Mediation analysis to determine if insurance status is a significant mediator between SES and CRC survival
| SES Group . | Cox multivariable regressiona . | Cox multivariable regression with insuranceb (2 levels, uninsured + Medicaid vs. insured) added . | Cox multivariable regression with insuranceb (2 levels, uninsured vs. insured) added . | Cox multivariable regression with insuranceb (2 levels, no private vs. private insurance) added . | ||||
|---|---|---|---|---|---|---|---|---|
| Hazard Ratio (95% CI) . | FDR-p . | Hazard Ratio (95% CI) . | FDR-p . | Hazard Ratio (95% CI) . | FDR-p . | Hazard Ratio (95% CI) . | FDR-p . | |
| <.0001 | <.0001 | <.0001 | <.0001 | |||||
| Low | 1.35 (1.26–1.46) | 1.27 (1.18–1.36) | 1.33 (1.24–1.43) | 1.24 (1.15–1.33) | ||||
| Mid-Low | 1.29 (1.20–1.38) | 1.24 (1.16–1.33) | 1.28 (1.20–1.37) | 1.22 (1.14–1.31) | ||||
| Mid-high | 1.15 (1.07–1.23) | 1.13 (1.05–1.20) | 1.14 (1.07–1.22) | 1.11 (1.04–1.19) | ||||
| High | reference | reference | reference | reference | ||||
| SES Group . | Cox multivariable regressiona . | Cox multivariable regression with insuranceb (2 levels, uninsured + Medicaid vs. insured) added . | Cox multivariable regression with insuranceb (2 levels, uninsured vs. insured) added . | Cox multivariable regression with insuranceb (2 levels, no private vs. private insurance) added . | ||||
|---|---|---|---|---|---|---|---|---|
| Hazard Ratio (95% CI) . | FDR-p . | Hazard Ratio (95% CI) . | FDR-p . | Hazard Ratio (95% CI) . | FDR-p . | Hazard Ratio (95% CI) . | FDR-p . | |
| <.0001 | <.0001 | <.0001 | <.0001 | |||||
| Low | 1.35 (1.26–1.46) | 1.27 (1.18–1.36) | 1.33 (1.24–1.43) | 1.24 (1.15–1.33) | ||||
| Mid-Low | 1.29 (1.20–1.38) | 1.24 (1.16–1.33) | 1.28 (1.20–1.37) | 1.22 (1.14–1.31) | ||||
| Mid-high | 1.15 (1.07–1.23) | 1.13 (1.05–1.20) | 1.14 (1.07–1.22) | 1.11 (1.04–1.19) | ||||
| High | reference | reference | reference | reference | ||||
| . | Causal mediation analysisc with insurance (2 levels, uninsured + Medicaid vs insured) as the mediator . | Causal mediation analysisc with insurance (2 levels, uninsured vs insured) as the mediator . | Causal mediation analysisc with insurance (2 levels, no private vs private insurance) as the mediator . | |||
|---|---|---|---|---|---|---|
| . | Percentage (95% CI) of SES effect mediated by insurance . | p value . | Percentage (95% CI) of SES effect mediated by insurance . | p value . | Percentage (95% CI) of SES effect mediated by insurance . | p value . |
| 19.45 (11.83–27.06) | <.0001 | 5.10 (2.14–8.07) | .0007 | 31.19 (20.95–41.44) | <.0001 | |
| . | Causal mediation analysisc with insurance (2 levels, uninsured + Medicaid vs insured) as the mediator . | Causal mediation analysisc with insurance (2 levels, uninsured vs insured) as the mediator . | Causal mediation analysisc with insurance (2 levels, no private vs private insurance) as the mediator . | |||
|---|---|---|---|---|---|---|
| . | Percentage (95% CI) of SES effect mediated by insurance . | p value . | Percentage (95% CI) of SES effect mediated by insurance . | p value . | Percentage (95% CI) of SES effect mediated by insurance . | p value . |
| 19.45 (11.83–27.06) | <.0001 | 5.10 (2.14–8.07) | .0007 | 31.19 (20.95–41.44) | <.0001 | |
aThe outcome of the model was survival time. The exposure was SES.
The confounders were age, sex, race, ethnicity, stage of diagnosis, grade, side of tumor, surgery of the primary tumor, chemotherapy, comorbidity, and area of living.
bInsurance was added to the model to test its mediation effect.
cCausal mediation analysis was performed to study if insurance is a significant mediator for the effect of SES on survival status (alive, death) when the confounders were adjusted.
