Abhishek Venkataramana, Dmitry Khodyakov, Jennifer T. Anger, Edwin Posadas, Howard Sandler, Stephen J. Freedland, Brennan Spiegel and Timothy J. Daskivich in Medical Decision Making

Supplemental Material

sj-xlsx-4-mdm-10.1177_0272989X241228612 – Supplemental material for Use of Persuasive Language in Communication of Risk during Prostate Cancer Treatment Consultations

Supplemental material, sj-xlsx-4-mdm-10.1177_0272989X241228612 for Use of Persuasive Language in Communication of Risk during Prostate Cancer Treatment Consultations by Aurash Naser-Tavakolian, Rebecca Gale, Michael Luu, John M. Masterson, Abhishek Venkataramana, Dmitry Khodyakov, Jennifer T. Anger, Edwin Posadas, Howard Sandler, Stephen J. Freedland, Brennan Spiegel and Timothy J. Daskivich in Medical Decision Making

Supplemental Material

sj-xlsx-5-mdm-10.1177_0272989X241228612 – Supplemental material for Use of Persuasive Language in Communication of Risk during Prostate Cancer Treatment Consultations

Supplemental material, sj-xlsx-5-mdm-10.1177_0272989X241228612 for Use of Persuasive Language in Communication of Risk during Prostate Cancer Treatment Consultations by Aurash Naser-Tavakolian, Rebecca Gale, Michael Luu, John M. Masterson, Abhishek Venkataramana, Dmitry Khodyakov, Jennifer T. Anger, Edwin Posadas, Howard Sandler, Stephen J. Freedland, Brennan Spiegel and Timothy J. Daskivich in Medical Decision Making

Footnotes

Authors’ Note

Abstract presented at American Urological Association Annual Meeting May 2022 in New Orleans, Louisiana.

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Financial support for this study was provided entirely by a by the Career Development Award (K08 CA230155 to Timothy J. Daskivich) from the National Cancer Institute. The funding agreement ensured the authors’ independence in designing the study, interpreting the data, writing, and publishing the report.

ORCID iDs

Aurash Naser-Tavakolian

Dmitry Khodyakov

References

Gärtner

Portielje

Langendam

, et al. Role of patient preferences in clinical practice guidelines: a multiple methods study using guidelines from oncology as a case. BMJ Open. 2019;9(12):e032483. DOI: 10.1136/bmjopen-2019-032483

Elwyn

Frosch

Thomson

, et al. Shared decision making: a model for clinical practice. J Gen Intern Med. 2012;27(10):1361–7. DOI: 10.1007/s11606-012-2077-6

Elwyn

Laitner

Coulter

Walker

Watson

Thomson

Implementing shared decision making in the NHS. BMJ. 2010;341:c5146. DOI: 10.1136/bmj.c5146

Hoffmann

Montori

Del Mar

The connection between evidence-based medicine and shared decision making. JAMA. 2014;312(13):1295–6. DOI: 10.1001/jama.2014.10186

Montori

Brito

Murad

MH.

The optimal practice of evidence-based medicine: incorporating patient preferences in practice guidelines. JAMA. 2013;310(23):2503–4. DOI: 10.1001/jama.2013.281422

Litwin

Tan

HJ.

The diagnosis and treatment of prostate cancer: a review. JAMA. 2017;317(24):2532–42. DOI: 10.1001/jama.2017.7248

Hamdy

Donovan

Lane

, et al. Fifteen-year outcomes after monitoring, surgery, or radiotherapy for prostate cancer. N Engl J Med. 2023;388(17):1547–58. DOI: 10.1056/NEJMoa2214122

Sanda

Dunn

Michalski

, et al. Quality of life and satisfaction with outcome among prostate-cancer survivors. N Engl J Med. 2008;358(12):1250–61. DOI: 10.1056/NEJMoa074311

Resnick

Koyama

Fan

, et al. Long-term functional outcomes after treatment for localized prostate cancer. N Engl J Med. 2013;368(5):436–45. DOI: 10.1056/NEJMoa1209978

10.

