Association of selected predictors on age at transfer in years
| Selected predictors | Absolute effect (95% CI) |
| Last pediatric hemoglobin A1c | −0.17 (−0.46 to 0.12) |
| Number of pediatric hospital admissions per year | −0.35 (−1.70 to 1.00) |
| Number of pediatric emergency department visits per year | −0.96 (−1.92 to 0.00) |
| Number of pediatric endocrinology visits per year | 0.07 (−0.64 to 0.79) |
| No depression diagnosis in pediatrics* | 0.24 (−1.33 to 1.80) |
| No anxiety diagnosis in pediatrics* | −1.82 (−4.24 to 0.60) |
| No hypertension diagnosis in pediatrics* | −0.86 (−3.32 to 1.60) |
| No hyperlipidemia diagnosis in pediatrics* | 0.46 (−1.40 to 2.31) |
| No obesity diagnosis in pediatrics*† | −1.82 (−3.62 to to 0.02) |
| No mental health provider encounter in pediatrics*† | −1.17 (−3.34 to 1.00) |
| Female versus male | 0.12 (−1.29 to 1.53) |
| No diabetes educator visit in pediatrics*† | 2.62 (0.93 to 4.32) |
| No nutritionist visit in pediatrics* | −0.16 (−1.96 to 1.65) |
| No continuous glucose monitor use in pediatrics* | 0.31 (−1.54 to 2.17) |
| No insulin pump use in pediatrics* | −1.26 (−2.74 to 0.21) |
Cohort includes those with at least two pediatric endocrinology encounters and one adult endocrinology encounter at Duke. Participants who took longer than 36 months to transfer to adult care were excluded from the analysis. Full model included gender, race, ethnicity, payor type, presence of depression, anxiety, hypertension, hyperlipidemia, obesity, visit with mental health provider, number of hospital admissions per year, number of emergency department visits per year, visit with diabetes educator, visit with nutritionist, continuous glucose monitor use, insulin pump use, and pre-transfer hemoglobin A1c.
*Those who did have diagnosis or condition used as a reference group.
†P<0.05.