Skip to main content
. Author manuscript; available in PMC: 2011 Jul 1.
Published in final edited form as: Ann Emerg Med. 2011 Feb 21;58(1):24–32.e3. doi: 10.1016/j.annemergmed.2011.01.009

Table 3.

Posterior predicted probability of left without being seen for model 2: hospital-level socioeconomic status and hospital structural factors.*

Continuous Predictor 10th Percentile Predicted Probability of LWBS 90th Percentile Predicted Probability of LWBS Mean Difference in % LBWS (=90%−10%; [SD])
Age, y 31.3 2.52 45.8 2.43 −0.10 (0.43)
Men, % 43.0 2.67 50.1 2.27 −0.41 (0.53)
Minority, % 16.8 2.21 80.4 2.75 0.54 (0.70)
Income, $1,000 33.2 3.38 65.1 1.65 −1.73 (1.39)
Insurance, % (Med-Cal+uninsured) 15.5 2.24 68.7 2.77 0.53 (0.87)
Operating margin, % −12.9 2.29 12.2 2.68 0.39 (0.47)
Binary Predictor Predicted Probability of LWBS Difference Assessed Mean Difference (SD)
Base case (not for profit, nonteaching, nontrauma center) 2.47
 Ownership County: for profit 2.39 (2.28)
  County 5.02 County: not for profit 2.55 (2.35)
  For profit 2.63 For profit: not for profit 0.16 (0.47)
 Teaching 5.13 Yes/No 2.66 (2.34)
 Trauma center 3.93 Yes/No 1.46 (1.29)
*

Means of the continuous predictors: Age: 38.82 years; men: 46.31%; nonwhite: 49.88%; income (per $1,000): 47.14; insurance (Med-Cal+uninsured: 41.83%; operating margin: −0.67%. Binary predictors are set: ownership=not for profit, teaching=no, trauma center=no.