Shared decision-making (SDM) is a foundational principle in the management of prostate cancer at all stages, from screening to treatment. Clinical guidelines consistently emphasize that treatment choice should incorporate individualized risk assessment, life expectancy, functional baseline, and patient preferences [1]. However, despite broad endorsement, the quality and consistency of risk communication in prostate cancer consultations remain highly variable and difficult to assess due to factors related to both the clinician and the patient [2]. Although satisfaction surveys are easy to administer, they don’t tell us whether patients actually received the information needed to make an informed decision: did the consultation include the specific, patient-relevant information required to make an informed choice (prognosis, life expectancy, side-effects, etc.), and was that information delivered with sufficient precision to support SDM? Evidence demonstrates that critical information is frequently omitted or discussed without timelines, even in specialized settings [3]. This also raises the question of the different methods used to deliver information other than oral communication.

Natural language processing (NLP) analysis of consultation transcripts offers a practical way to identify these gaps and move beyond subjective evaluations. Zheng et al. introduce an NLP-based framework to quantify the quality of physician communication in prostate cancer consultations [4]. The major shift of this work is that the authors focus on analyzing the actual content of communication and not measuring patients’ subjective feelings. The authors trained a number of NLP models (random forest, XGboost, etc.) to detect guideline-aligned communication domains, including cancer severity, life expectancy and prognosis, baseline urinary and erectile function, and treatment-related side effects. Across nearly 29,000 annotated sentences, Random-Forest demonstrated relatively high results (AUC 0.84 to 0.99) with more confusion concerning the functional baseline. Importantly, authors reported that consultation-level communication quality could be assessed by grading only the top 10 sentences per concept, with an accuracy of 90% in external validation [4].

Why measuring communication matters

Several studies have shown considerable variation in how treatment-related risks and harms are discussed in prostate cancer consultations. Bhojani et al. demonstrated that SDM for PSA testing remains suboptimal in the United States, with most men, regardless of age, having never engaged in SDM conversations with a healthcare provider about PSA testing [5, 6]. In a qualitative analysis of multidisciplinary consultations, Daskivich et al. showed that major side effects were sometimes not mentioned at all and were frequently not quantified or contextualized [3]. During surgery discussions, erectile dysfunction and urinary incontinence were omitted in 15% and 12% of consultations, respectively. For radiotherapy, omission rates were even higher. This lack of information on side effects may influence treatment decisions and subsequently lead to decision regret [7]. Some studies suggest that clinicians often use persuasive language in prostate cancer consultations that aligns with their own treatment preferences, influencing how prognosis, life expectancy, and side effects are framed and potentially affecting patient choice [8]. Other researchers highlight that SDM practices vary across populations and contexts. Masterson et al. found that Black and Hispanic men with localized prostate cancer were significantly less likely than White men to prioritize longer life expectancy over quality-of-life considerations when making treatment decisions [9]. This highlights the differences between patients and their expectations regarding information. Thus, developing a well-designed measure of doctor-patient communication may help identify weaknesses in communication with specific populations or on specific topics, ultimately improving accessibility and ensuring that all patients receive a comparable level of care.

Limitations and ethical considerations

Doctor-patient communication is a specific and, in many ways, sacred aspect of every healthcare professional’s work. Therefore, any attempt to quantify it must be approached carefully. It is important to recognize that NLP systems evaluate informational structure rather than relational depth. Empathy, emotional responsiveness, and alignment with patient values are core components of doctor-patient communication and remain difficult to measure directly. Besides, communication quality should not be reduced to numerical disclosure alone. Effective SDM requires contextualization, explanation of trade-offs, and openness to patient concerns [2]. This information can be provided in several consultations and in different forms to facilitate acceptance and understanding, and not to mention the time needed for reflection [10].

Overemphasis on measurable elements risks creating a checklist-like behavior rather than personal engagement. These measures take into account a specific moment in the doctor-patient relationship and do not consider the history of their relationship. Sociological articles explain that information provided in list form may unintentionally prevent SDM if they do not accommodate the preferences expressed [11]. Moreover, this tool assesses the SDM between one patient and one clinician, even though we know that patients rarely receive information and make decisions alone.

Conclusion

From the perspective of the Young Academic Urologists (YAU) of the European Association of Urology (EAU), the work by Zheng et al. marks a shift toward measurable standards of communication quality in urologic oncology. For early-career urologists, objective communication metrics offer opportunities to improve training, standardize SDM practices, and enable benchmarking across institutions—while ensuring that technology supports, rather than replaces, patient-centered care. Beyond satisfaction scores lies a more ambitious goal, evidence-based communication that provides every patient with clear, quantified, and individualized information; if carefully implemented, NLP-based assessment may become an important quality innovation in prostate cancer care. More than doctor-patient communication, this could enable innovation with consultation training modules similar to surgical simulations.