Table 3.
Patient characteristic | Race * age interaction | Adjusted relative rate ratio (95% CI) a |
---|---|---|
All |
|
|
Gender |
|
|
Female |
|
1.05 (1.04,1.07) |
Male |
|
REF |
Diabetes |
|
|
Yes |
|
1.95 (1.93,1.98) |
No |
|
REF |
Race/ethnicity |
|
|
Asian |
|
0.43 (0.41,0.44) |
African-American |
|
0.64 (0.61,0.67) |
Hispanic |
|
0.73 (0.71,0.75) |
Native American |
|
0.94 (0.85,1.04) |
Multiracial |
|
0.96 (0.92,1.00) |
White |
|
REF |
Race/ethnicity * Age |
|
|
Asian |
Under 5 |
1.86 (1.74,2.00) |
|
5- < 18 |
1.18 (1.12,1.24) |
|
18- < 50 |
1.05 (1.01,1.10) |
|
50- < 65 |
0.86 (0.82,0.90) |
|
65+ |
REF |
African-American |
Under 5 |
1.51 (1.40,1.62) |
|
5- < 18 |
0.99 (0.94,1.05) |
|
18- < 50 |
1.59 (1.52,1.66) |
|
50- < 65 |
1.19 (1.14,1.25) |
|
65+ |
REF |
Hispanic |
Under 5 |
1.22 (1.14,1.30) |
|
5- < 18 |
0.92 (0.88,0.96) |
|
18- < 50 |
0.96 (0.93,0.99) |
|
50- < 65 |
0.91 (0.87,0.94) |
|
65+ |
REF |
Native American |
Under 5 |
0.99 (0.82,1.20) |
|
5- < 18 |
0.81 (0.70,0.93) |
|
18- < 50 |
0.98 (0.87,1.10) |
|
50- < 65 |
0.95 (0.84,1.08) |
|
65+ |
REF |
Multiracial |
Under 5 |
0.94 (0.85,1.03) |
|
5- < 18 |
0.74 (0.69,0.78) |
|
18- < 50 |
0.85 (0.81,0.89) |
|
50- < 65 |
0.87 (0.83,0.91) |
|
65+ |
REF |
White |
Under 5 |
0.95 (0.89,1.00) |
|
5- < 18 |
0.78 (0.75,0.81) |
|
18- < 50 |
0.80 (0.78,0.82) |
|
50- < 65 |
0.78 (0.76,0.81) |
65+ | REF |
Kaiser Permanente of Northern California, 2009 to 2011
a Rate ratios were derived from a single multivariate negative binomial model. The entire KPNC membership for years 2009–2011 were summarized into strata by gender, age, race, and diabetes status. Each record in the analytic dataset was one of these strata. The dependent variable was the number of SSTI episodes for all members in the strata. Independent variables were gender, race, race*age, and diabetes status. The log of member years in each strata was used as an offset term. Ref = reference group; CI = 95% confidence interval.