TABLE 2

Race/Ethnicity Disparities in Diabetes Management and Clinical Outcomes

VariableNon-Hispanic WhiteNon-Hispanic BlackHispanicP: White Versus Black/White Versus Hispanic/Black Versus Hispanic
Overalln = 8841n = 697n = 1166
 Insulin pump usea5389 (61)181 (26)454 (39)<.001/<.001/<.001
 SMBGb6.1 ± 2.45.4 ± 2.35.6 ± 2.3<.001/<.001/.59
 HbA1cc8.4% ± 1.4%9.6% ± 1.9%8.7% ± 1.6%<.001/.01/<.001
 ≥1 SH event in past 12 mod328 (5)70 (13)49 (6)<.001/.62/<.001
 ≥1 DKA event in past 12 moe641 (7)153 (23)137 (12)<.001/.04/<.001
Annual household income <$50 000n = 1578n = 286n = 421
 Insulin pump usea702 (45)65 (23)120 (29)<.001/<.001/.07
 SMBGb5.8 ± 2.35.2 ± 2.55.4 ± 2.2.007/<.001/.44
 HbA1cc8.9% ± 1.5%9.9% ± 1.9%8.9% ± 1.6%<.001/.58/<.001
 ≥1 SH event in past 12 mod79 (7)38 (16)22 (6)<.001/.97/<.001
 ≥1 DKA event in past 12 moe200 (13)82 (29)67 (16)<.001/.18/<.001
Annual household income $50 000 to <$100 000n = 2481n = 101n = 198
 Insulin pump useb1549 (63)36 (36)102 (52)<.001/.02/<.001
 SMBGb6.1 ± 2.35.6 ± 1.95.8 ± 2.1.02/.03/.60
 HbA1cc8.4% ± 1.3%9.1% ± 1.6%8.4% ± 1.3%<.001/.94/<.001
 ≥1 SH event in past 12 mod85 (5)8 (11)9 (6).02/.22/.39
 ≥1 DKA event in past 12 moe168 (7)12 (12)17 (9).04/.85/.14
Annual household income ≥$100 000n = 2496n = 58n = 156
 Insulin pump useb1815 (73)26 (45)103 (66)<.001/.08/.01
 SMBGb6.5 ± 2.55.9 ± 1.86.4 ± 2.3.43/.12/.85
 HbA1cc8.1% ± 1.1%9.1% ± 1.9%8.2% ± 1.5%<.001/.17/<.001
 ≥1 SH event in past 12 mod78 (4)4(9)5 (4).35/.71/.64
 ≥1 DKA event in past 12 moe94 (4)12 (21)13 (8)<.001/.01/.02
Annual household income: not providedn = 2286n = 252n = 391
 Insulin pump useb1323 (58)54 (22)129 (33)<.001/<.001/.003
 SMBGb6.0 ± 2.55.3 ± 2.45.5 ± 2.5.04/.15/.52
 HbA1cc8.6% ± 1.6%9.7% ± 2.0%8.7% ± 1.7%<.001/.36/<.001
 ≥1 SH event in past 12 mod86 (5)20 (12)13 (5).003/.96/.03
 ≥1 DKA event in past 12 moe179 (9)47 (22)40 (11)<.001/.94/.002
  • Data are presented as n (%) or mean ± SD.

  • a A total of 44 participants were using both an insulin pump and injections and were not included in the analysis. Binary mixed models were adjusted for age, diabetes duration, interaction between age and diabetes duration, gender, highest parental education level (missing data imputed for ordinal variable and missing indicator included in model), annual household income (missing data imputed for ordinal variable and missing indicator included in model; not included in stratified models), and random site effect.

  • b A total of 394 participants were missing self-reported SMBG per day data (8552 white, 652 black, and 1106 Hispanic subjects). Linear mixed models were adjusted for age, insulin method, continuous glucose monitoring use, gender, highest parental education level (missing data imputed for ordinal variable and missing indicator included in model), annual household income (missing data imputed for ordinal variable and missing indicator included in model; not included in stratified models), and random site effect.

  • c Linear mixed models were adjusted for age, gender, BMI z score, highest parental education level (missing data imputed for ordinal variable and missing indicator included in model), annual household income (missing data imputed for ordinal variable and missing indicator included in model; not included in stratified models), and random site effect.

  • d A total of 4298 participants were missing SH data due to a change in how an SH event was defined during the enrollment period (6406 white, 525 black, and 1120 Hispanic subjects). Binary mixed models were adjusted for age, diabetes duration, gender, insulin method, SMBG per day, highest parental education level (missing data imputed for ordinal variable and missing indicator included in model), annual household income (missing data imputed for ordinal variable and missing indicator included in model; not included in stratified models), and random site effect.

  • e A total of 307 participants were missing DKA event data (8618 white, 659 black, and 1120 Hispanic subjects). Binary mixed models were adjusted for age, diabetes duration, gender, insulin method, SMBG per day, highest parental education level (missing data imputed for ordinal variable and missing indicator included in model), annual household income (missing data imputed for ordinal variable and missing indicator included in model; not included in stratified models), and random site effect.