Conversational stories & self organizing maps: Innovations for the scalable study of uncertainty in healthcare communication.

BACKGROUND: Understanding uncertainty in participatory decision-making requires scientific attention to interaction between what actually happens when patients, families and clinicians engage one another in conversation and the multi-level contexts in which these occur. Achieving this understanding will require conceptually grounded and scalable methods for use in large samples of people representing diversity in cultures, speaking and decision-making norms, and clinical situations. DISCUSSION: Here, we focus on serious illness and describe Conversational Stories as a scalable and conceptually grounded framework for characterizing uncertainty expression in these clinical contexts. Using actual conversations from a large direct-observation cohort study, we demonstrate how natural language processing and unsupervised machine learning methods can reveal underlying types of uncertainty stories in serious illness conversations. CONCLUSIONS: Conversational Storytelling offers a meaningful analytic framework for scalable computational methods to study uncertainty in healthcare conversations.

Palliative Care Clinician Overestimation of Survival in Advanced Cancer: Disparities and Association With End-of-Life Care.

CONTEXT: Clinicians frequently overestimate survival time in serious illness. OBJECTIVE: The objective of this study was to understand the frequency of overestimation in palliative care (PC) and the relation with end-of-life (EOL) treatment. METHODS: This is a multisite cohort study of 230 hospitalized patients with advanced cancer who consulted with PC between 2013 and 2016. We asked the consulting PC clinician to make their "best guess" about the patients' "most likely survival time, assuming that their illnesses are allowed to take their natural course" (<24 hours; 24 hours to less than two weeks; two weeks to less than three months; three months to less than six months; six months or longer). We followed patients for up to six month for mortality and EOL treatment utilization. Patients completed a brief interviewer-facilitated questionnaire at study enrollment. RESULTS: Median survival was 37 days (interquartile range: 12 days, 97 days) and 186/230 (81%) died during the follow-up period. Forty-one percent of clinicians' predictions were accurate. Among inaccurate prognoses, 85% were overestimates. Among those who died, overestimates were substantially associated with less hospice use (ORadj: 0.40; 95% CI: 0.16-0.99) and later hospice enrollment (within 72 hours of death ORadj: 0.33; 95% CI: 0.15-0.74). PC clinicians were substantially more likely to overestimate survival for patients who identified as Black or Latino compared to others (ORadj: 3.89; 95% CI: 1.64-9.22). EOL treatment preferences did not explain either of these findings. CONCLUSION: Overestimation is common in PC, associated with lower hospice use and a potentially mutable source of racial/ethnic disparity in EOL care.

Design of, and enrollment in, the palliative care communication research initiative: a direct-observation cohort study.

BACKGROUND: Understanding the characteristics of communication that foster patient-centered outcomes amid serious illness are essential for the science of palliative care. However, epidemiological cohort studies that directly observe clinical conversations can be challenging to conduct in the natural setting. We describe the successful enrollment, observation and data collection methods of the ongoing Palliative Care Communication Research Initiative (PCCRI). METHODS: The PCCRI is a multi-site cohort study of naturally occurring inpatient palliative care consultations. The 6-month cohort data includes directly observed and audio-recorded palliative care consultations (up to first 3 visits); patient/proxy/clinician self-report questionnaires both before and the day after consultation; post-consultation in-depth interviews; and medical/administrative records. RESULTS: One hundred fourteen patients or their proxies enrolled in PCCRI during Enrollment Year One (of Three). Seventy percent of eligible patients/proxies were invited to hear about a communication research study (188/269); 60% of them ultimately enrolled in the PCCRI (114/188), resulting in a 42% sampling proportion (114/269 eligible). All PC clinicians at study sites were invited to participate; all 45 participated. CONCLUSIONS: Epidemiologic study of patient-family-clinician communication in palliative care settings is feasible and acceptable to patients, proxies and clinicians. We detail the successful PCCRI methods for enrollment, direct observation and data collection for this complex "field" environment.

Using music[al] knowledge to represent expressions of emotions.

OBJECTIVE: Being able to identify expressions of emotion is crucial to effective clinical communication research. However, traditional linguistic coding systems often cannot represent emotions that are expressed nonlexically or phonologically (i.e., not through words themselves but through vocal pitch, speed/rhythm/tempo, and volume). METHODS: Using audio recording of a palliative care consultation in the natural hospital setting, two experienced music scholars employed Western musical notation, as well as the graphic realization of a digital audio program (Piano roll visualization), to visually represent the sonic features of conversation where a patient has an emotional "choke" moment. RESULTS: Western musical notation showed the ways that changes in pitch and rate correspond to the patient's emotion: rising sharply in intensity before slowly fading away. Piano roll visualization is a helpful supplement. CONCLUSIONS: Using musical notation to illustrate palliative care conversations in the hospital setting can render visible for analysis several aspects of emotional expression that researchers otherwise experience as intuitive or subjective. Various forms and formats of musical notation techniques and sonic visualization technologies should be considered as fruitful and complementary alternatives to traditional coding tools in clinical communications research. PRACTICE IMPLICATIONS: Musical notation offers opportunity for both researchers and learners to "see" how communication evolves in clinical encounters, particularly where the lexical and phonological features of interpersonal communication are concordant and discordant with one another.

