Improving Machine Learning 30-Day Mortality Prediction by Discounting Surprising Deaths.

BACKGROUND: Machine learning (ML) is an emerging tool for predicting need of end-of-life discussion and palliative care, by using mortality as a proxy. But deaths, unforeseen by emergency physicians at time of the emergency department (ED) visit, might have a weaker association with the ED visit. OBJECTIVES: To develop an ML algorithm that predicts unsurprising deaths within 30 days after ED discharge. METHODS: In this retrospective registry study, we included all ED attendances within the Swedish region of Halland in 2015 and 2016. All registered deaths within 30 days after ED discharge were classified as either "surprising" or "unsurprising" by an adjudicating committee with three senior specialists in emergency medicine. ML algorithms were developed for the death subclasses by using Logistic Regression (LR), Random Forest (RF), and Support Vector Machine (SVM). RESULTS: Of all 30-day deaths (n = 148), 76% (n = 113) were not surprising to the adjudicating committee. The most common diseases were advanced stage cancer, multidisease/frailty, and dementia. By using LR, RF, and SVM, mean area under the receiver operating characteristic curve (ROC-AUC) of unsurprising deaths in the test set were 0.950 (SD 0.008), 0.944 (SD 0.007), and 0.949 (SD 0.007), respectively. For all mortality, the ROC-AUCs for LR, RF, and SVM were 0.924 (SD 0.012), 0.922 (SD 0.009), and 0.931 (SD 0.008). The difference in prediction performance between all and unsurprising death was statistically significant (P < .001) for all three models. CONCLUSION: In patients discharged to home from the ED, three-quarters of all 30-day deaths did not surprise an adjudicating committee with emergency medicine specialists. When only unsurprising deaths were included, ML mortality prediction improved significantly.

Likelihood of admission to hospital from the emergency department is not universally associated with hospital bed occupancy at the time of admission.

BACKGROUND: The decision to admit into the hospital from the emergency department (ED) is considered to be important and challenging. The aim was to assess whether previously published results suggesting an association between hospital bed occupancy and likelihood of hospital admission from the ED can be reproduced in a different study population. METHODS: A retrospective cohort study of attendances at two Swedish EDs in 2015 was performed. Admission to hospital was assessed in relation to hospital bed occupancy together with other clinically relevant variables. Hospital bed occupancy was categorized and univariate and multivariate logistic regression were performed. RESULTS: In total 89,503 patient attendances were included in the final analysis. Of those, 29.1% resulted in admission within 24 h. The mean hospital bed occupancy by the hour of the two hospitals was 87.1% (SD 7.6). In both the univariate and multivariate analysis, odds ratio for admission within 24 h from the ED did not decrease significantly with an increasing hospital bed occupancy. CONCLUSIONS: A negative association between admission to hospital and occupancy level, as reported elsewhere, was not replicated. This suggests that the previously shown association might not be universal but may vary across sites due to setting specific circumstances.

Perspective of older African-American and Non-Hispanic white breast cancer survivors from diverse socioeconomic backgrounds toward physical activity: A qualitative study.

BACKGROUND: We sought to explore the perspective of older breast cancer survivors (BCS) from diverse racial and socioeconomic backgrounds toward physical activity (PA) to inform the design of a PA program that fosters acceptability. METHODS: Participants included sixty women, ≥65years, within two years of treatment completion for stage I-III breast cancer. We purposely sampled ≥ten patients in each race [African-American (AA) and Non-Hispanic White (NHW)] and socioeconomic status (SES) [SES disadvantaged and SES non-disadvantaged] group. Participants completed in-person interviews (n=60) and follow-up focus groups (n=45). Thematic analyses were employed. RESULTS: The median age was 71.0years (range: 65-87years). Five themes emerged: 1) importance of PA; 2) current PA participants engaged in; 3) influence of race and culture on PA attitudes and beliefs; 4) barriers to PA and facilitators to PA; and 5) PA preferences. Barriers included health issues (43%), particularly cancer treatment side effects such as fatigue. Facilitators included religious faith (38%) and family (50%). Preferences included group exercise (97%) and strength training (80%) due to concerns participants had with diminished upper body strength after cancer treatment. Although AA (59%) and SES non-disadvantaged (78%) participants reported that race and culture influenced their attitudes toward PA, it did not translate to racial and SES differences in preferences. CONCLUSION: Among older BCS, physical activity preferences were shaped by cancer experience, rather than by race and SES. Physical activity programs for older BCS should focus on addressing cancer treatment-related concerns and should include strength training to ensure PA programs are more acceptable to older BCS.

