Portal de Periódicos da CAPES

Is the time overdue for an international reporting standard for convulsive paediatric status epilepticus?

2006; Wiley; Volume: 18; Issue: 1 Linguagem: Inglês

10.1111/j.1742-6723.2006.00796.x

1742-6731

Ronald A. Dieckmann,

Neonatal and fetal brain pathology

See also pp. 45–50 In this issue of Emergency Medicine Australasia, Drs Lewena and Young1 provide new ED data on convulsive paediatric status epilepticus (CPSE) – a childhood disorder still clouded by many controversies in definition and treatment. CPSE is epidemiologically one of the commonest critical paediatric presentations in both the prehospital and ED populations2, 3 and is still associated with significant neurological sequelae. Experienced clinicians recognize CPSE as a challenging condition to treat, and paediatric life support courses in North America, the UK and Australasia promulgate specific management guidelines.4-7 Despite such attention, the evidence behind our treatment strategies is scanty, and current approaches are highly variable among practitioners.8 Even the term 'status epilepticus' (SE) is ambiguous and overly inclusive, and lumps together a broad spectrum of heterogeneous seizure presentations. Only recently have clinical neurologists and paediatric consensus groups from emergency medicine and critical care endorsed a functional definition of SE to guide emergency management.9-11 This newer definition of SE as 5 min, rather than 30 min, of ongoing convulsive seizures promotes a safer and more efficacious therapeutic strategy. However, the entire science behind CPSE remains murky, and most published reports are flawed by retrospective design, selection bias, methodological inconsistencies, incomplete reporting of patient data, small sample size and benchmarking with adult SE. The Lewena and Simon report now adds several new unanswered questions to this quagmire. The authors make several significant observations in their retrospective audit of 37 young children (mean age = 3.7 years) presenting to the Royal Children's Hospital ED in Melbourne, Australia over a 3 year period. Focusing exclusively on children in active convulsive SE of at least 10 min in duration, they indicate that only 30%11 of children responded to routine 'first line' therapy with benzodiazepines, or 'second line' therapy with phenytoin and/or phenobarbitone in the ED. This magnitude of treatment failure exceeds all previous studies of CPSE. Why? A study population at a tertiary children's hospital might be biased towards serious diagnoses and higher probability of poor treatment response, but this confounding factor is unlikely to alone account for the anomalous Melbourne experience. The published data simply do not explain these extraordinary findings, which are diluted by methodological problems rife among many similar studies seeking to unravel a 'best-practice' approach to CPSE: its design is retrospective and uncontrolled, treatment is non-standardized, the sample size is small and potentially incomplete, no aetiological associations are reported, and the outcome variable is imprecise. The report appears on solid ground to declare that early treatment is strongly associated with successful termination of CPSE. Eighty 6% of Melbourne children seizing less than 20 min at the time of initial benzodiazepine administration stopped, compared with only 15% termination among children seizing greater than 30 min at the time of the first drug intervention. This is a well-accepted feature of CPSE,12, 13 and justifies immediate prehospital drug therapy with rectal or i.v. benzodiazepines for children who are seizing at the time of ambulance arrival.14 One study of prehospital treatment of CPSE not only documented excellent success rates with early benzodiazepine administration, but also reduced hospital costs – including significantly lower expenses related to intubation, ICU care and hospitalization.15 Which benzodiazepine is chosen (clonazepam, diazepam, lorazepam or midazolam), and which administration route is used (transmucosal, i.v., i.o., rectal or i.m.), appear to be far less important clinical questions than the timing of drug therapy. Early administration of a rapidly absorbed benzodiazepine, as soon as CPSE is recognized, is the key to early termination. The third concerning observation in the Melbourne study is that 70% (26) of all children, who either failed 'first' and 'second line' drug therapy or developed airway or respiratory compromise, required 'third line' therapy with phenobarbitone and RSI. Why this intervention was necessary to treat refractory seizures or respiratory depression in such a large number of patients is unclear. RSI in children with CPSE is often difficult, sometimes unnecessary and occasionally dangerous. The complication rate from the procedure itself might be significant in inexperienced hands, and resource utilization for intubated children in the ICU and hospital is high. Also, bag-mask support might be sufficient therapy