eLetters to:

ARTICLE: David S. Greenes and Sara A. Schutzman Clinical Indicators of Intracranial Injury in Head-injured Infants Pediatrics 1999; 104: 861-867 [Abstract] [Full text] [PDF]

eLetters: Submit a response to this article

eLetters published:

much ado about nothing

Eddie Pont   (15 October 1999)

Re: much ado about nothing

David Greenes   (20 October 1999)

An odd conclusion

Don Seidman   (23 October 1999)

Re: An odd conclusion

David Greenes   (25 October 1999)

much ado about nothing 15 October 1999
Eddie Pont, community pediatrician AAP Send letter to journal: Re: much ado about nothing edpont{at}pol.net Eddie Pont	The real meat of this study is the 431 asymptomtic patients; no one would argue with imaging a symptomatic child. Of these kids, 179 were imaged, of which 14 were found to have significant findings. Assuming the same proportion of positive radiologic findings in the non-imaged group (14/179=7.8%), an additional 20 positive studies would have occurred had all the asymptomatic kids been imaged. Since none of these kids had any disease as per the follow-up, of 34 (20+14) positive radiologic findings in 431 asymptomatic kids, only one needed any significant intervention. This yields a positive predictive value (the presence of disease given a positive test) of 2.9%. Actually I'm being generous, as in even that one positive case (an epidural hematoma requiring evacuation), the authors admit the necessity of the intervention was questionable. Nonetheless, in a startling contradiction of their own evidence, the authors conclude that we should be concerned with these radiologic phantoms anyway despite their clinical irrelevence. Struggling to find some semblence of significance to their study, they vaguely warn us of, "long-term neurologic or developmental deficits" lest we ignore the plague of asymptomatic intracranial injury. I knew there was a reason I didn't get into Harvard--it was that school bully who pushed me down in 5th grade. Give me a break!! The sad part is, now that this has been published in a respected, peer-reviewed journal, imaging asymptomatic head traumas will likely become a clinical guideline that I will be obligated to follow. I hope the ERs of Chicago are ready for the onslaught of perfectly healthy toddlers whose only crime was falling down as they were running headfirst toward their next adventure.
Re: much ado about nothing 20 October 1999
David Greenes, Pediatric Emergency Medicine Children''''s Hospital Send letter to journal: Re: Re: much ado about nothing greenes{at}a1.tch.harvard.edu David Greenes	To the editors: Attached is a response from the authors. ----------------------------------------------------------- We read with interest Dr. Pont's response to our article. We disagree with him on several points. Dr. Pont's assumption that those patients who were not imaged had the same rate of ICI as those who were imaged would suggest that clinicians' decisions about imaging were totally random. In fact, this was by no means the case, as those patients with no scalp hematoma were far less likely to be imaged than those with a scalp hematoma. Given that scalp hematoma was significantly associated with ICI, we suspect that those patients who weren't imaged likely had a much lower rate of ICI. All we can say with certainty is that 1 of the 14 cases of asymptomatic ICI that were detected resulted in an emergent surgical procedure. This represents a rate of surgical intevention of 7%, with a 95% confidence interval ranging from 2 to 32%. Dr. Pont argues that these lesions are "clinically irrelevant." We argue that the data are insufficient to support his point of view. Dr. Pont claims that we "vaguely warn" about long-term neurologic or developmental deficits. Nothing that we have written suggests in any way that we believe these lesions are associated with long-term deficits. We merely point out that the long-term prognosis after asymptomatic ICI in infants has not been studied, and that the data would be of interest. Several recent publications recommend head-imaging for all head- injured infants presenting for medical attention.(1,2) Our data show that approximately 43% of these patients may be managed safely without head imaging. We have found no published guidelines for imaging of head- injured infants that are more restrictive than our own. Our goal is not to require pediatricians to do more testing, as Dr. Pont fears, but rather to collect data, to inform our clinical practice. Dr. Pont may be right that it is not worthwhile to image asymptomatic children with head trauma. We look forward to seeing the data that prove his case. David Greenes, MD Sara Schutzman, MD REFERENCES 1. Lloyd D, Carty H, Patterson M, Butcher C, Roe D. Predictive value of skull radiography for intracranial injury in children with blunt head injury. Lancet 1997; 349: 821-4. 2. Ros S, Cetta F. Are skull radiographs useful in the evaluation of asymptomatic infants following minor head injury? Peditr Emerg Care 1992; 8:328-30.