A Community-Based Exercise and Support Group Program Improves Quality of Life in African-American Breast Cancer Survivors: A Quantitative and Qualitative Analysis.

African-American (AA) breast cancer (BCa) survivors have higher mortality rates, more comorbidities and are less likely to meet national physical activity guidelines after diagnosis compared to Caucasian BCa survivors. We previously reported that a 20-week resistance exercise intervention coupled with a support group and home walking program, conducted using facilities and personnel at a community cancer support center, in Stage I-III AA BCa survivors improved strength, fitness and circulating C-peptide levels. Here, we report our findings on changes in quality of life (QoL) and other behavioral measures associated with this 20-week intervention and, discuss findings from a qualitative analysis of semi-structured patient interviews. We found a clinically relevant improvement in QoL using the Functional Assessment of Cancer Therapy for Breast Cancer (FACT-B) (Baseline, B: 101.1 ± 21.5; End-of-Intervention, EOI: 108.5 ± 21.6; p = 0.05) and, a significant decrease in depression using the Beck Depression Inventory-II (B: 11.9 ± 8.1; EOI: 9.0 ± 5.5; p = 0.03). Our analysis of the patient interviews support improvements in these behavioral measures in that participants stated that they "feel better", were "more motivated" and "uplifted" after the program. The patient interviews also provided insights to the primary motivators (e.g., social support, improvements in strength and function, weight loss) and barriers (e.g., family and health issues) in adhering to the program and provided suggestions for improving the program (e.g., incorporating nutritional and treatment related side-effect discussions). Our results suggest that community-based lifestyle interventions may improve QoL and depression in AA BCa survivors and lend insights for improving future programs.

A Community-Based Exercise and Support Group Program in African-American Breast Cancer Survivors (ABCs).

African-American (AA) women have higher rates of breast cancer (BCa) mortality than Caucasian women, and a recent study using data from the Surveillance, Epidemiology and End Results (SEER) registry suggests that this disparity may be due, in part, to the poorer health status of AAs at diagnosis and not treatment related issues. Randomized controlled trials involving supervised aerobic and resistance exercise have shown improved body composition and improvement in cancer-related biomarkers in BCa patients and may lead to improved recurrence and survival rates; however, most trials have focused on Caucasians and many have been conducted in academic- and clinic-based settings. We evaluated the feasibility of conducting a 20-week, supervised, resistance training, group exercise intervention coupled with a support group and home walking program utilizing facilities and personnel at a community cancer support center (The Gathering Place, Beachwood, Ohio) in AA Stage I-III BCa survivors who were within 12 months of completing treatment (surgery, chemotherapy, and/or breast irradiation); and, evaluated the potential effects of this intervention on physical measures and cancer-related biomarkers. 27 patients provided informed consent and 19 participated in the program. On average, attendance rates were 70.0% ± 19.1% for the exercise sessions and 63.1% ± 13.8% for the support group. We observed a significant decrease in circulating C-peptide levels (B: 893.9 ± 399.1 pg/mL; EOI: 723.9 ± 319.0 pg/mL; p=0.01). Although we did not observe a significant decrease in weight in the entire sample, there was a significant decrease in waist circumference and percent total body fat among those who attended 70% or more of the exercise sessions. In summary, we demonstrated that conducting lifestyle interventions in AA BCa survivors in a community setting is feasible. Future interventions should invoke strategies to enhance adherence and include a structured dietary intervention to enable greater weight loss.

A naturalistic evaluation of psychosocial interventions for cancer patients in a community setting.

We used a naturalistic methodology to examine associations between change in cancer patients' emotional functioning and their use of interventions in a community organization. One-hundred ninety-two patients completed measures at baseline and 6 months later. During this time, they utilized the organization's various interventions as they wished. Attendance at educational events was associated with decreased well-being. Use of art therapy groups was not associated with decreases in negative emotion but was consistently associated with increases in positive emotion. Improved functioning on some measures was associated with use of psychoeducational groups, expressive/supportive groups, movement classes, healing arts, and a buddy-matching service.

Coping-related variables associated with individual differences in adjustment to cancer.

This study examined relationships between emotional adjustment and a number of coping styles and strategies in people with cancer. Two-hundred eighty-three adults completed measures of positive and negative emotions, subjective ratings of cancer-related symptoms and functional impairment, coping strategies, hope, benefit finding, emotional approach/avoidance, and cancer-related social support. Among the coping strategies, self-blame and behavioral disengagement were consistently associated with poor adjustment, while acceptance and humor were consistently associated with good adjustment. Among the broader measures of coping style, there were associations between poor adjustment and emotional processing, and between good adjustment and hope, benefit finding, and cancer-related social support.