NEW ORLEANS – Hospitals with palliative care programs had lower treatment intensity on average at the end of life than did those without palliative care, in a national sample of 3,593 hospitals.
ICU length of stay in the last 6 months of life was 0.4 days shorter (P less than .001) and hospice length of stay 1.6 days longer (P = .013) at hospitals with palliative care versus those without.
The study strengthens claims that palliative care cuts costs, and is the first to examine the impact of palliative care in such a large national sample of hospitals, Jay R. Horton said at the annual meeting of the American Academy of Hospice and Palliative Medicine.
Prior studies focused on the effects of palliative care. Mr. Horton’s study considers outcomes for the entire older adult population in the hospital.
Patrice Wendling/IMNG Medical Media
Jay R. Horton
The researchers considered data from 3,593 hospitals with a palliative care status noted on the 2008 American Hospital Association survey. In all, 1,657 hospitals had palliative care programs and 1,936 hospitals did not. The researchers then linked the hospital data with the data from the Dartmouth Atlas on 896,097 fee-for-service Medicare patients, aged 67-99 years, with one or more chronic illnesses, who died in 2007. Patients were assigned to a hospital where they received the majority of their care in the last 2 years of life.
Covariates predictive of outcomes such as age, sex, race and comorbidities were already corrected for in the Dartmouth Atlas. Propensity scoring was used for variables predictive of outcomes or the presence of palliative care in the AHA survey such as Joint Commission Accreditation and total bed count. Finally, propensity scores were used to reweight the sample to reduce selection bias.
The effect of palliative care would very likely be stronger if the data had identified those patients who actually received palliative care, said Mr. Horton , director of the palliative care consult service at The Lilian and Benjamin Hertzberg Palliative Care Institute, Icahn School of Medicine at Mount Sinai Hospital, New York. Ongoing research uses data from the National Palliative Care Registry to better identify palliative care programs and socioeconomic factors to further reduce potential selection bias.
Treatment intensity for patients with serious illness varies widely across the country. One study showed that more than 40% of the variation is due to the supply of specialists and hospital capacity (BMJ 2002;325:961-4). Put another way, the greater the supply of physicians, the greater the utilization, even after adjustment for factors that should drive utilization, such as patient preference and disease severity.
"This supply-sensitive care, as it is sometimes called, is at the discretion of clinicians and to a certain extent at the discretion of patients, but more troubling is that much of this care may be unnecessary," said Mr. Horton.
Mr. Horton reported having no financial disclosures.