Skip to main content

Table 3 Multilevel Oaxaca–blinder decomposition of differences for Mexican Americans and non-latino whites in treatment completion

From: Identifying and reducing disparities in successful addiction treatment completion: testing the role of Medicaid payment acceptance

  b SE 95% CI p
Overall
 Mexican American 0.175 0.022 0.132, 0.218 < .001
 White 0.144 0.026 0.093, 0.196 < .001
 Difference 0.031 0.021 −0.012, 0.073 .154
 Explained 0.069 0.019 0.031, 0.106 < .001
 Unexplained −0.038 0.014 −0.066, −0.010 .009
Program variables
 Wave 2a −0.002 0.002 −0.006, 0.003 .492
 Medicaid payment 0.013 0.005 0.003, 0.023 .009
 Licensed 0.000 0.001 −0.001, 0.002 .669
 Accreditedb −0.001 0.006 −0.012, 0.010 .857
Cross-level interaction
 Wave × Mexican American 0.008 0.007 −0.005, 0.021 .230
 Medicaid × Mexican American 0.015 0.008 −0.001, 0.032 .065
Client variables
 Female 0.000 0.000 0.000, 0.001 .458
 Age −0.003 0.002 −0.006, 0.001 .147
 Education level 0.000 0.002 −0.005, 0.004 .843
Primary drugc
 Alcohol −0.005 0.001 −0.008, −0.002 < .001
 Methamphetamine 0.009 0.003 0.004, 0.015 .001
 Marijuana or hashish 0.005 0.002 0.002, 0.008 .003
 Other −0.003 0.001 −0.005, −0.002 < .001
 Days usedd 0.025 0.010 0.006, 0.044 .012
Age at first use 0.000 0.001 −0.001, 0.002 .714
Medicaid eligibility −0.001 0.002 −0.004, 0.002 .345
Mental health disorder 0.005 0.002 0.002, 0.009 .006
Treatment typee     
Methadone −0.003 0.003 −0.009, 0.003 .296
Residential 0.005 0.004 −0.002, 0.013 .154
  1. aWave 1 (2011) used as reference
  2. bAccreditation by the Joint Commission
  3. cHeroin used as reference
  4. dDuring 30 days prior to admission
  5. eOutpatient used as reference
  6. The 1.1% disparity is explained by differences in programs accepting Medicaid payments, professional accreditation, client Medicaid eligibility and treatment type, and differences unexplained by model covariates