Predictors of Family Dissatisfaction with Support During Neurocritical Care Shared Decision-Making.

BACKGROUND: There is a critical need to improve support for families making difficult shared decisions about patient care with clinicians in the neuroscience ICU (neuro-ICU). The aim of this study is to identify patient- and family-related factors associated with dissatisfaction with shared decision-making support among families of neuro-critically ill patients. METHODS: We conducted a retrospective observational cohort study using survey data that had been collected from a consecutive sample of family members of patients in the neuro-ICU (one family member per patient) at two US academic centers. Satisfaction with shared decision-making support on ICU discharge had been measured among family members using one specific Likert scale item on the Family Satisfaction in the ICU 24 survey, a validated survey instrument for families of patients in the ICU. We dichotomized top-box responses for this particular item as an outcome variable and identified available patient- and family-related covariates associated with dissatisfaction (i.e., less than complete satisfaction) via univariate and multivariate analyses. RESULTS: Among 355 surveys, 180 (49.5%) of the surveys indicated dissatisfaction with support during decision-making. In a multivariate model, no preexisting characteristics of families or patients ascertainable on ICU admission were predictive of dissatisfaction. However, among family factors determined during the ICU course, experiencing three or fewer formal family meetings (odds ratio 1.93 [confidence interval 1.13-3.31]; p = 0.01) was significantly predictive of dissatisfaction with decisional support in this cohort with an average patient length of stay of 8.6 days (SD 8.4). There was also a trend toward a family's decision to keep a patient as full code, without treatment limitations, being predictive of dissatisfaction (odds ratio 1.80 [confidence interval 0.93-3.51]; p = 0.08). CONCLUSIONS: Family dissatisfaction with neuro-ICU shared decision-making support is not necessarily predicted by any preexisting family or patient variables but appears to correlate with participating in fewer formal family meetings during ICU admission. Future studies to improve family satisfaction with neurocritical care decision-making support should have broad inclusion criteria for participants and should consider promoting frequency of family meetings as a core strategy.

Dedicated Afternoon Rounds for ICU Patients' Families and Family Satisfaction With Care.

OBJECTIVE: It was hypothesized that adding dedicated afternoon rounds for patients' families to supplement standard family support would improve overall family satisfaction with care in a neuroscience ICU. DESIGN: Pre- and postimplementation (pre-I and post-I) design. SETTING: Single academic neuroscience ICU. PATIENTS: Patients in the neuroscience ICU admitted for longer than 72 hours or made comfort measures only at any point during neuroscience ICU admission. INTERVENTION: The on-service attending intensivist and a neuroscience ICU nursing leader made bedside visits to families to address concerns during regularly scheduled, advertised times two afternoons each week. MEASUREMENTS AND MAIN RESULTS: One family member per patient during the pre-I and post-I periods was recruited to complete the Family Satisfaction in the ICU 24 instrument. Post-I respondents indicated whether they had participated in the afternoon rounds. For primary outcome, the mean pre-I and post-I composite Family Satisfaction in the ICU 24 scores (on a 100-point scale) were compared. A total of 146 pre-I (March 2013 to October 2014; capture rate, 51.6%) and 141 post-I surveys (October 2014 to December 2015; 47.2%) were collected. There was no difference in mean Family Satisfaction in the ICU 24 score between groups (pre-I, 89.2 ± 11.2; post-I, 87.4 ± 14.2; p = 0.6). In a secondary analysis, there was also no difference in mean Family Satisfaction in the ICU 24 score between the pre-I respondents and the 39.0% of post-I respondents who participated in family rounds. The mean Family Satisfaction in the ICU 24 score of the post-I respondents who reported no participation trended lower than the mean pre-I score, with fewer respondents in this group reporting complete satisfaction with emotional support (75% vs. 54%; p = 0.002), coordination of care (82% vs. 68%; p = 0.03), and frequency of communication by physicians (60% vs. 43%; p = 0.03). CONCLUSIONS: Dedicated afternoon rounds for families twice a week may not necessarily improve an ICU's overall family satisfaction. Increased dissatisfaction among families who do not or cannot participate is possible.

Deep Learning to Estimate RECIST in Patients with NSCLC Treated with PD-1 Blockade.

Real-world evidence (RWE), conclusions derived from analysis of patients not treated in clinical trials, is increasingly recognized as an opportunity for discovery, to reduce disparities, and to contribute to regulatory approval. Maximal value of RWE may be facilitated through machine-learning techniques to integrate and interrogate large and otherwise underutilized datasets. In cancer research, an ongoing challenge for RWE is the lack of reliable, reproducible, scalable assessment of treatment-specific outcomes. We hypothesized a deep-learning model could be trained to use radiology text reports to estimate gold-standard RECIST-defined outcomes. Using text reports from patients with non-small cell lung cancer treated with PD-1 blockade in a training cohort and two test cohorts, we developed a deep-learning model to accurately estimate best overall response and progression-free survival. Our model may be a tool to determine outcomes at scale, enabling analyses of large clinical databases. SIGNIFICANCE: We developed and validated a deep-learning model trained on radiology text reports to estimate gold-standard objective response categories used in clinical trial assessments. This tool may facilitate analysis of large real-world oncology datasets using objective outcome metrics determined more reliably and at greater scale than currently possible.This article is highlighted in the In This Issue feature, p. 1.

Primary Care Physician Involvement in Shared Decision Making for Critically Ill Patients and Family Satisfaction with Care.

PURPOSE: An intensive care unit (ICU) patient's primary care physician (PCP) may be able to assist family with certain ICU shared medical decisions. We explored whether families of patients in nonopen ICUs who nevertheless report involvement of a patient's PCP in medical decision making are more satisfied with ICU shared decision making than families who do not. METHODS: Between March 2013 and December 2015, we administered the Family Satisfaction in the ICU 24 survey to family members of adult neuroscience ICU patients. We compared the mean score for the survey subsection regarding shared decision making (graded on a 100-point scale), as well as individual survey items, between those who reported the patient's PCP involvement in any medical decision making versus those who did not. RESULTS: Among 263 respondents, there was no difference in mean overall decision-making satisfaction scores for those who reported involvement (81.1; SD = 15.2) versus those who did not (80.1; SD = 12.8; P = .16). However, a higher proportion reporting involvement felt completely satisfied with their 1) inclusion in the ICU decision making process (75.9% vs 61.4%; P = .055), and 2) control over the care of the patient (73.6% vs 55.6%; P = .02), with no difference regarding consistency of clinical information provided by the medical team (64.8% vs 63.5%; P = 1.00). CONCLUSIONS: Families who report involvement of a patient's PCP in medical decision making for critically ill patients may be more satisfied than those who do not with regard to specific aspects of ICU decision making. Further research would help understand how best to engage PCPs in shared decisions.