WE ARE THANKFUL TO Dr Fernandez-Ortega and colleagues1 for the interest in our work, as they have acknowledged the need for further studies involving pediatric patients and with a long-term rehabilitation perspective.
First of all, we agree with the authors of the letter that all future reports of possible paroxysmal sympathetic hyperactivity (PSH) episodes should be made according to the consensus paper2 in order to avoid dispersion of useful data or, on the contrary, inclusion of phenomena that are not PSH episodes.
Limited to the case of our work,3 unfortunately, the peer review was finalized before the consensus paper was published. Still, a practical limit to the application of the diagnostic guidelines would have been that the consensus paper did not yet formulate scores for pediatric life signs. Therefore, in light of the consensus paper, we later proposed the reference values and criteria that we used.4 The diagnoses of PSH made for this article were based on those criteria and, as we are going to explain, were largely consistent with the method of the consensus, concerning both the Clinical Feature Scale (CFS) and the Diagnosis Likelihood Tool (DLT).
Analyzing in detail the CFS, the episodes we considered to be PSH scored at least 2 to 5 points for the life signs and at least 2 points for sweating/posturing. We can infer that all included episodes had a minimum CFS score of 4 whereas those with lower scores were excluded. The only discrepancy is that we used, as a separate parameter, the diastolic blood pressure, which was excluded from the consensus work. It must be noted, however, that all patients were also taking neurologically relevant drugs that may have blunted symptoms; therefore, the CFS score we observed may be underestimated. Indeed, work is in progress regarding the possible influence on PSH of several drugs administered on a regular basis and for other neurological indications.
Concerning the DLT, all the episodes that we considered to be PSH were, by definition, compliant to items 1, 2, 3, and 11. Features persisted more than 3 days (item 4); indeed, we chose a minimum of 7 days. Features lasted more than 2 weeks after injury (item 5) since all patients were undergoing rehabilitation. Features always persisted despite the treatment of other conditions (item 6); in fact, patients with unclear manifestations were excluded and included patients were evaluated only when no other medical problem or parasympathetic feature was recognizable (items 9 and 10). Although the need of medications was not included in our definition of PSH, we evaluated only patients who were discharged from the intensive care unit with a preset therapy that also aimed to mitigate PSH, thereby complying with item 7. The only discrepancy between the cohort we analyzed and the consensus definitions is that we included patients with a single episode per day but who required a longer duration of stay (see item 4). Overall, we can confirm that included episodes scored at least 9 points on the DLT. Adding all scores together, the sample we described places itself around 13 points; therefore, in the top half of the "possible" category of PSH likelihood. Moreover, on the basis of our higher cutoff on symptoms, it may be that our criteria operated on a more strict selection; indeed, we excluded 12 of the possible 38 cases because of diagnostic uncertainty.
Regarding the assessment of drug efficacy, we agree with Dr Fernandez-Ortega and colleagues on the intrinsic limitations of observational studies. Yet, no previous work performed an evaluation of acute drug efficacy in the context of PSH, specifying drug doses and odds ratios; thus, we felt that our observational analysis was a significant addition to what is already known.
The time limit of half an hour, during which we chose to evaluate the resolution of PSH episodes, had 2 rationales: first, it was a reasonable compromise comprising the Tmax (time after drug administration when the peak plasmatic concentration is reached) of most investigated drugs; second, it takes into account the fact that PSH episodes may self-resolve, an aspect highlighted by Dr Fernandez-Ortega and colleagues, which we agree upon. Had we chosen an excessively long time frame, the chance of spontaneous resolution would have indeed increased over time, adding confusion to the results. Furthermore, the fact that most drugs did not provide results on PSH during this time frame strengthens the significance of the fewer positive results we observed. Concerning the therapeutic alternatives that have been proposed previously, we would like to point out that they were not aimed at acute care treatment, rather at the stabilization of patients, until PSH episodes subsided or discharge was possible. What we are now proposing is a panel of therapies to be administered only when needed. For instance, it would not be feasible to conduct a therapy on an as-needed basis using baclofen, gabapentin, or bromocriptine, all requiring accurate dose titration. Conversely, this is feasible with hydroxyzine and benzodiazepines, which are already used on an as-needed basis for neuropsychiatric and pediatric neurological indications. Regarding opioids, we agree with Dr Fernandez-Ortega and colleagues on their high value in intensive care contexts, but it should be considered that in cases in which patients are discharged from the intensive care unit to the rehabilitation unit with unresolved PSH episodes, the opioid therapy may represent an additional risk factor and may complicate patient management, especially in the pediatric age. In a more conservative future perspective, we are convinced that therapies studied in rehabilitation should be initiated during intensive care treatment to minimize possible issues after the transfer and prevent emergency interventions due to uncontrolled PSH.
We also agree with the authors that more studies on the cause of PSH are required to clarify anatomical and functional correlates, although how to perform imaging during PSH episodes remains to be defined. Without sedation, the examination would be hardly feasible whereas the use of sedatives could generate artifacts and different sedation protocols may lead to different results.
Overall, we believe that despite the intrinsic limitations, our study add significantly to the current scenario, in which there is still a lack of appropriate bases on which to design a randomized trial. It would also be ethically questionable to propose a randomized study with treatments not supported by prior observations, or to include untreated/placebo control groups, in a delicate condition such as PSH. In this view, we hope our results may provide a basis on which to develop future systematic studies for the therapy of PSH on an as-needed basis.
-Marco Pozzi, PhD
Federica Locatelli, MD
Sara Galbiati, MD
Scientific Institute IRCCS Eugenio Medea
Lecco, Italy
Emilio Clementi, MD
Scientific Institute IRCCS Eugenio Medea
Lecco, Italy
Unit of Clinical Pharmacology
CNR Institute of Neuroscience
Department of Biomedical and Clinical Sciences L. Sacco
"Luigi Sacco" University Hospital, Universita di Milano
Milan, Italy
Sandra Strazzer, MD
Scientific Institute IRCCS Eugenio Medea
Lecco, Italy
REFERENCES