Moving Toward a Basic Science of Prognostic Communication

In this issue of Pediatrics, Kaye et al¹  explore longitudinal patterns of prognostic disclosure in pediatric oncology. Disclosing a prognosis to parents and children was once considered antithetical to good clinical practice.²  Today, a large body of evidence supports transparent disclosure in pediatric oncology, especially with parents and adolescent and young adult patients. Parents and adolescent and young adult patients largely desire transparency, and truthful disclosure of a prognosis has been associated with increased hope.^3–5  Lack of prognostic disclosure has been linked to increased regret for bereaved family members and delayed palliative care.⁶  Yet some parents and patients report unmet information needs and provide overly optimistic estimates of a prognosis.^7–11 

Despite this body of evidence, our understanding of prognostic disclosure practices is limited mostly to (1) whether clinicians have disclosed a prognosis and (2) how accurately patients and parents have estimated the child’s prognosis. In 2017, a group of communication experts proposed a research agenda for communication with patients who have serious illness. One focus of their agenda was a new emphasis on exploring the mechanisms of communication. They argued that “to develop more targeted interventions requires a broader understanding of the ‘basic science’ of communication.”¹²  In this issue of Pediatrics, Kaye et al¹  make an important contribution to this basic science of prognostic communication using longitudinal recordings of disease progression discussions to characterize patterns of disclosure.

Kaye et al¹  audio-recorded serial discussions at disease reevaluation time points and administered surveys across a child’s advancing illness course. They analyzed 141 recordings and 40 hours of conversations, finding that <4% of recorded dialogue represented prognostic disclosure. The majority of prognostic communication occurred during discussions at the time of frank disease progression, with statements of curability often being unclear. In addition to this finding, they also identified 3 distinct patterns of prognostic communication: (1) absent, (2) deferred, and (3) seed-planting. “Absent” indicated a complete lack of direct discussions about a poor prognosis or incurability. “Deferred” indicated a delay in discussions about a poor prognosis or incurability until the last few discussions before death. “Seed-planting” indicated that clinicians laid the groundwork for future discussions with conversations of a poor prognosis early in the child’s illness course.

These findings provide an evidence-informed framework of potential approaches to prognostic disclosure, adding a new layer of depth to our knowledge about prognostic disclosure in pediatric oncology. The authors rightly note that larger studies are needed to validate this framework and assess for additional patterns of disclosure. Future studies might also determine if these approaches correlate with other outcomes of communication, such as prognostic understanding, decisional regret, and quality of life. If these patterns of prognostic disclosure are confirmed, future researchers might also leverage natural language processing in large data sets to determine relationships between disclosure patterns and outcomes.

As the basic science of prognostic disclosure continues to advance, appreciation for the parent’s and child’s unique communication preferences and needs will be essential. There is likely no single right way to disclose a prognosis. Instead, we need to advance toward precision communication, in which the approach to communication meets the unique needs and desires of each particular family at that particular time. This type of precision communication requires a deeper understanding of parent and child health literacy and numeracy, the impact of culture on communication, and the role of previous illness experiences, among other considerations. Furthermore, clinicians need validated tools and approaches to ascertaining these unique needs of patients and parents. This study by Kaye et al¹  opens an exciting new trajectory for research that will enrich our understanding of prognostic disclosure in pediatrics, yet much work remains to be done.