The researchers analyzed data from 740,890 patients with 1,024,666 hospitalizations (82% had one hospitalization). They divided subjects into three cohorts, one of which was a category of patients with criteria for palliative care based on previous research (PC-ICU). The PC-ICU cohort included patients with an expected length of stay more than 2 weeks, patients receiving extracorporeal membrane oxygenation (ECMO), severe brain injuries, acute respiratory failure with serious comorbidity, hematologic or oncologic disease, metabolic disease, renal failure that required continuous renal replacement therapy, hepatic failure, or serious chromosomal abnormality. A second cohort included chronic complex conditions not found in the PC-ICU cohort (additional criteria), and a third cohort had no criteria for palliative care.
Thirty percent of hospitalizations met the PC-ICU cohort criteria, 40% met the additional cohort criteria, and 30% fell in the no criteria cohort. The PC-ICU group had the highest mortality, at 8.03%, compared with 1.08% in the additional criteria group and 0.34% in the no criteria group (P less than .00001).
Palliative care consultations occurred more frequently in 5-12 year olds (odds ratio 1.06; 95% confidence interval, 1.01-1.13) and in those aged 13 years or older (OR, 1.38; 95% CI, 1.3-1.46), in females (OR, 1.13; 95% CI, 1.06-1.15), and in patients with government insurance (OR, 1.23; 95% CI, 1.17-1.29). Compared with those in the no criteria cohort, PC-ICU patients were more likely to receive a palliative care consult (OR, 75.5; 95% CI, 60.4-94.3), as were those in the additional criteria group (OR, 19.1; 95% CI, 15.3-23.9).
Cross-institutional palliative care frequency varied widely among patients in the PC-ICU group, ranging from 0% to 44%. The frequency ranged from 0% to 12% across institutions for patients in the additional criteria group.
SOURCE: O’Keefe S et al. Critical Care Congress 2019, Abstract 418.
© Frontline Medical Communications 2018-2021. Reprinted with permission, all rights reserved.
