Skip to main content
. 2016 May 17;31(10):1119–1126. doi: 10.1007/s11606-016-3728-9

Table 2.

Logistic regression model for completion of a POLST for the full population and stratified by long and short stay

Completion of a POLST OR (95 % CI)
Overall, N = 275,605 Long-stay, N = 92,376 Short-stay, N = 183,229
Long-stay resident 2.36 (2.30, 2.42)
Cognitive impairment
 None REF REF REF
 Mild/moderate 0.98 (0.95, 1.01) 0.88 (0.83, 0.93) 1.01 (0.98, 1.05)
 Severe 0.89 (0.87, 0.92) 0.76 (0.72, 0.80) 0.95 (0.92, 0.98)
ADL score
 0–7 (least dependent) REF REF REF
 8–14 1.04 (1.00, 1.08) 1.14 (1.06, 1.24) 1.01 (0.97, 1.06)
 15–20 1.03 (0.99, 1.07) 1.19 (1.10, 1.27) 0.97 (0.93, 1.02)
 21–28 (most dependent) 0.97 (0.93, 1.01) 1.07 (0.99, 1.15) 0.94 (0.83, 0.996)
Age (per additional 5 years) 1.03 (1.03, 1.04) 1.04 (1.04, 1.05) 1.03 (1.03, 1.04)
Female 1.02 (1.003, 1.05) 1.01 (0.98, 1.05) 1.03 (1.01, 1.06)
Race/ethnicity
 White, non-Hispanic REF REF REF
 Hispanic 0.99 (0.96, 1.02) 0.98 (0.93, 1.04) 0.995 (0.96, 1.04)
 Black 1.03 (0.99, 1.07) 1.01 (0.95, 1.09) 1.04 (0.99, 1.10)
 Asian 1.05 (1.01, 1.10) 1.05 (0.97, 1.04) 1.05 (0.99, 1.10)
 Other/unknown 1.07 (1.02, 1.13) 1.06 (0.96, 1.18) 1.09 (1.02, 1.16)

The logistic regression model includes the variables listed plus the nursing home facility as a random effect. McFadden’s pseudo R2 for the models range from 0.34 to 0.38 (likelihood ratio test of the nursing home random effect p < 0.001)

Odds ratios (ORs) in bold are significantly different from the reference at p < 0.05. CI = confidence interval

Model of completion of a POLST excludes 4.9 % (N = 14,148) of residents missing cognitive status or ADL score and excludes one nursing home facility (N = 1,219 nursing facilities). Long-stay model excludes 1104 residents (1.2 %) and short-stay model excludes 13,044 residents (6.6 %)

Long stay is defined as >100 consecutive days in the nursing home. Data are from the last MDS Section S completed in 2011 for each resident