Detecting goals of care conversations in clinical notes with active learning

OBJECTIVE

Goals of care (GOC) discussions are an increasingly used quality metric in serious illness care and research. Wide variation in documentation practices within the Electronic Health Record (EHR) presents challenges for reliable measurement of GOC discussions. Novel natural language processing approaches are needed to capture GOC discussions documented in real-world samples of seriously ill hospitalized patients' EHR notes, a corpus with a very low event prevalence.

METHODS

To automatically detect sentences documenting GOC discussions outside of dedicated GOC note types, we proposed an ensemble of classifiers aggregating the predictions of rule-based, feature-based, and three transformers-based classifiers. We trained our classifier on 600 manually annotated EHR notes among patients with serious illnesses. Our corpus exhibited an extremely imbalanced ratio between sentences discussing GOC and sentences that do not. This ratio challenges standard supervision methods to train a classifier. Therefore, we trained our classifier with active learning.

RESULTS

Using active learning, we reduced the annotation cost to fine-tune our ensemble by 70% while improving its performance in our test set of 176 EHR notes, with 0.557 F1-score for sentence classification and 0.629 for note classification.

CONCLUSION

When classifying notes, with a true positive rate of 72% (13/18) and false positive rate of 8% (13/158), our performance may be sufficient for deploying our classifier in the EHR to facilitate bedside clinicians' access to GOC conversations documented outside of dedicated notes types, without overburdening clinicians with false positives. Improvements are needed before using it to enrich trial populations or as an outcome measure.

Clinician-Patient Relationships in Virtual Care: A Dimensional Analysis of the Symbolic World of Cancer Care

Rapid uptake of telehealth technologies has shifted clinician-patient relationships, a well-studied topic of sociological inquiry. The purpose of this dimensional analysis study was to understand the symbolic interaction of clinicians and patients within virtual supportive cancer care. Seventeen clinicians, eighteen patients, and three care partners receiving or providing care at a multi-site cancer center in the United States participated in interviews. Our analysis of supportive cancer care experiences reveals a key tension: clinicians need to rely on patients in order to share clinical tasks in a virtual care setting but can be reluctant to do so. We posit that dimming the light on cancer is a process that enables clinicians to overcome their reluctance to engage in clinician-patient task sharing by strengthening the relationship. Taken together, these findings reconceptualize the symbolic interaction of the clinician-patient relationship and highlight opportunities to actualize models of relationship-centered virtual care. We discuss implications for clinical practice, ethical relational care, and the literature on clinician-patient relationships and trust.

The impact and implications of virtual supportive cancer care during the COVID-19 pandemic: integrating patient and clinician perspectives

Purpose

Many cancer centers made rapid shifts in supportive care delivery modalities at the onset of the COVID-19 pandemic. Improving virtual supportive cancer care requires deeply understanding both patient’s and clinician’s experiences. We aimed to integrate the perspectives of clinicians and patients to describe the transition to virtual supportive cancer care during COVID-19.

Methods

In clinical-academic partnership between a multi-site cancer care center in the Northeastern USA and a school of nursing, we conducted a study using dimensional analysis method. Theoretical sampling drove recruitment of patients and clinicians who engaged in virtual supportive cancer care from March 15, 2020 to December 15, 2020. In this sub-analysis, we coded the dimensional analysis data from semi-structured interviews using a descriptive approach with inductive conventional content analysis.

Results

We interviewed 17 clinicians, 18 patients, and 3 care partners about their experiences. We integrate patient and clinician perspectives in four in vivo categories: “When COVID hit,” “Not an IT expert,” “Those little moments,” and “The mothership.”

Conclusion

The findings uncover shared patient and clinician fears of missing or sub-optimal care at the onset of COVID-19, technological and relational challenges to engaging in care, and the mixed impacts of virtual care on access, convenience, and efficiency. This analysis suggests concrete action items to improve virtual supportive care for patients and clinicians. The findings corroborate the importance of convenience, access, and efficiency as care quality indicators and suggest potential to emphasize the clinician-patient relationship as an additional indicator of care quality.

Palliative care delivery changes during COVID-19 and enduring implications in oncology nursing: a rapid review

PURPOSE OF REVIEW

A rapid review was conducted to synthesize evidence of palliative care delivery changes during the COVID-19 pandemic. Changes were synthesized according to the eight domains of high-quality palliative care and enduring implications for oncology nurses beyond the pandemic discussed.

RECENT FINDINGS

The most significant changes occurred in the structure and processes of palliative care (Domain 1), where increased utilization of telehealth was critical in circumventing barriers imposed by COVID-19 mitigation. The suboptimal availability of community-based psychosocial supports for patients and caregivers and inadequate health system-based psychosocial supports for healthcare providers were highlighted (Domains 3-5). The pandemic also ushered in an increased emphasis on the need for advance care planning (ACP), where integrating its delivery earlier in the outpatient setting and shifting policy to promote subsequent virtual documentation (Domain 8) were essential to ensure care preferences were clarified and accessible before health crises occurred.

SUMMARY

Continuing to embrace and sustain systems-level changes with respect to telehealth, psychosocial supports, and ACP are critical to bridging gaps in palliative care delivery underscored by the pandemic. Oncology nurses are well positioned to fill these gaps in care beyond the pandemic by providing evidence-based, palliative care throughout the cancer continuum.