Eastham

Auffenberg

Barocas

, et al. Clinically localized prostate cancer: AUA/ASTRO guideline, part I: introduction, risk assessment, staging, and risk-based management. J Urol. 2022;208(1):10–8. DOI: 10.1097/JU.0000000000002757

11.

Daskivich

Litwin

Penson

DF.

Effect of age, tumor risk, and comorbidity in a U.S. Population-based cohort of men with prostate cancer. Ann Intern Med. 2013;159(5):370. DOI: 10.7326/0003-4819-159-5-201309030-00018

12.

Albertsen

Moore

Shih

Lin

Lu-Yao

GL.

Impact of comorbidity on survival among men with localized prostate cancer. J Clin Oncol. 2011;29(10):1335–41. DOI: 10.1200/JCO.2010.31.2330

13.

Hamdy

Donovan

Lane

, et al. 10-year outcomes after monitoring, surgery, or radiotherapy for localized prostate cancer. N Engl J Med. 2016;375(15):1415–24. doi: 10.1056/NEJMoa1606220

14.

Bill-Axelson

Holmberg

Garmo

, et al. Radical prostatectomy or watchful waiting in prostate cancer - 29-year follow-up. N Engl J Med. 2018;379(24):2319–29. DOI: 10.1056/NEJMoa1807801

15.

Wilt

Langsetmo

, et al. Radical prostatectomy or observation for clinically localized prostate cancer: extended follow-up of the Prostate Cancer Intervention Versus Observation Trial (PIVOT). Eur Urol. 2020;77(6):713–24. DOI: 10.1016/j.eururo.2020.02.009

16.

Chamie

Williams

JC.

Population-based assessment of determining treatments for prostate cancer. JAMA Oncol. 2015;1(1):60–7. DOI: 10.1001/jamaoncol.2014.192

17.

Hoffman

Niu

Shen

, et al. Physician variation in management of low-risk prostate cancer: a population-based cohort study. JAMA Intern Med. 2014;174(9):1450–9. DOI: 10.1001/jamainternmed.2014.3021

18.

Daskivich

Gale

Luu

, et al. Variation in communication of competing risks of mortality in prostate cancer treatment consultations. J Urol. 2022;208(2):301–8. DOI: 10.1097/JU.0000000000002675

19.

Naser-Tavakolian

Gale

Luu

, et al. Variation in communication of side effects in prostate cancer treatment consultations. J Urol. 2022;207(suppl 5):e527. DOI: 10.1097/JU.0000000000002580.14

20.

Andreadis

Chan

Park

, et al. Imprecision and preferences in interpretation of verbal probabilities in health: a systematic review. J Gen Intern Med. 2021;36(12):3820–9. DOI: 10.1007/s11606-021-07050-7

21.

van de Water

van Kleef

Dijksterhuis

WPM

, et al. Communicating treatment risks and benefits to cancer patients: a systematic review of communication methods. Qual Life Res. 2020;29(7):1747–66. DOI: 10.1007/s11136-020-02503-8

22.

Mottet

Bellmunt

Bolla

, et al. EAU-ESTRO-SIOG guidelines on prostate cancer. Part 1: screening, diagnosis, and local treatment with curative intent. Eur Urol. 2017;71(4):618–29. DOI: 10.1016/j.eururo.2016.08.003

23.

Schaeffer

Srinivas

Adra

, et al. NCCN guidelines^® insights: prostate cancer, Version 1.2023. J Natl Compr Canc Netw. 2022;20(12):1288–98. DOI: 10.6004/jnccn.2022.0063

24.

Sanda

Cadeddu

Kirkby

, et al. Clinically localized prostate cancer: AUA/ASTRO/SUO guideline. Part I: risk stratification, shared decision making, and care options. J Urol. 2018;199(3):683–90. DOI: 10.1016/j.juro.2017.11.095

25.

Lane

Ellimoottil

Wallner

Meeks

Mbassa

Clemens

JQ.

