Multivariable Logistic Regression Models for Ever, High Blood Sugar, and Low Blood Sugar Rehospitalization

VariableORs (95% Confidence Interval) vs Never RH
Ever Rehospitalization (N = 508 [227 Case Subjects])PHigh Blood Sugar Rehospitalization (N = 463 [182 Case Subjects])PLow Blood Sugar Rehospitalization (N = 369 [88 Case Subjects])P
Duration of diabetes, per year1.26 (1.18–1.35)<.011.25 (1.17–1.35)<.011.26 (1.16–1.36)<.01
    Female1.67 (1.12–2.48)1.56 (1.02–2.39)1.90 (1.10–3.27)
    Nonwhitea1.94 (0.96–3.92)3.03 (1.27–7.26)1.50 (0.62–3.64)
Insurance status at questionnaireb<.01<.01.03
    Underinsured1.79 (1.20–2.67)1.85 (1.20–2.85)1.79 (1.04–3.27)
Presumed diabetes phenotype<.01<.01<.01
    Type 1ReferenceReferenceReference
    Non–type 10.32 (0.17–0.59)0.28 (0.14–0.56)0.15 (0.05–0.48)
Hospital of diabetes diagnosisc<.01<.01.28
    Tertiary care facilityReferenceReferenceReference
    Community hospital1.96 (1.30–2.97)2.42 (1.56–3.75)1.37 (0.77–2.44)
  • The number of case subjects in each multivariable regression model are lower than the overall number because of missing data on covariates. Other variables initially entered into models included head of household work status, head of household education, year of diabetes diagnosis, saw a health care provider in past year, and did not get medical attention when needed in past year.

  • a Nonwhite race/ethnicity includes NHB, Hispanic, and other/mixed.

  • b Private insurance consisted of those who reported using any health management organization, preferred provider organization, or fee-for-service plan not affiliated with Medicaid or the Illinois Department of Public Aid. Underinsurance included those lacking health insurance and those receiving Medicaid or Public Aid.

  • c Tertiary care facilities(n = 11) included facilities affiliated with a university, the large public hospital serving the medically indigent, and hospitals with pediatric endocrinology units. All of the other hospitals and clinics were deemed community hospitals.