Life-sustaining Orders and End-of-Life Care

End-of-life care is often a focus for supposed wasteful health spending.  One way to limit intensive and probably futile end-of-life care is to have a patient provide guidance on what kind of treatment they prefer in certain situations.  The effectiveness of those orders depends upon their being followed.  A study in the Journal of the American Medical Association examines use of and compliance with Physician Orders for Life-Sustaining Treatment, a sort of advance directive worked out with patients and supposedly available to all treating physicians.   (JAMA Article)   The researchers were particularly interested in use of ICU care, which is very expensive.  They conducted a retrospective study of patients who had such advance orders and were hospitalized near end-of-life at a large academic medical center.  The patients included all had what were considered to be life-limiting diseases or conditions.  In addition to admission to an ICU, use of certain treatments like mechanical ventilation, vasoactive infusions, dialysis or CPR were used as outcome measures.  About 1800 patients ended up being included in the analysis.  The average age was about 71.  39% had cancer and 26% had dementia.  36% had advance directives for full treatment, 42% put limits on treatment and 22% wanted comfort measures only.  48% of the patients ended up being admitted to the ICU during their hospitalization, including 31% of those with comfort measures only POLSTs.  Treatment-limiting advance directives were associated with significantly less use of ICU and and other life-extending treatments.  However, a large number of patients who requested comfort only measures still ended up in the ICU and ended up receiving other interventions.  This POLST discordant care rate was 30% for the comfort-measures only group and 41% in the limited interventions group.  Patients with cancer were less likely to receive POLST-discordant care, as were those with dementia, potentially because physicians and family recognize that these are almost certainly end-of-life conditions.  Patients admitted for traumatic injury were more likely to receive discordant treatment.    Since all the patients included in this study by definition died, all this extra care was also by definition futile.  The results suggest that some savings could be achieved, and likely patient quality of life near death would be improved, by greater compliance with advance directives regarding life-sustaining care.