Skip to main content

Table 2 Baseline characteristics of the analytic sample (N = 2674), before and after propensity score weighting on t1

From: Using predicted length of stay to define treatment and model costs in hospitalized adults with serious illness: an evaluation of palliative care

 

TG (n = 243)

CG (n = 2431)

Absolute Standardized difference

Unweighted

Weighted

Unweighted

Weighted

Age: over 75 years

490 (20%)

373 (15%)

490 (20%)

13%

< 0.01%

Gender: female

970 (40%)

970 (40%)

971 (40%)

< 0.5%

< 0.01%

Race: white

1341 (55%)

1123 (46%)

1341 (55%)

18%

< 0.01%

Surgery: first day

100 (4%)

374 (15%)

99 (4%)

39%

< 0.01%

ICU: first day

450 (19%)

882 (36%)

449 (18%)

41%

< 0.01%

Admitted: via ED

1371 (56%)

1180 (49%)

1371 (56%)

16%

< 0.01%

1ary dx: Solid tumor

1130 (47%)

480 (20%)

1131 (47%)

59%

< 0.01%

1ary dx: haematological cancer

10 (<0.5%)

58 (2%)

10 (<0.5%)

17%

< 0.01%

Charlson score: Mean (SD)

7.5 (2.4)

5.3 (2.9)

7.5 (2.9)

84%

< 0.01%

  1. TG Treatment group (t1 = 1), CG Comparison group (t1 = 0). Absolute standardized difference compares prevalence for binary variables, and mean and standard deviation for continuous variables, without taking into account sample size. It’s a standard measure of propensity score balance where < 10% is taken as a rule of thumb for acceptable balance [40].