An odd conclusion 23 October 1999
Don Seidman, pediatrician private practice Send letter to journal: Re: An odd conclusion dseidmann{at}pol.net Don Seidman	Dear Sirs: I am puzzled by the conclusions reached in this article. There were 431 children who were asymptomatic after head injury. Of them, two children recieved interventions (of questionable value) based on CT scan results. One child recieved prophylactic anticonvulsants. I do not think that one can consider this case as an outcome. One underwent surgery that the article's authors admit was of unknown need. The remaining children were all clinically well on follow up; any radiological result was of no clinical utility. If we persume that the surgery, and early diagnosis, was of real benefit to that child (one could argue that it was a deleterious result)then one can only conclude that a clinically asymptomatic child is extremely unlikely to have his/her clinical course significantly and beneficially altered by the results of a CT scan. To put this another way: a child asymptomatic after head injury will have no benefit to his/her clinical management at least 99.77% of the time. In no case was there unequivocal benefit gained by screening in the abscence of a clinical indicator. The author's reference to a previous study documenting the rare risk of significant event is disingenuous at best. The study cited had one "adverse event" - a single isolated seizure. Hardly sufficient to justify scanning all toddlers with bumped heads.
Re: An odd conclusion 25 October 1999
David Greenes, Pediatric Emergency Medicine Children''s Hospital, Boston Send letter to journal: Re: Re: An odd conclusion greenes{at}a1.tch.harvard.edu David Greenes	To the editors: We are writing in response to Dr. Seidman's letter. It is well-recognized that patients with epidural hematoma may have a "lucid interval" in which they are minimally, if at all, symptomatic, before serious neurologic deterioration sets in. In cases of large or expanding epidural hematomas, it is standard neurosurgical practice to intervene emergently, before such deterioration occurs. There are several reports in the neurosurgical literature describing criteria for nonoperative management of "asymptomatic" epidural hematomas.(1,2) All of the cases they describe involve small epidural hematomas, typically measuring less than 30 cc in estimated volume. The case we describe had a large epidural hematoma noted on head CT, with mass effect, and some "swirling " indicating active bleeding. In the OR a volume of 100 cc of clotted blood (as well as 100 cc of fresh bleeding) was evacuated. The decision to evacuate this hematoma was completely in keeping with the neurosurgical standard of care and potentially life-saving. To characterize this operation a "deleterious result" for the patient, as Dr. Seidman does, is a viewpoint utterly without basis in the literature. We also disagree with Dr. Seidman's use of 431 as the denominator for his calculations. We would recommend radiographic imaging only for 166/431 (38%) of the asymptomatic study subjects. If one performed skull x-ray as the initial study, there would be 166 skull x-rays, of which 43 (26%) would be positive. Of the 43 head CTs performed in follow-up, 14 (33%) would be positive for ICI. One of these 14 cases (7%) would have an emergent, possibly life-saving surgical procedure. A formal cost-benefit analysis would be of interest, but in our opinion, the yield of this management strategy is sufficient to justify its use. David Greenes, MD Sara Schutzman, MD REFERENCES (1) Bezircioglu H, Ersahin Y, Demircivi F et al. Nonoperative treatment of acute extradural hematomas: analysis of 80 cases. J Trauma 1996; 41:696-8. (2) Knuckley NW, Gelbard S, Epstein MH et al. The management of "asymptomatic" epidural hematomas. J Neurosurg 1989; 70:392-6.