Chatbot Performance in Defining and Differentiating Palliative Care, Supportive Care, Hospice Care.

CONTEXT: Artificial intelligence (AI) chatbot platforms are increasingly used by patients as sources of information. However, there is limited data on the performance of these platforms, especially regarding palliative care terms. OBJECTIVES: We evaluated the accuracy, comprehensiveness, reliability, and readability of three AI platforms in defining and differentiating "palliative care," "supportive care," and "hospice care." METHODS: We asked ChatGPT, Microsoft Bing Chat, Google Bard to define and differentiate "palliative care," "supportive care," and "hospice care" and provide three references. Outputs were randomized and assessed by six blinded palliative care physicians using 0-10 scales (10 = best) for accuracy, comprehensiveness, and reliability. Readability was assessed using Flesch Kincaid Grade Level and Flesch Reading Ease scores. RESULTS: The mean (SD) accuracy scores for ChatGPT, Bard, and Bing Chat were 9.1 (1.3), 8.7 (1.5), and 8.2 (1.7), respectively; for comprehensiveness, the scores for the three platforms were 8.7 (1.5), 8.1 (1.9), and 5.6 (2.0), respectively; for reliability, the scores were 6.3 (2.5), 3.2 (3.1), and 7.1 (2.4), respectively. Despite generally high accuracy, we identified some major errors (e.g., Bard stated that supportive care had "the goal of prolonging life or even achieving a cure"). We found several major omissions, particularly with Bing Chat (e.g., no mention of interdisciplinary teams in palliative care or hospice care). References were often unreliable. Readability scores did not meet recommended levels for patient educational materials. CONCLUSION: We identified important concerns regarding the accuracy, comprehensiveness, reliability, and readability of outputs from AI platforms. Further research is needed to improve their performance.

Quality-of-Life (QOL) during Screening for Phase 1 Trial Studies in Patients with Advanced Solid Tumors and Its Impact on Risk for Serious Adverse Events.

Background: Serious adverse events (SAEs) and subject replacements occur frequently in phase 1 oncology clinical trials. Whether baseline quality-of-life (QOL) or social support can predict risk for SAEs or subject replacement among these patients is not known. Methods: Between 2011-2013, 92 patients undergoing screening for enrollment into one of 22 phase 1 solid tumor clinical trials at Roswell Park Cancer Institute were included in this study. QOL Questionnaires (EORTC QLQ-C30 and FACT-G), Medical Outcomes Study Social Support Survey (MOSSSS), Charlson comorbidity scores (CCS) and Royal Marsden scores (RMS) were obtained at baseline. Frequency of dose limiting toxicities (DLTs), subject replacement and SAEs that occurred within the first 4 cycles of treatment were recorded. Fisher's exact test and Mann-Whitney-Wilcoxon test were used to study the association between categorical and continuous variables, respectively. A linear transformation was used to standardize QOL scores. p-value ≤ 0.05 was considered statistically significant. Results: Baseline QOL, MOSSSS, CCS and RMS were not associated with subject replacement nor DLTs. Baseline EORTC QLQ-C30 scores were significantly lower among patients who encountered SAEs within the first 4 cycles (p = 0.04). Conclusions: Lower (worse) EORTC QLQ-C30 score at baseline is associated with SAE occurrence during phase 1 oncology trials.

Longitudinal perceptions of prognosis and goals of therapy in patients with metastatic non-small-cell lung cancer: results of a randomized study of early palliative care.

PURPOSE: Understanding of prognosis among terminally ill patients impacts medical decision making. The aims of this study were to explore perceptions of prognosis and goals of therapy in patients with metastatic non-small-cell lung cancer (NSCLC) and to examine the effect of early palliative care on these views over time. PATIENTS AND METHODS: Patients with newly diagnosed metastatic NSCLC were randomly assigned to receive either early palliative care integrated with standard oncology care or standard oncology care alone. Participants completed baseline and longitudinal assessments of their perceptions of prognosis and the goals of cancer therapy over a 6-month period. RESULTS: We enrolled 151 participants on the study. Despite having terminal cancer, one third of patients (46 of 145 patients) reported that their cancer was curable at baseline, and a majority (86 of 124 patients) endorsed getting rid of all of the cancer as a goal of therapy. Baseline perceptions of prognosis (ie, curability) and goals of therapy did not differ significantly between study arms. A greater percentage of patients assigned to early palliative care retained or developed an accurate assessment of their prognosis over time (82.5% v 59.6%; P = .02) compared with those receiving standard care. Patients receiving early palliative care who reported an accurate perception of their prognosis were less likely to receive intravenous chemotherapy near the end of life (9.4% v 50%; P = .02). CONCLUSION: Many patients with newly diagnosed metastatic NSCLC hold inaccurate perceptions of their prognoses. Early palliative care significantly improves patient understanding of prognosis over time, which may impact decision making about care near the end of life.

Early palliative care for patients with metastatic non-small-cell lung cancer.