Shared decision-making in urologic practice: results from the 2019 AUA census. Urology. 2020;145:66–72. DOI: 10.1016/j.urology.2020.06.078

26.

Maes-Carballo

Munoz-Nunez

Martin-Diaz

Mignini

Bueno-Cavanillas

Khan

KS.

Shared decision making in breast cancer treatment guidelines: development of a quality assessment tool and a systematic review. Health Expect. 2020;23(5):1045–64. DOI: 10.1111/hex.13112

27.

Lohia

Hanson

Tuttle

Morris

LGT

. Active surveillance for patients with very low-risk thyroid cancer. Laryngoscope Investig Otolaryngol. 2020;5(1):175–82. DOI: 10.1002/lio2.356

28.

Keating

Pace

LE.

Breast cancer screening in 2018: time for shared decision making. JAMA. 2018;319(17):1814–5. DOI: 10.1001/jama.2018.3388

29.

Forner

Noel

Shuman

, et al. Shared decision-making in head and neck surgery: a review. JAMA Otolaryngol Head Neck Surg. 2020;146(9):839–44. DOI: 10.1001/jamaoto.2020.1601

30.

Kunneman

Stiggelbout

Pieterse

AH.

Do clinicians convey what they intend? Lay interpretation of verbal risk labels used in decision encounters. Patient Educ Couns. 2020;103(2):418–22. DOI: 10.1016/j.pec.2019.08.035

31.

Mazur

Hickam

Mazur

MD.

How patients’ preferences for risk information influence treatment choice in a case of high risk and high therapeutic uncertainty: asymptomatic localized prostate cancer. Med Decis Making. 1999;19(4):394–8. DOI: 10.1177/0272989X9901900407

32.

Daskivich

Gale

Luu

, et al. Patient preferences for communication of life expectancy in prostate cancer treatment consultations. JAMA Surg. 2022;157(1):70–2. DOI: 10.1001/jamasurg.2021.5803

33.

Vromans

Tillier

Pauws

van der Poel

van de Poll-Franse

Krahmer

EJ.

Communication, perception, and use of personalized side-effect risks in prostate cancer treatment-decision making: an observational and interview study. Patient Educ Couns. 2022;105(8):2731–9. DOI: 10.1016/j.pec.2022.04.017

34.

Vromans

van Eenbergen

Geleijnse

Pauws

van de Poll-Franse

Krahmer

EJ.

Exploring cancer survivor needs and preferences for communicating personalized cancer statistics from registry data: qualitative multimethod study. JMIR Cancer. 2021;7(4):e25659. DOI: 10.2196/25659

35.

Galesic

Garcia-Retamero

Gigerenzer

Using icon arrays to communicate medical risks: overcoming low numeracy. Health Psychol. 2009;28(2):210–6. DOI: 10.1037/a0014474

36.

Gillespie

The experience of risk as ‘measured vulnerability’: health screening and lay uses of numerical risk. Sociol Health Illn. 2012;34(2):194–207. DOI: 10.1111/j.1467-9566.2011.01381.x

37.

Elwyn

O’Connor

Stacey

, et al. Developing a quality criteria framework for patient decision aids: online international Delphi consensus process. BMJ. 2006;333(7565):417. DOI: 10.1136/bmj.38926.629329.AE

38.

Stacey

Legare

Lewis

, et al. Decision aids for people facing health treatment or screening decisions. Cochrane Database Syst Rev. 2017;4(4):CD001431. DOI: 10.1002/14651858.CD001431.pub5

39.

Trevena

Bonner

Okan

, et al. Current challenges when using numbers in patient decision aids: advanced concepts. Med Decis Making. 2021;41(7):834–47. DOI: 10.1177/0272989X21996342

40.

Bonner

Trevena

Gaissmaier

, et al. Current best practice for presenting probabilities in patient decision aids: fundamental principles. Med Decis Making. 2021;41(7):821–33. DOI: 10.1177/0272989X21996328

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.05 MB

0.07 MB

0.01 MB