Table 4. Oaxaca-Blinder decomposition analysis for the racial disparity in influenza vaccination between Black and White* KPMAS patients.
| Effect estimate** | Probability of Vaccine | SE | P-value | Lower 95% CL | Upper 95% CL | Percentage of Total Difference | |
|---|---|---|---|---|---|---|---|
| PANEL A: | White | 0.5793 | 0.0017 | 0.0000 | 0.5759 | 0.5826 | |
| Overall effect | Black | 0.4082 | 0.0015 | 0.0000 | 0.4052 | 0.4112 | |
| Total difference | 0.1711 | 0.0023 | 0.0000 | 0.1666 | 0.1756 | 100.00 | |
| Difference-in-covariates | 0.0503 | 0.0022 | 0.0000 | 0.0460 | 0.0546 | 29.42 | |
| Difference-in-coefficients | 0.1208 | 0.0030 | 0.0000 | 0.1149 | 0.1266 | 70.58 | |
| PANEL B: | Age (per 10 years) | 0.0280 | 0.0008 | 0.0000 | 0.0265 | 0.0294 | 16.34 |
| Covariate effect | Female | -0.0025 | 0.0002 | 0.0000 | -0.0030 | -0.0021 | -1.47 |
| Service area | |||||||
| Greater Baltimore area | 0.0012 | 0.0002 | 0.0000 | 0.0008 | 0.0016 | 0.72 | |
| District of Columbia | -0.0012 | 0.0006 | 0.0558 | -0.0025 | 0.0000 | -0.71 | |
| Suburban Maryland | -0.0103 | 0.0012 | 0.0000 | -0.0127 | -0.0080 | -6.05 | |
| Current smoker | 0.0011 | 0.0001 | 0.0000 | 0.0008 | 0.0013 | 0.64 | |
| Registered on KP.org | 0.0214 | 0.0008 | 0.0000 | 0.0198 | 0.0230 | 12.51 | |
| Neighborhood median income (per $10,000) | 0.0190 | 0.0013 | 0.0000 | 0.0163 | 0.0216 | 11.09 | |
| Miles to nearest medical center | -0.0022 | 0.0002 | 0.0000 | -0.0026 | -0.0019 | -1.31 | |
| CCI = 1 | -0.0016 | 0.0002 | 0.0000 | -0.0020 | -0.0013 | -0.95 | |
| CCI ≥ 2 | -0.0027 | 0.0003 | 0.0000 | -0.0032 | -0.0021 | -1.55 | |
| Discordant patient/provider race | 0.0003 | 0.0001 | 0.0001 | 0.0001 | 0.0004 | 0.16 | |
| PANEL C: | Age (per 10 years) | 0.0500 | 0.0075 | 0.0000 | 0.0353 | 0.0646 | 29.21 |
| Coefficient effect | Female | 0.0060 | 0.0028 | 0.0312 | 0.0005 | 0.0115 | 3.51 |
| Service area | |||||||
| Greater Baltimore area | -0.0019 | 0.0010 | 0.0501 | -0.0037 | 0.0000 | -1.09 | |
| District of Columbia | 0.0061 | 0.0014 | 0.0000 | 0.0033 | 0.0088 | 3.54 | |
| Suburban Maryland | 0.0267 | 0.0033 | 0.0000 | 0.0203 | 0.0332 | 15.63 | |
| Current smoker | -0.0036 | 0.0007 | 0.0000 | -0.0049 | -0.0023 | -2.11 | |
| Registered on KP.org | 0.0290 | 0.0038 | 0.0000 | 0.0216 | 0.0364 | 16.95 | |
| Neighborhood median income (per $10,000) | 0.0211 | 0.0051 | 0.0000 | 0.0111 | 0.0311 | 12.34 | |
| Miles to nearest medical center | -0.0149 | 0.0029 | 0.0000 | -0.0206 | -0.0092 | -8.71 | |
| CCI = 1 | -0.0052 | 0.0011 | 0.0000 | -0.0075 | -0.0030 | -3.06 | |
| CCI ≥ 2 | -0.0114 | 0.0014 | 0.0000 | -0.0142 | -0.0087 | -6.67 | |
| Discordant patient/provider race | -0.0337 | 0.0029 | 0.0000 | -0.0393 | -0.0280 | -19.67 | |
| Intercept | 0.0526 | 0.0128 | 0.0000 | 0.0274 | 0.0777 | 30.72 |
KPMAS = Kaiser Permanente Mid-Atlantic States; SE = standard error; CCI = Charlson Comorbidity Index; CL = confidence limit
*White is the reference group
**The "Covariate effect" represents differences in probability of vaccination due to covariate values; the "Coefficient effect" represents differences in probability of vaccination due to other effects, including discrimination and unmeasured covariates.