Mediation analysis to determine if insurance status is a significant mediator between SES and CRC survival
| SES Group . | Cox multivariable regressiona . | Cox multivariable regression with insuranceb (2 levels, uninsured + Medicaid vs. insured) added . | Cox multivariable regression with insuranceb (2 levels, uninsured vs. insured) added . | Cox multivariable regression with insuranceb (2 levels, no private vs. private insurance) added . | ||||
|---|---|---|---|---|---|---|---|---|
| Hazard Ratio (95% CI) . | FDR-p . | Hazard Ratio (95% CI) . | FDR-p . | Hazard Ratio (95% CI) . | FDR-p . | Hazard Ratio (95% CI) . | FDR-p . | |
| <.0001 | <.0001 | <.0001 | <.0001 | |||||
| Low | 1.35 (1.26–1.46) | 1.27 (1.18–1.36) | 1.33 (1.24–1.43) | 1.24 (1.15–1.33) | ||||
| Mid-Low | 1.29 (1.20–1.38) | 1.24 (1.16–1.33) | 1.28 (1.20–1.37) | 1.22 (1.14–1.31) | ||||
| Mid-high | 1.15 (1.07–1.23) | 1.13 (1.05–1.20) | 1.14 (1.07–1.22) | 1.11 (1.04–1.19) | ||||
| High | reference | reference | reference | reference | ||||
| SES Group . | Cox multivariable regressiona . | Cox multivariable regression with insuranceb (2 levels, uninsured + Medicaid vs. insured) added . | Cox multivariable regression with insuranceb (2 levels, uninsured vs. insured) added . | Cox multivariable regression with insuranceb (2 levels, no private vs. private insurance) added . | ||||
|---|---|---|---|---|---|---|---|---|
| Hazard Ratio (95% CI) . | FDR-p . | Hazard Ratio (95% CI) . | FDR-p . | Hazard Ratio (95% CI) . | FDR-p . | Hazard Ratio (95% CI) . | FDR-p . | |
| <.0001 | <.0001 | <.0001 | <.0001 | |||||
| Low | 1.35 (1.26–1.46) | 1.27 (1.18–1.36) | 1.33 (1.24–1.43) | 1.24 (1.15–1.33) | ||||
| Mid-Low | 1.29 (1.20–1.38) | 1.24 (1.16–1.33) | 1.28 (1.20–1.37) | 1.22 (1.14–1.31) | ||||
| Mid-high | 1.15 (1.07–1.23) | 1.13 (1.05–1.20) | 1.14 (1.07–1.22) | 1.11 (1.04–1.19) | ||||
| High | reference | reference | reference | reference | ||||
| . | Causal mediation analysisc with insurance (2 levels, uninsured + Medicaid vs insured) as the mediator . | Causal mediation analysisc with insurance (2 levels, uninsured vs insured) as the mediator . | Causal mediation analysisc with insurance (2 levels, no private vs private insurance) as the mediator . | |||
|---|---|---|---|---|---|---|
| . | Percentage (95% CI) of SES effect mediated by insurance . | p value . | Percentage (95% CI) of SES effect mediated by insurance . | p value . | Percentage (95% CI) of SES effect mediated by insurance . | p value . |
| 19.45 (11.83–27.06) | <.0001 | 5.10 (2.14–8.07) | .0007 | 31.19 (20.95–41.44) | <.0001 | |
| . | Causal mediation analysisc with insurance (2 levels, uninsured + Medicaid vs insured) as the mediator . | Causal mediation analysisc with insurance (2 levels, uninsured vs insured) as the mediator . | Causal mediation analysisc with insurance (2 levels, no private vs private insurance) as the mediator . | |||
|---|---|---|---|---|---|---|
| . | Percentage (95% CI) of SES effect mediated by insurance . | p value . | Percentage (95% CI) of SES effect mediated by insurance . | p value . | Percentage (95% CI) of SES effect mediated by insurance . | p value . |
| 19.45 (11.83–27.06) | <.0001 | 5.10 (2.14–8.07) | .0007 | 31.19 (20.95–41.44) | <.0001 | |
aThe outcome of the model was survival time. The exposure was SES.
The confounders were age, sex, race, ethnicity, stage of diagnosis, grade, side of tumor, surgery of the primary tumor, chemotherapy, comorbidity, and area of living.
bInsurance was added to the model to test its mediation effect.
cCausal mediation analysis was performed to study if insurance is a significant mediator for the effect of SES on survival status (alive, death) when the confounders were adjusted.
This PDF is available to Subscribers Only
View Article Abstract & Purchase OptionsFor full access to this pdf, sign in to an existing account, or purchase an annual subscription.
