Article Text

Download PDFPDF

“Sentinel” or “early warning” bleed
  1. N H Horwitz
  1. R J Davenport
  1. Department of Clinical Neurosciences, Western General Hospital, Edinburgh EH4 2XU, UK; rjd{at}

Statistics from

I am writing to take issue with Richard Davenport’s prejorative objection to the term “sentinel” or “early warning” bleed in patients found subsequently to have had a subarachnoid haemorrhage.1 His dismissive comment that the term was “coined by an Edinburgh neurosurgeon” is particularly offensive. The neurosurgeon in question was no journeyman but in fact was Professor F John Gillingham, an internationally esteemed vascular neurosurgeon, the protege and successor of the legendary Mr Norman Dott.

In a 1953 address to the Harvey Cushing Society (predecessor of the American Association of Neurological Surgeons) Professor Gillingham reported that over half of the patients who were admitted in coma or stupor—or were otherwise severely disabled from a ruptured intracranial aneurysm—had a clear cut history of a minor haemorrhage a few days or weeks before the major episode. Admittedly this was a retrospective survey in an era that pre-dated modern diagnostic imaging modalities and a time when there was less awareness on the part of general practitioners and emergency room personnel of the importance of new onset acute headache.

Professor Gillingham’s observation was supported by Dr W M Lougheed, a highly regarded Canadian vascular neurosurgeon. My own experience, and that of many of my neurosurgical colleagues, is consonant with the above viewpoint. Finally, I suspect that some of the 37 subjects presenting with subarachnoid haemorrhage in the prospective study alluded to may, in fact, have been experiening a minor (“sentinel”) bleed that was correctly diagnosed rather than being overlooked.


Author’s reply

I am perplexed that Dr Horwitz should have concluded that my comments regarding “sentinel” bleeds were “offensive”. Having worked in Edinburgh for over a decade, I am well aware of Professor Gillingham’s reputation. Although I doubt many other readers will have similarly misinterpreted my words, I am happy to reassure Dr Horwitz that no disrespect was remotely intended, to Professor Gillingham or anyone else.

However, I stand by my message that the term “warning leak” should be abandoned. The prospective evidence now available, and quoted in my paper, indicates that most such “warning leaks” are probably the result of recall bias from previous retrospective studies. In some cases there may indeed have been a true subarachnoid haemorrhage that was unrecognised at the time (by either the patient or their medical attendants), although the study by Linn et al1 suggests these account for a minority—in which case let us be honest and call them “missed subarachnoid haemorrhage”, rather than “warning leaks”. The key educational point is that all doctors should appreciate the potential importance of a true sudden onset headache, and refer them accordingly to the appropriate unit for further investigations.


Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.