Increased Use of Hyperosmolar Therapy for Suspected Clinically Apparent Brain Injury in Pediatric Patients with Diabetic Ketoacidosis during the Peak of the COVID-19 Pandemic

Table 2

Comparison of biochemical characteristics of patients presenting with suspected clinically apparent brain injury (CABI) before the COVID-19 pandemic, patients presenting with suspected CABI during the pandemic, and cases presenting with diabetic ketoacidosis but without suspected CABI during the pandemic.

Characteristics	Group			Test statistic	value
	Suspected CABI before pandemic (group 1, n = 18)	Suspected CABI during pandemic (group 2, n = 18)	DKA without suspected CABI during pandemic (group 3, n = 141)
Biochemical variables ^† (mean ± SD^‡)
pH	6.89 ± 0.09	6.91 ± 0.13	7.14 ± 0.11	H = 61.032
pCO2 (mmHg)	24.0 ± 6.0	21.1 ± 7.5	28.9 ± 7.5	F = 10.871	<0.001
Bicarbonate (mmol/L)	4.6 ± 1.7	4.4 ± 2.5	9.1 ± 3.4	H = 50.885	<0.001
BOHB (mmol/L)	9.4 ± 3.6	9.0 ± 4.0	7.3 ± 2.3	H = 4.419	0.110
Measured sodium (mmol/L)	134 ± 7	133 ± 5	133 ± 5	H = 1.667	0.43
Corrected sodium (mmol/L)	144 ± 7	142 ± 6	140 ± 6	H = 6.068	0.048
Potassium (mmol/L)	4.6 ± 1.1	4.4 ± 1.3	4.5 ± 0.8	H = 2.335	0.31
Anion gap (mmol/L)	33.3 ± 6.7	31.3 ± 5.4	28.1 ± 4.8	H = 13.333	<0.001
Measured osmolality (mOsm/kg)	356 ± 27	335 ± 22	326 ± 19	H = 9.429	0.009
Effective osmolality (mOsm/kg)	307 ± 17	302 ± 18	295 ± 13	H = 10.637	0.005
Glucose (mg/dL)	695 ± 275	658 ± 311	541 ± 217	H = 13.225	0.001
BUN (mg/dL)	19 ± 10	23 ± 10	16 ± 6	H = 12.381	0.002
Creatinine (mg/dL)	1.11 ± 0.45	0.90 ± 0.53	0.80 ± 0.38	H = 8.117	0.017
BUN/creatinine ratio	18.9 ± 9.2	28.5 ± 9.4	23.7 ± 15.1	H = 8.891	0.012
Initial HbA1c (%) (mmol/mol)	12.1 ± 1.8 (108.3 ± 19.6)	11.9 ± 2.2 (106.7 ± 23.9)	12.0 ± 2.3 (107.5 ± 24.7)	H = 0.086	0.96
Future HbA1c (%) (mmol/mol)	8.7 ± 2.1 (71.8 ± 23.3)	7.2 ± 0.7 (54.7 ± 7.4)	9.0 ± 3.0 (73.5 ± 37.2)	H = 4.550	0.10
HbA1c change from baseline^§	−3.3 ± 2.1	−3.9 ± 2.0	−3.0 ± 3.6	F = 2.263^¶	0.11

^†On presentation, unless otherwise indicated. ^‡Although means and standard deviations were reported for all variables, nonparametric testing was performed for comparison of mean ranks in the cases where data in at least one group were not normally distributed and and/or there were unequal variances across the groups. ^§Adjusted for baseline HbA1c, interval between the two HbA1c measurements (in days), and new versus established diabetes status. ^¶Due to unequal variances across groups, parameter estimates with robust standard error were calculated, with the following results: group 1–group 3: t = 1.063 and  = 0.290; group 2-group 3: t = −2.616 and  = 0.010. On post hoc analysis, group 3 was significantly different from groups 1 and 2. On post hoc analysis, group 3 was significantly different from group 1 but not group 2. On post hoc analysis, group 3 was significantly different from group 2 but not group 1. On post hoc analysis, group 2 was significantly different from groups 1 and 3. BOHB, beta-hydroxybutyrate; CABI, clinically apparent brain injury; DKA, diabetic ketoacidosis; SD, standard deviation; HbA1c, hemoglobin A1c.