Emotional distress and compassionate responses in palliative care decision-making consultations.

BACKGROUND: Seriously ill hospitalized patients and their loved ones are frequently faced with complex treatment decisions laden with expressions of emotional distress during palliative care (PC) consultations. Little is known about these emotional expressions or the compassionate responses providers make and how common these are in PC decision-making conversations. OBJECTIVES: To describe the types and frequency of emotional distress that patients and loved ones express and how providers respond to these emotions during PC decision-making consultations with seriously ill hospitalized patients. METHODS: We used a quantitative descriptive approach to analyze 71 audio-recorded inpatient PC decision-making consultations for emotional distress and clinicians' responses to those emotions using reliable and established methods. RESULTS: A total of 69% of conversations contained at least one expression of emotional distress. The per-conversation frequency of expressions of emotional distress ranged from 1 to 10. Anxiety/fear were the most frequently encountered emotions (48.4%) followed by sadness (35.5%) and anger/frustration (16.1%). More than half of the emotions related to the patient's feelings (53.6%) and 41.9% were related to the loved ones' own emotions. The majority of emotions were moderate in intensity (65.8%) followed by strong (20.7%) and mild (13.5%). Clinicians responded to a majority of emotions with a compassionate response (75.7%) followed by those with medical content (21.9%) and very few were ignored (1.3%). CONCLUSIONS: Expressions of emotional distress are common during PC consultations and are usually met with compassionate responses by the clinician.

Direct observation of prognosis communication in palliative care: a descriptive study.

CONTEXT: Palliative care (PC) consultations result in improved patient understanding of prognosis and better quality of life, yet the content and processes of prognosis communication during PC consultations remain unknown. OBJECTIVES: To describe prognosis communication during PC consultation with seriously ill hospitalized patients. METHODS: We audio recorded 71 sequential inpatient PC consultations (initial visit) with seriously ill patients and their families who were referred for "goals of care" clarification or help with "end-of-life decision making." Conversations were coded using reliable methods and we then linked conversation codes to clinical record and clinician interview data. RESULTS: Ninety-three percent of consultations contained prognosis communication. Participants communicated prognoses regarding quality of life more frequently than survival; focused prognosis estimates on the unique patient more frequently than on a general population; and framed prognosis using pessimistic cues more frequently than optimistic ones. Prognoses were more commonly spoken by PC clinicians than by patients/families. The following two factors demonstrated an association with the rate of prognostic communication and with the pessimistic framing of that information: whether the patient, family, or both participated in the conversation, and shorter expected survival (as estimated by the attending physician). CONCLUSION: Prognoses are routinely communicated in PC consultations with hospitalized patients and their families. The rate and characteristics of prognosis communication differ based on the length of time the patient is expected to live.

Self-rated cardiovascular risk and 15-year cardiovascular mortality.

PURPOSE: Many individuals perceive their cardiovascular disease (CVD) risk to be lower than established clinical tools would estimate, yet little is known about the long-term consequences of holding such optimistic beliefs. We evaluated whether lower self-ratings of CVD risk are associated with lower rates of CVD death after addressing potential confounding by an extensive set of social and biologic CVD risk factors. METHODS: We conducted a 15-year mortality surveillance study of adults aged 35 to 75 years from southeastern New England (n = 2,816) who had no history of myocardial infarction. Baseline evaluation in 1990-1992 included household interview, anthropomorphic measures, and laboratory analyses. Outcomes were obtained using the National Death Index records through December 2005. RESULTS: Rating oneself to be at lower-than-average risk for one's age and sex was associated with lower rates of CVD mortality among men (hazard ratio [HR]=0.3; 95% confidence interval [CI], 0.2-0.7) but not among women (HR = 0.9; 95% CI, 0.5-1.7). None of the following weakened the findings among men: adjustment for baseline Framingham Risk Score, propensity score adjustment for both social and biologic factors, and censoring the first 2 years of surveillance. CONCLUSIONS: Lower self-ratings of CVD risk are independently associated with lower rates of CVD death among men.