Skip to main content
. 2019 Sep 11;1(3):210–218. doi: 10.1089/aut.2018.0036

Table 3.

Comparison of Two Zero-Inflated Negative Binomial Regression Models of Factors Associated with the Number of Primary Care Visits among Young Adults on the Autism Spectrum, 2010

  Regression 1+ sample a Regression 2+ sample b
IRR SE Lower 95% CI Upper 95% CI IRR SE Lower 95% CI Upper 95% CI
Primary care visits
 ASD and ID 0.986 0.089 0.826 1.178 0.926 0.100 0.750 1.144
 Age 22–27 1.046 0.110 0.851 1.284 1.003 0.127 0.782 1.286
 Minority 1.027 0.100 0.849 1.242 0.993 0.114 0.792 1.244
 Female 1.136 0.116 0.930 1.388 1.106 0.139 0.864 1.416
 Dual eligibility: Medicaid 1–12 months 0.903 0.122 0.693 1.176 0.937 0.149 0.687 1.278
 Psychiatric primary claim 0.914 0.102 0.735 1.136 0.814 0.106 0.630 1.051
 Log unscaled concurrent ACG score 1.810*** 0.086 1.649 1.988 1.883*** 0.117 1.666 2.128
 Constant 1.910*** 0.326 1.370 2.670 1.993*** 0.421 1.318 3.014
Inflation model (0 PCP visits)                
 Dual eligibility: Medicaid 1–12 months 0.906** 0.029 0.850 0.964 0.926 0.042 0.847 1.011
 Log unscaled concurrent ACG score 1.052 0.178 0.755 1.467 0.968 0.242 0.593 1.579
 Constant (coefficient) −0.710* 0.341 −1.370 −0.038 −1.066* 0.529 −2.102 −0.300
 Log alpha (coefficient) −1.150*** 0.200 −0.162 −0.036 −1.103*** 0.238 −1.57 −0.637
 Alpha (coefficient) 0.318 0.063 0.215 0.470 0.332 0.079 0.208 0.529

Data source: Centers for Medicare and Medicaid, Medicare Limited Data Set 2010 file for 5% carrier file (professional service file).

Reference group for indicator variables: ASD-only, 18–21 years, white, male, no dual Medicaid coverage, no primary psychiatric claim in past year.

*

p < 0.05; **p < 0.01; ***p < 0.001.

a

Number of observations = 527, number of zero observations = 168, Wald χ2 = 142.2, p < 0.001.

b

Number of observations = 364, number of zero observations = 107, Wald χ2 = 96.0, p < 0.001.

CI, confidence interval; IRR, incidence rate ratio; LR, likelihood ratio; SE, standard error.