Table 3.

Association between treatment-limiting directive and surrogate decision-maker with treatments in the last month of life and site of death among patients with ESRD

TreatmentTreatment-Limiting DirectiveAdjusted Risk Difference, % (95% CI)Surrogate Decision-MakerAdjusted Risk Difference, % (95% CI)
Absent % (n=21,364)Present % (n=9352)Absent % (n=25,221)Present % (n=5495)
Hospitalization7162−7 (−8 to −6)7062−4 (−6 to −3)
Intensive care unit admission4732−12 (−13 to −10)4434−5 (−6 to −3)
Mechanical ventilation196−10 (−11 to −9)1610−2 (−3 to −1)
CPR62−3 (−3 to −2)53−1 (−2 to 0)
Gastrostomy tube21−1 (−1 to 0)220 (0, 1)
Inpatient death4634−9 (−10 to −8)4435−4 (−6 to −3)
Hospice admission23335 (4 to 6)25311 (−1 to 2)
Dialysis discontinuation27416 (5 to 7)30412 (1 to 3)
  • Models adjusted for age, sex, race, duration of dialysis, length of nursing home stay, days in hospital in prior year, dialysis modality, functional status, diabetes, ischemic heart disease, heart failure, stroke, chronic liver disease, chronic lung disease, cancer, depression, dementia, impaired decision-making skills, hospital referral region spending quintile, treatment-limiting directive, and surrogate decision-maker, in addition to correlation of subjects within nursing home facilities. 95% CI, 95% confidence interval; CPR, cardiopulmonary resuscitation.