BACKGROUND: Patients with metastatic non-small-cell lung cancer have a substantial symptom burden and may receive aggressive care at the end of life. We examined the effect of introducing palliative care early after diagnosis on patient-reported outcomes and end-of-life care among ambulatory patients with newly diagnosed disease. METHODS: We randomly assigned patients with newly diagnosed metastatic non-small-cell lung cancer to receive either early palliative care integrated with standard oncologic care or standard oncologic care alone. Quality of life and mood were assessed at baseline and at 12 weeks with the use of the Functional Assessment of Cancer Therapy-Lung (FACT-L) scale and the Hospital Anxiety and Depression Scale, respectively. The primary outcome was the change in the quality of life at 12 weeks. Data on end-of-life care were collected from electronic medical records. RESULTS: Of the 151 patients who underwent randomization, 27 died by 12 weeks and 107 (86% of the remaining patients) completed assessments. Patients assigned to early palliative care had a better quality of life than did patients assigned to standard care (mean score on the FACT-L scale [in which scores range from 0 to 136, with higher scores indicating better quality of life], 98.0 vs. 91.5; P=0.03). In addition, fewer patients in the palliative care group than in the standard care group had depressive symptoms (16% vs. 38%, P=0.01). Despite the fact that fewer patients in the early palliative care group than in the standard care group received aggressive end-of-life care (33% vs. 54%, P=0.05), median survival was longer among patients receiving early palliative care (11.6 months vs. 8.9 months, P=0.02). CONCLUSIONS: Among patients with metastatic non-small-cell lung cancer, early palliative care led to significant improvements in both quality of life and mood. As compared with patients receiving standard care, patients receiving early palliative care had less aggressive care at the end of life but longer survival. (Funded by an American Society of Clinical Oncology Career Development Award and philanthropic gifts; ClinicalTrials.gov number, NCT01038271.)

Code status documentation in the outpatient electronic medical records of patients with metastatic cancer.

BACKGROUND: Advanced care planning (ACP) is considered an essential component of medical care in the United States, especially in patients with incurable diseases. However, little is known about clinical practices in outpatient oncology settings related to discussing end-of-life care and documenting code status preferences in ambulatory medical records. OBJECTIVE: To assess the rate of documentation of code status in the electronic longitudinal medical records (LMR) of patients with metastatic cancer. DESIGN: Retrospective review of 2,498 patients with metastatic solid tumors at an academic cancer center. An electronic patient database and the LMR were queried to identify demographic information, cancer type, number of clinic visits, and documentation of code status. PARTICIPANTS: The sample consisted of adult patients with metastatic prostate, breast, ovarian, bladder kidney, colorectal, non-colorectal gastrointestinal (GI), and lung cancers. MEASUREMENTS: Primary outcome was the percentage of documented code status in the LMR. MAIN RESULTS: Among the 2,498 patients, 20.3% had a documented code status. Code status was designated most frequently in patients with non-colorectal GI (193/609, 31.7%) and lung (179/583, 30.7%) cancers and least frequently in patients with genitourinary malignancies [bladder/kidney (4/89, 4.5%), ovarian (4/93, 4.3%), and prostate (7/365, 1.9%) cancers]. Independent predictors of having documented code status included religious affiliation, cancer type, and a greater number of visits to the cancer center. Younger patients and black patients were less likely to be designated as DNR/DNI. CONCLUSIONS: Despite the incurable nature of metastatic cancer, only a minority of patients had a code status documented in the electronic medical record.

Clinical features and outcome of patients with non-small-cell lung cancer who harbor EML4-ALK.

PURPOSE: The EML4-ALK fusion oncogene represents a novel molecular target in a small subset of non-small-cell lung cancers (NSCLC). To aid in identification and treatment of these patients, we examined the clinical characteristics and treatment outcomes of patients who had NSCLC with and without EML4-ALK. PATIENTS AND METHODS: Patients with NSCLC were selected for genetic screening on the basis of two or more of the following characteristics: female sex, Asian ethnicity, never/light smoking history, and adenocarcinoma histology. EML4-ALK was identified by using fluorescent in situ hybridization for ALK rearrangements and was confirmed by immunohistochemistry for ALK expression. EGFR and KRAS mutations were determined by DNA sequencing. RESULTS: Of 141 tumors screened, 19 (13%) were EML4-ALK mutant, 31 (22%) were EGFR mutant, and 91 (65%) were wild type (WT/WT) for both ALK and EGFR. Compared with the EGFR mutant and WT/WT cohorts, patients with EML4-ALK mutant tumors were significantly younger (P < .001 and P = .005) and were more likely to be men (P = .036 and P = .039). Patients with EML4-ALK-positive tumors, like patients who harbored EGFR mutations, also were more likely to be never/light smokers compared with patients in the WT/WT cohort (P < .001). Eighteen of the 19 EML4-ALK tumors were adenocarcinomas, predominantly the signet ring cell subtype. Among patients with metastatic disease, EML4-ALK positivity was associated with resistance to EGFR tyrosine kinase inhibitors (TKIs). Patients in the EML4-ALK cohort and the WT/WT cohort showed similar response rates to platinum-based combination chemotherapy and no difference in overall survival. CONCLUSION: EML4-ALK defines a molecular subset of NSCLC with distinct clinical characteristics. Patients who harbor this mutation do not benefit from EGFR TKIs and should be directed to trials of ALK-targeted agents.