“Whatever mother says!” was the title of a brief 1985 editorial written by Dr. E. Bruce Hendrick.1 In it, he eschewed the emerging trend toward reliance on imaging technology, which he warned might come at the expense of the art of medicine; in particular, carefully listening to patients and their families. In all branches of pediatric medicine, we heavily rely on parental input, not just for objective accounts of symptoms, but often to help put the symptoms in the appropriate context for the child and his or her history. But when parents are asked to take on this role—of interpreting symptoms, not just reporting them—how often are they correct? In their thoughtful research paper, Naftel et al.3 provide us with some scientific evidence behind this issue as it relates to shunt failure in children. The results, although mitigated by some important caveats, suggest that parental interpretation of symptoms related to shunt failure is not particularly good. In essence, when participating parents in this study believed their child was not in shunt failure, they were usually correct, but if they thought their child was experiencing shunt failure, they were often wrong. This paper has several hallmarks of good research: the primary question is simple and something neurosurgeons ask daily, the prospective methodology is appropriately applied, the analysis is clear and simple, and their conclusions are reasonable without overreaching. For this, the authors should be congratulated. Behind the headlines, there are some other interesting results. For example, we are shown prospective data that the diagnostic accuracy of CT and shunt series, although generally better than parental assessment, is only approximately 85%. These imaging modalities will miss a reasonable minority of children in true shunt failure. While this phenomenon is well recognized by pediatric neurosurgeons, it is a lesson that should be emphasized widely to emergency and primary care physicians who, on occasion, might take too great comfort in the “unchanged CT scan” report. Although mostly mentioned by the authors in their discussion, the caveats to this paper are worth highlighting because they affect how we should interpret and use their results. First, within the study, there were other motives for parents to claim that symptoms were shuntrelated (such as speed of receiving medical attention), regardless of what they honestly believed. So, this paper might be overestimating the number of parents who truly believed their child had a shunt problem, which would at least partially explain their low positive predictive value. Second, the parents were forced into a dichotomy: having to choose “yes” or “no” to a shunt problem, leaving no room to assess their degree of uncertainty. This does not translate directly into how parental opinion is typically used by the neurosurgeon in assessing the probability of a shunt problem. More often, we might ask how closely the current symptoms resemble previous episodes of shunt malfunction, acknowledging that parents, like us, are not always certain. Third, it is not clear how one should interpret their finding that observations by “experienced” parents had a lower positive predictive value than those of less experienced parents, a finding at odds with previous work by Kim et al.2 As the authors state, this might relate to a sensitization to shunt problems or experience in better negotiating the medical triage system to expedite care. The latter is perhaps more explanatory, since a key difference in the study of Kim et al. is that parental opinion about the shunt was assessed by the emergency physician (once medical care was already being received) rather than by the triage nurse, so there would be less reason for parents to overstate their concerns of shunt failure. Ultimately, how should we use the results of Naftel et al.? Neurosurgeons can perhaps take some comfort in how consistently parents are able to rule out shunt failure. In their conclusions, however, the authors harken back to the principle of medicine as art and suggest that clinicians need to weigh all clinical and imaging factors, recognizing that none, individually, are absolute in their diagnostic accuracy. In doing so, the authors remind us that the essence of Dr. Hendrick’s original warning1 remains as relevant today as it was nearly 3 decades ago. (http://thejns.org/doi/abs/10.3171/2011.11.PEDS11462) 361 362
Search from 250 M+ research papersSearch from 250 M+ research papers
Jun 7, 2023
Jia Xu Lim + 7
Open Access