In this randomized blind controlled study, we found decreased incidence of PR by changing the order of midazolam and fentanyl administration. However, administration of fentanyl seems invalid to relieve when these reactions have already burst out. Additionally, we found that BIS was significantly higher when children displayed PR.
Paradoxical reactions following midazolam were reported as an incidence varies from 1.4% to 11.1% in different studies.1, 2, 22–27 This may be related to heterogeneity in terms of study design, dose and route of administration, purpose of medication, judgment criteria for PR, age, and selectivity bias.2, 14, 28 In our study, 5.98% patients occurred PR in the midazolam group and the incidence was significant reduced by preferentially and slowly injecting fentanyl. The mean time of onset of the reaction occurred at 1.8 minutes (standard deviation ±1.4 minutes) after administration of midazolam in our study. The mechanism of paradoxical reactions has not yet been elucidated. Various mechanisms proposed to explain paradoxical reactions. Our study found that BIS was significantly higher when children developed PR and it is in fact with that paradoxical
reactions occurred after an initial sedative period of following midazolam administration. As a GABA receptor agonist, midazolam has sedative and anti-epileptic effects.29 Activation of the GABAA receptor induces inward chloride currents, hyperpolarizing post-synaptic neurons, ultimately leading to pyramidal neuronal suppression and GABAA receptor-mediated inhibition is essential for inhibition of brain excitability.30 However, the midazolam-induced excitatory response seems to contradict the GABAA receptor-mediated inhibition of midazolam. Interestingly, sevoflurane and propofol can also induce excitatory response, which are epileptic spikes on the electroencephalogram during anesthesia induction.17 Similarly, excitability in the cerebral cortex appears to be conflict with the strong GABAergic effect of sevoflurane and propofol. Recently studies showed that all anesthetics that enhance GABAergic neurotransmission have both inhibitory and convulsive effects because overloaded GABAergic activity inhibits neuronal chloride channels and eventually leads to inhibition potentials being converted to excitatory potentials, which we named the double-sided effect of GABAergic neurotransmission.18 This double-sided effect may explain why midazolam causes contradictory reactions. Even more interesting is that epileptiform discharges during anesthesia induction are similarly to the PR of midazolam. On one hand, we found that PR usually occurs after a period of sedation in our study. BIS usually drops first and then rises sharply when PR occurs. This is very similar to the appearance of suppression with spike in EEG, which is outbreak of brain waves after a period of brain wave suppression.19 On the other hand, suppression with spikes is only seen during anesthesia induction, which varies from other types of epileptic discharges, such as interictal spike events, also seen in other functional brain disorder. (E.g. ADHD, migraine, Epilepsy).27, 31, 32 Finally, PR and epileptiform discharges both belong to hyperexcitability.28 Hence the mechanism of PR might be further clarified by EEG.
Our study has several limitations. First, the judgment of PR is objective and might lead to measurement bias. Fortunately, judgment of paradoxical reaction was performed by two anesthesiologists, who were engaged in pediatric anesthesia for more than 10 years. Secondly, most anesthesiologists do not induce with midazolam, however, many sites still premedication midazolam, which might also cause paradoxical reactions. In the present study, preferentially and slowly administrated fentanyl was helpful to prevent occurrence of PR, which was beneficial to patients with malignant hyperthermia or susceptibility, patients allergic to fluoride, and hospitals in developing countries.
In summary, we observed that preferentially and slowly injecting fentanyl during anesthesia induction resulted in lower incidence of paradoxical reactions following midazolam. BIS was significantly higher when children displayed PR. There were no significant differences in terms of extubation duration, recovery duration and safety endpoint between fentanyl group and midazolam group.