Health care utilization at end of life among patients with lung or pancreatic cancer. Comparison between two Swedish cohorts

Initial treatment is associated with improved survival and end-of-life outcomes for patients with pancreatic cancer: a cohort study

Background

We describe the association between initial treatment and end-of-life (EOL) outcomes among patients with pancreatic ductal adenocarcinoma (PDAC).

Methods

This population-based cohort study included patients with PDAC who died from April 2010–December 2017 in Ontario, Canada using administrative databases. We used multivariable models to explore the association between index cancer treatment (no cancer-directed therapy, radiation, chemotherapy, surgery alone, and surgery and chemotherapy), and primary (mortality, healthcare encounters and palliative care) and secondary outcomes (location of death, hospitalizations, and receipt of chemotherapy within the last 30 days of life).

Results

In our cohort (N = 9950), 56% received no cancer-directed therapy, 5% underwent radiation, 27% underwent chemotherapy, 7% underwent surgery alone, and 6% underwent surgery and chemotherapy. Compared to no cancer-directed therapy, radiation therapy (HR = 0.63), chemotherapy (HR = 0.43) surgery alone (HR = 0.32), and surgery and chemotherapy (HR = 0.23) were all associated with decreased mortality. Radiation (AMD = − 3.64), chemotherapy (AMD = -6.35), surgery alone (AMD = -6.91), and surgery and chemotherapy (AMD = -6.74) were all associated with fewer healthcare encounters per 30 days in the last 6 months of life. Chemotherapy (AMD = -1.57), surgery alone (AMD = -1.65), and surgery and chemotherapy (AMD = -1.67) were associated with fewer palliative care visits (all p-values for estimates above < 0.05). Treatment groups were associated with lower odds of institutional death and hospitalization at EOL, and higher odds of chemotherapy at EOL.

Conclusions

Receiving cancer-directed therapies was associated with higher survival, fewer healthcare visits, lower odds of dying in an institution and hospitalization at EOL, fewer palliative care visits, and higher odds of receiving chemotherapy at EOL.

Exploring Health Care Professionals' Perceptions Regarding Shared Clinical Decision-Making in Both Acute and Palliative Cancer Care

Developments in cancer care have resulted in improved survival and quality of life. Integration of acute and palliative cancer care is desirable, but not always achieved. Fragmented care is associated with sub-optimal communication and collaboration, resulting in unnecessary care transitions. The aim of this study was to explore how health care professionals, from both acute and palliative care, perceive clinical decision-making when caring for patients undergoing active cancer treatment in parallel with specialized palliative care at home.

METHODS

Qualitative explorative design, using online focus-group interviews, based on patient-cases, among health care professionals (physicians and nurses) and Framework Analysis.

RESULTS

Six online focus-group interviews were performed. Few signs of systematic integration were found, risking fragmented care, and putting the patients in a vulnerable situation. Different aspects of uncertainty related to mandates and goals-of-care impacted clinical decision-making. Organizational factors appeared to hinder mutual clinical decision-making as well as the uncertainty related to responsibilities. These uncertainties seemed to be a barrier to timely end-of-life conversations and clinical decisions on optimal care, for example, the appropriateness of transfer to acute care.

CONCLUSIONS

Lack of integration between acute and palliative care have negative consequences for patients (fragmented care), health care professionals (ethical stress), and the health care system (inadequate use of resources).

Evaluation of Whole Brain Radiotherapy among Lung Cancer Patients with Brain Metastases in Relation to Health Care Level and Survival

Whole-brain radiotherapy (WBRT) as a treatment for brain metastases has been questioned over the last years. This study aimed to evaluate health care levels and survival after WBRT in a cohort of lung cancer patients with brain metastases receiving WBRT in Stockholm, Sweden, from 2008 to 2019 (n = 384). If the patients were able to come home again was estimated using logistic regression and odds ratios (OR) and survival by using Cox regression. The median age in the cohort was 65.6 years, the median survival following WBRT was 2.4 months (interquartile range (IQR) 1.2–6.2 months), and 84 (22%) patients were not able to come home after treatment. Significantly more males could come home again after WBRT compared to women (OR = 0.37, 95%CI 0.20–0.68). Patients with performance status scores WHO 3–4 had a median survival of 1.0 months, hazard ratio (HR) = 4.69 (95%CI 3.31–6.64) versus WHO score 0–1. Patients admitted to a palliative ward before WBRT had a median survival of 0.85 months, HR = 2.26 (95%CI 1.53–3.34) versus being at home. In conclusion, patients treated with WBRT had a short median survival and 20% could not be discharged from the hospital following treatment. Significantly more women did not come home again.

Machine Learning-Based Short-Term Mortality Prediction Models for Patients With Cancer Using Electronic Health Record Data: Systematic Review and Critical Appraisal

BACKGROUND

In the United States, national guidelines suggest that aggressive cancer care should be avoided in the final months of life. However, guideline compliance currently requires clinicians to make judgments based on their experience as to when a patient is nearing the end of their life. Machine learning (ML) algorithms may facilitate improved end-of-life care provision for patients with cancer by identifying patients at risk of short-term mortality.

OBJECTIVE

This study aims to summarize the evidence for applying ML in ≤1-year cancer mortality prediction to assist with the transition to end-of-life care for patients with cancer.

METHODS

We searched MEDLINE, Embase, Scopus, Web of Science, and IEEE to identify relevant articles. We included studies describing ML algorithms predicting ≤1-year mortality in patients of oncology. We used the prediction model risk of bias assessment tool to assess the quality of the included studies.

RESULTS

We included 15 articles involving 110,058 patients in the final synthesis. Of the 15 studies, 12 (80%) had a high or unclear risk of bias. The model performance was good: the area under the receiver operating characteristic curve ranged from 0.72 to 0.92. We identified common issues leading to biased models, including using a single performance metric, incomplete reporting of or inappropriate modeling practice, and small sample size.

CONCLUSIONS

We found encouraging signs of ML performance in predicting short-term cancer mortality. Nevertheless, no included ML algorithms are suitable for clinical practice at the current stage because of the high risk of bias and uncertainty regarding real-world performance. Further research is needed to develop ML models using the modern standards of algorithm development and reporting.