The association of hypoglycemia with outcome of critically ill children in relation to nutritional and blood glucose control strategies

Table 2 Multivariable regression analyses investigating the independent association of hypoglycemia in PICU with outcome

	Patients with vs. patients without hypoglycemia	P	Patients with vs. without hypoglycemia; after further adjustment for baseline NSE/S100B	P
Mortality	Odds ratio (95% CI)		Odds ratio (95% CI)
In PICU	1.20 (0.51–2.83)	0.69	–
At 90 days	0.66 (0.31–1.43)	0.30	–
At 4 years	0.71 (0.37–1.36)	0.31	–
Neurodevelopmental outcome after 4 years¹	β-Estimate (95% CI)		β-Estimate (95% CI)
Clinical neurological evaluation score², range 0–8	0.09 (− 0.06 to 0.23)	0.25	0.07 (− 0.09 to 0.22)	0.39
Executive functioning as reported by parents or caregivers, T-score²
Inhibition	1.26 (− 0.32 to 2.84)	0.12	1.09 (− 0.59 to 2.77)	0.20
Flexibility	0.01 (− 1.41 to 1.44)	0.99	0.16 (− 1.31 to 1.64)	0.83
Emotional control	− 0.08 (− 1.52 to 1.35)	0.91	− 0.17 (− 1.70 to 1.36)	0.83
Working memory	2.17 (0.60 to 3.75)	0.007	2.36 (0.68 to 4.04)	0.006
Planning and organization	2.03 (0.53 to 3.53)	0.008	2.27 (0.65 to 3.89)	0.006
Metacognition index	1.85 (0.28 to 3.42)	0.021	2.08 (0.39 to 3.76)	0.016
Total score	1.36 (− 0.25 to 2.96)	0.10	1.52 (− 0.20 to 3.23)	0.084
Emotional and behavioral problems as reported by parents or caregivers, T-score²
Internalizing problems	− 0.11 (− 1.60 to 1.37)	0.88	− 0.38 (− 1.95 to 1.19)	0.64
Externalizing problems	− 0.16 (− 1.54 to 1.22)	0.82	− 0.34 (− 1.80 to 1.11)	0.64
Total problems	− 0.14 (− 1.60 to 1.32)	0.85	− 0.31 (− 1.84 to 1.22)	0.69
Intelligence³
Total IQ, range 45–155	0.24 (− 1.85 to 2.33)	0.82	0.52 (− 1.69 to 2.72)	0.65

Odds ratios for mortality outcomes are calculated with multivariable logistic regression analyses adjusted for age, sex, treatment center, randomization group, pediatric index of mortality (PIM3 score) at time of admission, Screening Tool for Risk on Nutritional Status and Growth (STRONGkids) risk level at time of admission, diagnostic category, history of malignancy, diabetes, and predefined syndrome. β-Estimates for neurodevelopmental outcomes are calculated with multivariable linear regression analyses performed on the 31 imputed datasets. These multivariable analyses are additionally adjusted for race, geographic origin, language, socioeconomic status (parental educational and occupational level), parental smoking before admission, left-hand preference, and (for the two columns on the right, as indicated) for upon-admission concentrations of NSE and S100B
¹For executive functioning and emotional and behavioral problems reported by parents or caregivers, data from 19 (neurodevelopmental follow-up cohort), respectively, 17 (NSE/S100B cohort) patients without hypoglycemia are missing
²Higher scores reflect worse performance
³Higher scores reflect better performance

ISSN: 1364-8535