Rationalizing Emotions: A Secondary Analysis of the Supportive and Palliative Cancer Care Clinicians’ Experiences During COVID-19 (GP766)

Outcomes

1. Recall two emotions that oncology clinicians experienced during COVID-19

2. Recognize one pattern of mixed emotions experienced by clinicians during COVID-19

Background

The COVID-19 pandemic has called for a radical change in the delivery of oncology and palliative care services. In the first phase of the pandemic, oncology care was disrupted, with delayed screening and postponement of life-saving treatments including chemotherapy, radiation, and surgeries. Psycho-oncology care shifted due to social isolation, visitor restrictions, fear of infection, and transitions to telehealth. Emerging research is examining these effects on the patient and family unit; however, there is limited research exploring the emotional impact on cancer and palliative care clinicians. We know that healthcare professionals worldwide are experiencing immense strain, depression, anxiety, PTSD, and burnout, and many are considering leaving their profession. Understanding the emotional experience of clinicians in the early phase of the pandemic is key to addressing these issues.

Methods

Through a secondary analysis of qualitative data collected in the Virtual Supportive Cancer Care Research study, clinician interview transcripts were analyzed using an interpretive descriptive approach. Clinician participants represented various supportive and palliative roles in oncology care, including the disciplinary perspectives of nurse practitioners, social workers, dieticians, nurses, and physicians.

Findings

Clinicians experienced a plethora of emotions. These emotions were experienced on four axes: overt or latent; independent, sequential, or simultaneous; mixed emotions, which were experienced in three main patterns (contradictory, interwoven, and rationalizing); and internalized or externalized. All emotions could exist on multiple axes and ultimately guided clinicians’ reactions, responses, and reflections on their emotions as well as their behaviors.

Conclusion

Undoubtedly, clinicians have been significantly impacted both physically and emotionally during the pandemic. We are now faced with a unique opportunity to support clinicians holistically, as we do with patients. This study is a first step in unpacking what is behind clinician emotions and processing patterns in extreme care contexts such as the COVID-19 pandemic.

Clinician Distress: A Concept Clarification

In 2017, The National Academy of Medicine created the Action Collaborative on Clinician Well-Being and Resilience given the staggering statistics around distress and burnout, which is particularly high while caring for seriously ill older adults. This distress may be physical, emotional, spiritual, moral, ethical, existential or in degrees and combinations of multiple forms. Many crossover concepts are present in the literature, including moral distress, role stress, compassion fatigue, empathy, emotional labor, post-traumatic stress disorder, vicarious traumatization, and second victim. Because of the number of crossover concepts, we employed the Norris concept clarification method examining similar concepts to create an operational definition, construct a model, and develop possible research hypotheses of clinician distress. Articles were identified through PubMed, CINAHL, PsyhINFO, and SCOPUS with the expert assistance of a nurse librarian. Similar concepts in the literature were compared and contrasted with an emphasis on what encompasses clinician distress. Based on synthesis of the literature and existing concepts, clinician distress is defined as experiencing one or more negative emotions before, during, or after a clinical encounter that impacts the quality of care the clinician is able to provide to the patient. The result of this analysis provides some clarity around clinician distress and its surrogate terms, which presents opportunities for further investigation.

Palliative Care Consult for Clinician Distress Through the Philosophical Lenses of Gender Norms and Phenomenology

The National Academy of Medicine has raised significant concerns on clinician health and well-being as many experiencing burnout, post-traumatic stress, and depression. Indeed, clinicians experience a range of human emotions when caring for older adults with severe, life-limiting illnesses. These emotions may manifest in multiple ways and from various sources. Uncertain of how to attend to such distress, clinicians may consult a trusted resource, including the palliative care team. Palliative care specialists are trained to support the complexities and needs of patients and families; increasingly, however, palliative care consults are rooted in clinician distress. This session uses clinical case examples to explore the palliative care consult for distressed clinicians from two different philosophical perspectives: (1) phenomenology and (2) the social construct of gender norms. A phenomenological lens respects the unique, subjective lived experience of each individual in their day-to-day interactions with patients, families, and health care systems. Therefore, when caring for seriously ill older adults, clinicians may bring their own subjective experiences to the patient encounter and react differently to ethical dilemmas and conflicts that arise. The social construct of gender norms asks us to examine clinician distress from a different perspective. Here, the postmodern rejection of gender binarism allows clinicians to experience a spectrum of emotions and distress regardless of gender. Exploration through clinical cases will highlight the unique, varied experience of clinician distress and offer opportunities for future research into the role of palliative care teams in supporting distressed clinicians who care for seriously ill older adults.

Omega-3 fatty acid intake results in a relationship between the fatty acid composition of LDL cholesterol ester and LDL cholesterol content in humans

The relationship between the fatty acid composition of the low density lipoprotein (LDL) cholesterol ester and LDL cholesterol content was assessed in 26 free-living, normal subjects. Dietary intakes of 14:0, 16:0, 18:0, 18:1, 18:2omega6, 18:3omega3, 20:4omega6, 20:5omega3, 22:6omega3 were calculated from seven-day food records kept by each subject at baseline and after three months of supplementation with olive, flaxseed or fish oil, respectively. A randomized cross-over design was used. The fatty acid content of specific foods was calculated. Fasting blood samples, taken at the beginning and end of each supplementation period, were analyzed for the fatty acid content present in individual lipoproteins. There was a significant correlation between 20:5omega3 and 22:6omega3 intake and the content of these fatty acids in the LDL cholesterol ester fraction. During the fish oil treatment period the 16:0 and 18:0 content of the LDL cholesterol ester was highly predictive of LDL cholesterol content. This relationship was not observed during the baseline or placebo (olive oil) supplement period.