Greater preferences for death in hospital and mechanical ventilation at the end of life among non-whites recently diagnosed with cancer

Factors and Barriers to Goals-of-Care Conversations for Patients With Cancer and Inpatient Mortality

PURPOSE

Conversations about personal values and goals of care (GOC) at the end of life are essential in caring for patients with advanced cancer. However, GOC conversations may be influenced by patient and oncologist factors during transitions of care.

METHODS

We electronically administered surveys to medical oncologists of inpatients who died from May 1, 2020, to May 31, 2021. Primary outcomes included oncologists' knowledge of inpatient death, anticipation of patient death, and recollection of GOC discussions. Secondary outcomes, including GOC documentation and advance directives (ADs), were collected retrospectively from electronic health records. Outcomes were analyzed for association with patient, oncologist, and patient-oncologist relationship factors.

RESULTS

For 75 patients who died, 104/158 (66%) surveys were completed by 40 inpatient and 64 outpatient oncologists. Eighty-one oncologists (77.9%) were aware of patients' deaths, 68 (65.4%) anticipated patients' deaths within 6 months, and 67 (64.4%) recalled having GOC discussions before or during the terminal hospitalization. Outpatient oncologists were more likely to report knowledge of patient death (P < .001), as were those with longer therapeutic relationships (P < .001). Inpatient oncologists were more likely to correctly anticipate patient death (P = .014). Secondary outcomes revealed 21.3% of patients had documented GOC discussions before admission and 33.3% had ADs; patients with a longer duration of cancer diagnosis were more likely to have ADs (P = .003). Oncologist-reported barriers to GOC included unrealistic expectations from patients or family (25%) and decreased patient participation because of clinical conditions (15%).

CONCLUSION

Most oncologists recalled having GOC discussions for patients with inpatient mortality, yet documentation of serious illness conversations remained suboptimal. Further studies are needed to examine barriers to GOC conversations and documentation during transitions of care and across health care settings.

Characteristics Associated With Mexican-American Hospice Use: Retrospective Cohort Study Using the Hispanic Established Population for the Epidemiologic Study of the Elderly (H-EPESE)

BACKGROUND

Hospice use is lower among ethnic/racial minorities in the United States, though little is known about trends, associated factors and duration of hospice use by Mexican-Americans.

AIM

The purpose of this study is to examine Mexican-American characteristics associated with hospice stay, both ≤ and > 7 days.

DESIGN

This retrospective cohort study used data from the Hispanic Established Population for the Epidemiological Study of the Elderly (H-EPESE) and the Centers for Medicare and Medicaid Services. Multivariate logistic regression models were used to estimate the ORs and 95% CIs for hospice stay among Mexican-Americans, both ≤ and > 7 days.

SETTING AND PARTICIPANTS

The first cohort (N = 970) includes H-EPESE participants who died between 2004 and 2016 who had Medicare parts A and B. The second cohort (N = 403) includes participants who completed the H-EPESE survey within the last 2 years of life.

RESULTS

Although hospice use increased among Mexican-Americans between 2004 and 2016 (OR 1.88, 95% CI 1.19-2.97), 38% of participants died within the first week of hospice care. Mexican-Americans in New Mexico and Arizona were 2-4 times more likely to use hospice than those in Texas and Colorado. Dementia was associated with hospice use (OR 1.47, 95% CI 1.11-1.94). Characteristics, like church attendance and living alone, were not associated with hospice use.

CONCLUSIONS

The substantial proportion of Mexican-Americans with 7 days or less of hospice use underscores the need for early palliative/hospice intervention to mitigate variation in use.