Skip to main content
. 2021 Feb 17;4(3):e00235. doi: 10.1002/edm2.235

TABLE 2.

Multivariable logistic regression model of 385 patients in respect to DKA at time of diagnosis of T1D

Parameter OR (95% CI) p‐value
Age
0‐3 4.16 (1.64‐10.6) 0.003
4‐8 1.26 (0.64‐2.48) 0.50
9‐13 2.21 (1.22‐4.00) 0.009
14‐18 (ref) 1
Sex
Male 1.26 (0.78‐2.01) 0.34
Female (ref) 1
Race
White 1.70 (0.97‐2.99) 0.07
Other (ref) 1
BMI percentile
<15th percentile 1.11 (0.57‐2.17) 0.75
15th to 85th percentile 0.69 (0.39‐1.23) 0.21
>85th percentile (ref) 1
Insurance
Private (ref) 1
Medicaid 1.21 (0.70‐2.07) 0.50
Other 1.00 (0.45‐2.25) 0.99
Outcome of PCP healthcare encounter
Referred to ED for concern of T1D (ref) 1
Misdiagnosis and discharged without concern of T1D 3.73 (1.72‐8.11) 0.001
No PCP Visit (went to ED directly) 1.76 (1.04‐2.96) 0.034
A1c
≤10% (13.4mmol/L) (ref) 1
>10% (13.4mmol/L) 5.81 (2.96‐11.4) <0.001
Altered mental status
Yes 9.73 (1.21‐78.0) 0.032
No (ref) 1
Autism
Yes 0.94 (0.3‐2.34) 0.90
No (ref) 1
COVID era
Pre‐COVID (ref) 1
Post‐COVID 0.94 (0.40‐2.21) 0.89

A total of 27 patients (6.6%) had a missing value for at least one of the variables included in the model. These patients were excluded when fitting the model. Odds ratios (ORs) and corresponding 95% CIs for parameters in the model are shown. Bold font signifies p‐value ≤ 0.05.

Abbreviations: ED, emergency department; PCP, primary care provider.