during the common but often limited period of respiratory depression from benzodiazepines. Furthermore, the benefit of RSI in refractory PSE is unknown. Neuromuscular paralysis will stop the motor convulsion, but might not terminate electrical SE – a condition with neurological morbidity apparently independent of convulsive SE.16 The downside of neuromuscular paralysis is that, while stopping the motor manifestations of electrical SE, RSI also abolishes the best bedside clinical guide for further anticonvulsant treatment. Another important limitation in discerning benefits from RSI in these patients is the outcome variable in the study design. The gold standard must be neurological outcome for the child, not simply seizure termination. If long seizure times do not consistently effect ultimate neurological outcome, and if good bag-mask ventilatory support will suffice in some cases of respiratory depression, what is our rush to intervene with RSI? Studies in larger populations of CPSE patients suggest that the length of the seizure has a variable relationship to brain injury,17-19 and seizure duration alone does not portend bad brain outcome. Seizure duration becomes strongly associated with worse neurological outcome in patients who have long seizure duration and acute progressive aetiologies, such as massive traumatic brain injury, overdose of highly neurotoxic drug, or fulminant infection (meningitis/encephalitis).20, 21 Young age, less than 3 years, might also confer added injury risks. In contrast, duration of the seizure in older, chronic epileptics with static, non-progressive aetiologies might have minimal association with long-term neurological outcome, especially if seizures are for less than 1 h.22 Hence, aetiology of CPSE, not seizure duration, appears to be the primary factor in forecasting outcome. The Melbourne article reminds us of the clinical dilemmas regarding drugs and RSI in this critical condition. Who should be intubated, when, and with what pharmacological adjuncts? Available data suggest that decisions about drug treatment and timing of RSI might require a more discriminating risk/benefit approach, that first stratifies patients for treatment based on probability of neurological injury. Hence, young children with CPSE who present with serious acute underlying aetiologies probably require highly aggressive care – with maximum drug doses, repeated drug administrations, more extensive drug choices and early RSI. Alternatively, older known epileptics with chronic, static aetiologies might do well with incremental treatment and less aggressive RSI. In the end, our evidence base is grossly inadequate to attempt to answer the clinical questions raised by the Lewena and Simon study, and by our day-to-day ED practices. 'Best practice' for CPSE remains obscured by the absence of evidence. We have no reporting standard to compare current treatments and outcomes or to test newer therapies, such as valproate or propofol. CPSE, like adult convulsive status epilepticus, is a condition begging for well structured, prospective randomized clinical investigations testing old assumptions and new treatment options. The time appears ripe for an international CPSE database, comprised of agreed-upon patient and treatment characteristics (e.g. age, sex, aetiology, duration of seizure, drugs, drug doses, drug sequences, responses to treatment, RSI and short- and long-term outcomes). Association of CPSE interventions with quantifiable, neurological outcomes using reproducible scoring methodologies is essential. One decade ago, there was a similar level of confusion in our study methodologies and documentation of adult and paediatric approaches to cardiac arrest. Retrospective studies, animal data and adult data provided the 'scientific' foundations for treatment. Inability to interpret the data led to international adoption of the celebrated Utstein criteria for reporting paediatric and adult cardiac arrest information using a standardized format.23 As a result, aggregate uniform data on cardiac arrest patients, treatments and outcomes can now be better evaluated. Application of Utstein criteria to cardiac arrest studies in both children and adults has enriched the evidence-base for current approaches and for future treatment modifications. Given the many ambiguities in our understanding of CPSE, are we still confusing ourselves by comparing apples with oranges? The answers for clinicians will emerge from adoption of an international standardized reporting format, and from pooling prospective, randomized studies for collective analysis. Ideally, a consistent reporting format should be adopted for both children and adults. Then, statistically significant cohorts of CPSE patients can be more fully evaluated with the benefit of measurable neurological outcomes. And valid definitions, and critical treatments can be reviewed with confidence and adjusted to different populations with different risk levels for neurological injury. None declared.