Skip to main content

Table 4 Logistic regression main effects and interactions of neighborhood poverty and primary payer by breast cancer stage at diagnosis on survival

From: Mediation of the effects of living in extremely poor neighborhoods by health insurance: breast cancer care and survival in California, 1996 to 2011

 

Local-Regional

Distally Metastasized

 

Node negative disease

Node positive disease

 
 

8 year survival

5 year survival

3 year survival

Predictor Variables

Sample

OR

(95% CI)

Sample

OR

(95% CI)

Sample

OR

(95% CI)

 

Single predictor models

Neighborhood poverty

  < 5% poor

1,435 

1.00

 

561 

1.00

 

55

1.00

 

  5-29% poor

1,353 

0.73

(0.61, 0.88)

574 

0.70

(0.53, 0.93)

97

0.76

(0.37, 1.59)

  ≥ 30% poor

1,247 

0.54

(0.45, 0.65)

639 

0.41

(0.31, 0.53)

135

0.63

(0.31, 1.29)

Primary payer

  Uninsured or Medicaid

572 

1.00

 

382 

1.00

 

84 

1.00

 

  Medicare or private

3,463 

1.57

(1.27, 1.96)

1,392 

1.86

(1.44, 2.41)

203 

2.51

(1.28, 4.92)

 

Full models

Neighborhood poverty

  < 5% poor

1,435 

1.00

 

561 

1.00

 

55 

1.00

 

  5-29% poor

1,353 

0.92

(0.67, 1.26)

574 

0.72

(0.55, 0.96)

97 

0.79

(0.38, 1.68)

  ≥ 30% poor

1,247 

0.84

(0.50, 1.40)

639 

0.44

(0.34, 0.58)

135 

0.76

(0.36, 1.59)

Primary payer

  Uninsured or Medicaid

572 

1.00

 

382 

1.00

 

84 

1.00

 

  Medicare or private

3,463 

1.89

(1.26, 2.84)

1,392 

1.59

(1.22, 2.07)

203 

2.40

(1.21, 4.79)

  Poverty by payer

4,035 

0.79*

(0.60, 1.04)

1,774 

1.20a

(0.85, 1.69)

287 

1.17a

(0.39, 3.54)

 

Poverty by payer interaction among women with node negative disease on 8 year survival

 

> 30% poor

5-29% poor

< 5% poor

 

Sample

OR

(95% CI)

Sample

OR

(95% CI)

Sample

OR

(95% CI)

Primary payer

  Uninsured or Medicaid

277 

1.00

 

186 

1.00

 

109 

1.00

 

  Medicare or private

970 

1.16

(0.82, 1.62)

1,167 

1.57

(1.09, 2.27)

1,326 

1.81

(1.11, 2.95)

  1. Notes. OR = odds ratio, CI = confidence interval. All effects were age-adjusted across these categories: 25–44, 45–54, 55–64, 65–74 and 75 or older. After age, poverty and payer were accounted for, place (large urban, smaller urban or rural) and race/ethnicity (person of color [32.1%] or non-Hispanic white) did not enter any of the full models. Bolded ORs were statistically significant at p < .05.
  2. a Null interaction was removed from the full model. *p < .10.