Prevalence of Sudden Death in Palliative Care: Data From the Australian Palliative Care Outcomes Collaboration.

CONTEXT

Advanced, life-limiting illnesses are likely to have a predictable functional decline through a terminal phase to death, but some patients may also die suddenly. To date, empirical evidence characterizing "sudden death" in hospice/palliative care is lacking.

OBJECTIVES

The aim of this study was to determine prevalence and clinicodemographic predictors of sudden death in hospice/palliative care.

METHODS

This is a longitudinal consecutive cohort study of prospectively collected national data in 104 specialist palliative care services from the Australian Palliative Care Outcomes Collaboration. Patients who died between July 1, 2013, and June 30, 2014, with one or more measurement of Australian-modified Karnofsky Performance Status (AKPS) in the last 30 days of life were included. "Sudden death" was defined as a lowest AKPS score of 50 or more in the last seven days of life and excluded anyone with "terminal phase" as their last phase before death. Predictors were defined using logistic regression.

RESULTS

In total, 13,966 patients were included, mean age 73.6 (SD 13.6) years, 46% women, and 77% had cancer. During the seven days before death, there were 20,992 AKPS measurements; median 1 (interquartile range 1-2) per patient. Four percent of deaths (one of 25) were sudden, predicted independently by having lung cancer (odds ratio [OR] 2.64), cardiovascular disease (OR 1.94), other cancers (OR 1.63), being male (OR 1.23), younger, worse fatigue, and worse breathlessness. Sudden death was associated with higher rates of death at home (OR 3.2; 95% CI 2.9 to 3.6).

CONCLUSION

This study quantifies rates of sudden death in hospice/palliative care and has implications for conversations about prognosis between clinicians, patients, and their families.