Impact of advanced healthcare directives on treatment decisions by physicians in patients with acute stroke.

BACKGROUND: The implementation of advanced healthcare directives, prepared by almost half of the adult population in United States remains relatively under studied. We determined the impact of advanced healthcare directives on treatment decisions by multiple physicians in stroke patients.
METHODS: A deidentified summary of clinical and radiological records of 28 patients with stroke was given to six stroke physicians who were not involved in the care of the patients. Each physician independently rated 28 treatment decisions per patient in the presence or absence of advanced healthcare directives 1 month apart to allow memory washout. The percentage agreement to treat/intervene per patient and proportion of treatment withheld as a group were estimated for each of the 28 treatment decision items. We also determined the interobserver reliability between the two raters (attorneys) in interpretation of six items characterizing the adequacy of documentation within the 28 advanced healthcare directives.
RESULTS: The percentage agreement among physician raters for treatment decisions in 28 stroke patients was highest for treatment of hyperpyrexia (100%, 100%) and lowest for ICU monitoring duration based on family-physician considerations outside of accepted criteria within institution (68%, 69%) in presence and absence of advanced healthcare directives. The physician rater agreement in choosing "yes" was highest for "routine-complexity" treatment decisions and lowest for "moderate-complexity" treatment decisions. The choice of withholding treatment in "routine-complexity," "moderate-complexity," or "high-complexity" treatment decisions was remarkably similar among raters in presence or absence of advanced healthcare directives. The only treatment decision that showed an impact of advanced healthcare directives was ICU monitoring withheld in 32% of treatment decisions in presence of directives (compared with 8% in the absence of directives). IV medication and defibrillation for cardiac arrest was withheld in 29% (compared with 19%) of the treatment decisions in the presence of advanced healthcare directives. The two attorney raters found the description of acceptable outcome inadequate in 14 and 21 of 28 advanced healthcare directives reviewed, respectively. The overall mean kappa for agreement regarding adequacy of documentation was modest (43%) for "does the advanced healthcare directive specify which treatments the patient would choose, or refuse to receive if they were diagnosed with an acute, terminal condition?" and lowest (3%) for "description of acceptable outcome."
CONCLUSIONS: We did not find any prominent differences in most "routine-complexity," "moderate-complexity," or "high-complexity" treatment decisions in patient management in the presence of advanced healthcare directives. Presence of advanced healthcare directives also did not reduce the prominent variance among physicians in treatment decisions.