Down Syndrome Reduces the Sedative Effect of Midazolam in Pediatric Cardiovascular Surgical Patients

Abstract

Down syndrome (DS) is frequently comorbid with congenital heart disease and has recently been shown to reduce the sedative effect of benzodiazepine (BDZ)-class anesthesia but this effect in a clinical setting has not been studied. Therefore, this study compared midazolam sedation after heart surgery in DS and normal children. We retrospectively reviewed patient records in our pediatric intensive care unit (PICU) of pediatric cardiovascular operations between March 2015 and March 2018. We selected five days of continuous post-operative data just after termination of muscle relaxants. Midazolam sedation was estimated by Bayesian inference for generalized linear mixed models. We enrolled 104 patients (average age 26 weeks) of which 16 (15%) had DS. DS patients had a high probability of receiving a higher midazolam dosage and dexmedetomidine dosage over the study period (probability = 0.99, probability = 0.97) while depth of sedation was not different in DS patients (probability = 0.35). Multi regression modeling included severity scores and demographic data showed DS decreases midazolam sedation compared with controls (posterior OR = 1.32, 95% CrI = 1.01-1.75). In conclusion, midazolam dosages should be carefully adjusted as DS significantly decreases midazolam sedative effect in pediatric heart surgery patients.

Figure 1

Participant flow chart. Participant flow chart. This figure shows participant flow chart including exclusion criteria, and final enrollment patients for the investigation.

Figure 2

Relationship between operation severity and sedative dosage. The figure shows the relationship between operation risk score measured by RACHES-1 and sedative dosage. Category 1 operations tended to not differ in total amounts of sedative, but Category 2 and Category 3 had a high probability that DS patients would need a higher total amount of sedative (the probability of the DS group dosage of midazolam was greater than normal: Category 1 = 0.57, Category 2 = 0.99, Category 3 = 0.98; the probability of the DS group dosage of dexmedetoimidine was greater than normal: Category 1 = 0.84, Category 2 = 0.95, Category 3 = 0.92).

Figure 3

Figure of multiple regression model for sedation. The figure shows the main effect for model 1 in blue, the main effect (including interaction for midazolam [MDZ] and down syndrome [DS] interaction) for model 2 in green and the main effect in model 2 for dexmedetomidine (DEX) in red. An posterior odds ratio (OR) less than 1.0 indicates that a factor has a sedative effect. All the models estimate sedative effect by using 520 continuous data points from 104 patients.

Figure 4

Interaction between sedatives and Down syndrome. The figure shows the interaction of sedative effect of dexmedetomidine (DEX) and midazolam (MDZ) for normal and down syndrome (DS) patients as estimated by Bayesian inference for GLMM using No-U-turn sampler (NUTS). Shaded area indicates 95% credible intervals. (A) Describes associations between sedation status as estimated by SBS score and dose of DEX in both DS and normal patients. (B) Describes association between sedation status as estimated by SBS score and dose of MDZ in both DS and normal patients.