Midazolam dose is associated with recurrence of paradoxical reactions during endoscopy

Table 1 Baseline characteristics of enrolled participants (n = 361)

	Non-recurrent group (n = 250), n (%)	Recurrent group (n = 111), n (%)	P value
Age (mean ± SD)	49.57 ± 9.22	51.15 ± 9.67	0.1395
Sex, male	165 (66)	67 (60.36)	0.3022
BMI (kg/m2) (mean ± SD)	23.74 ± 3.3	24.3 ± 3.69	0.1552
Smoking			0.8010
Never	82 (40.39)	33 (37.93)
Ex-smoker	68 (33.5)	28 (32.18)
Current smoker	53 (26.11)	26 (29.89)
Alcohol consumption			0.9465
None to minimal (< 70 g/wk)	120 (51.28)	51 (52.58)
Moderate (70-279 g/wk)	82 (35.04)	34 (35.05)
Heavy (> 280 g/wk)	32 (13.68)	12 (12.37)
Medication1	8 (3.4)	3 (3.09)	1.000
HTN (Yes)	54 (21.6)	30 (27.03)	0.2601
DM (Yes)	19 (7.6)	6 (5.41)	0.4485
Hyperlipidemia	56 (22.4)	29 (26.13)	0.4413

¹Medication: Antidepressants, anxiolytics, or hypnotics. SD: Standard deviation; BMI: Body mass index; HTN: Hypertension; DM: Diabetes mellitus.

Table 2 Univariable analysis of procedure-related factors for recurrent paradoxical reactions

	Non-recurrent group (n = 250)	Recurrent group (n = 111)	P value
Endoscope type, n (%)			0.0765
H260	157 (62.8)	59 (53.64)
H290	90 (36)	51 (46.36)
Endoscopist experience, n (%)			0.4251
Less experienced	75 (30)	41 (36.94)
Expert	106 (42.4)	43 (38.74)
Highly expert	69 (27.6)	27 (24.32)
Midazolam dose (mean ± SD)
Initial (mg)	3.98 ± 1	4.19 ± 1.16	0.1804
Initial (mg/kg)	0.06 ± 0.02	0.06 ± 0.02	0.4070
Total (mg)	5.49 ± 2.04	6.74 ± 2.58	< 0.0001
Total (mg/kg)	0.08 ± 0.04	0.1 ± 0.04	< 0.0001
Midazolam dose administered in previous endoscopy1 (mean ± SD)
Initial (mg)	3.78 ± 0.95	3.84 ± 1.04	0.7277
Initial (mg/kg)	0.06 ± 0.02	0.06 ± 0.02	0.8448
Total (mg)	7.43 ± 2.79	7.4 ± 2.37	0.6087
Total (mg/kg)	0.11 ± 0.05	0.11 ± 0.04	0.4360
Δ Midazolam dose difference (total) (mg)
Median value (quartile 25%, 75%)	-2 (-3, 0)	-1 (-2, 1)	< 0.0001

¹All participants showed paradoxical responses to midazolam in previous endoscopies. Δ Midazolam dose difference = midazolam dose (mg) - previous midazolam dose (mg). Endoscopist experience was defined as: Less experienced, < 10 years (n = 14); expert, 10-15 years (n = 9); highly expert, > 15 years (n = 7). SD: Standard deviation.

Table 3 Multivariable analysis of risk factors for recurrent paradoxical reactions in 350 participants

	OR	95%CI		P value1
Age	1.005	0.978	1.034	0.7025
Female gender (vs male)	1.521	0.731	3.164	0.262
BMI	1.072	0.992	1.158	0.0797
Smoking
Ex-smoker (vs never smoker)	1.07	0.484	2.366	0.8673
Current smoker (vs never smoker)	1.311	0.572	3.007	0.522
Alcohol
Moderate (vs none to minimal)	1.283	0.683	2.409	0.438
Heavy (vs none to minimal)	0.925	0.413	2.075	0.8504
Medication2: Yes (vs no)	1.055	0.242	4.598	0.9436
Endoscope type: H260 (vs H290)	1.405	0.854	2.312	0.1808
Endoscopist experience
Expert (vs less experienced)	0.712	0.403	1.256	0.2407
Highly expert (vs less experienced)	0.822	0.434	1.557	0.5477
Δ Midazolam dose difference (total) (mg)	1.213	1.099	1.338	0.0001

¹Adjusted for age, sex, body mass index, smoking, alcohol, medication, endoscope type, and endoscopist’s experience.

²Medication: Antidepressants, anxiolytics, or hypnotics. Δ Midazolam dose difference = midazolam dose (mg) - previous midazolam dose (mg). OR: Odds ratio; CI: Confidence interval; BMI: Body mass index.