Healthcare access and mammography screening in Michigan: a multilevel cross-sectional study

Table 3 Multivariate Multilevel Models for Recent Mammography by Individual and County Level Characteristics

Characteristic	Model 1: Demographics ^a	Model 2: + County^b
Race
White	0.63 (0.34-1.16)	0.64 (0.32-1.28)
Black (Ref.)	-	-
Age
50-60	1.16 (0.71-1.91)	1.12 (0.68-1.85)
61-74 (Ref.)	-	-
Income
< 35,000	0.33 (0.14-0.76)**	0.32 (0.14-0.74)**
35,000-74,999	0.47 (0.21-1.05)	0.46 (0.21-1.03)
> 75,000 (Ref.)	-	-
Employment
Unemployed	0.85 (0.46-1.56)	0.90 (0.48-1.67)
Retired/Unable	1.15 (0.66-2.00)	1.13 (0.65-1.99)
Employed (Ref.)	-	-
Education
Less than High School	2.19 (1.21-3.96)**	2.05 (1.12-3.75)*
Some College	1.49 (0.85-2.60)	1.45 (0.82-2.56)
College plus	-	-
Marital Status
Single	0.50 (0.25-1.02)	0.51 (0.25-1.03)
Divorced/Separated	0.74 (0.43-1.27)	0.75 (0.43-1.31)
Widowed	1.24 (0.62-2.48)	1.22 (0.60-2.44)
Married (Ref.)	-	-
Usual Source of Care
No	0.46 (0.21-0.99)*	0.46 (0.21-1.02)
Yes (Ref.)	-	-
Usual Healthcare Provider
No	0.32 (0.15-0.69)**	0.33 (0.15-0.71)**
Yes (Ref.)	-	-
Health Insurance Status
No	0.27 (0.14-0.54)**	0.26 (0.13-0.52)**
Yes (Ref.)	-	-
Facilities ^c
Low		1.55 (0.76-3.14)
High (Ref.)		-
Personnel ^c
Low		1.16 (0.48-2.80)
High (Ref.)		-
Concentrated Affluence ^d
Low		2.23 (0.89-5.56)
High (Ref.)		-
Concentrated Disadvantage ^e
Low		1.26 (0.55-2.87)
High (Ref.)		-
Concentrated Immigration ^f
Low		0.82 (0.32-2.14)
High (Ref.)		-
Percent black ^g
< 6%		0.88 (0.37-2.09)
> = 6% (Ref.)		-

^aModel adjusting for individual demographic variables only
^bModel adjusting for county level variables including healthcare access and SES in addition to individual demographic variables
^cPersonnel and Facilities, two measures of health care access were defined using principal components analysis on the count per 10,000 population of several variables by county: hospitals, mammography facilities, MDs, DOs and nurse practitioners
^dConcentrated affluence was based on PCA of county level proportion of individuals making over $100,000, proportion over 25 years with a college degree and the proportion of white collar workers
^eConcentrated disadvantage was based on PCA of county level proportion of persons in poverty, proportion of families with a female-headed household, proportion of households that are food stamp recipients, and proportion of unemployed
^fConcentrated immigration was based on PCA of the proportion of foreign born individuals and the proportion of non-English speaking individuals
*P < 0.05, **p < 0.01, ***p < 0.001; Ref, reference group

ISSN: 1475-9276