With constitutional diseases like syphilis, sepsis and hamophilia, we assume that damage to the vessel walls is the primary and causal factor. In children otherwise healthy, however, it is unwarranted to assume a pathological change in the vessel walls. In such cases we must, therefore, predicate an abnormal stasis in the vascular system of the gastro-intestinal tract, which leads to hyperemia and hemorrhages in the mucosa or directly to the outpouring of blood on its surface. The hyperemia and stasis can in turn be caused: (1) by pression of the cord during labor; (2) by post partum asphyxia; (3) by insufficient respiration, such as occurs with pulmonary atelectasis, (4) by congenital anomalies of the heart and vessels; (5) by cerebral and meningeal lumorrhages; (6) by local circulatory disturbances in the abdomen (e.g., with cirrhosis of the liver). It is well established that haemorrhages do occur in the various organs of asphyctic children, of children that, in consequence of some obstruction to respiration, breathe only very superficially. But it has not yet been demonstrated that cerebral haemorrhage in the newborn can actually give rise to mehuna. The interesting experiments instituted by Brown-Sequard, Schiff and Epstein and elaborated by Klosterhalfen, were carried out on young animals by v. Preuschen and Poinorski; they showed that injuries of, and haemorrhages into, various parts of the brain lead to hemorrhages in the gastro-intestinal canal. This probably depends upon a disturbance or a paralysis of the vasomotor centre, which in turn leads to Molly of the vessels, stasis and haemorrhages. Therefore-, since v. Preusehen found cerebral haemorrhage in two cases of mehena, he considers himself justified in attributing mehena, in a part of the cases, to cerebral hemorrhage. It is, however, possible that the cere bral and intestinal haemorrhages exist coordinately, both being pro duced by the same cause; this question has not yet. been clearly settled.
The Diagnostic Import of Gastro-Intestinal Haemorrhage.—In every case of mehena an effort must be made to establish its cause. Spurious melxna, is rather easily excluded; to this end careful inspection of the nose, buccal cavity and the posterior pharyngeal wall must be made in order to exclude these as the source of the blood. Examination of the maternal breast gives evidence as to fissures of the nipples. In case spurious inellena can be excluded, we must assume that the seat of the bleeding is in the gastro-intestinal tract. Then attention must be given to the signs by which hereditary syphilis can manifest itself in the newborn (see chapter on hereditary syphilis). Should there be neither clinical nor ananmestic evidence of the presence of syphilis, we must by means of the history, endeavor to exclude or establish a family and hereditary luemophilia. However, in every case, even though evidence for the existence of syphilis or haemophilia be at hand, we must endeavor to settle the question, whether sepsis may or may not be the principal causal factor. Consideration of the symptoms of sepsis, as detailed in the preceding chapter, will decide for us in such instances (in doubtful cases, the diagnosis might possibly be cleared up by means of venesec tion and the aseptic withdrawal of 1-2 c.c. of blood, for bacteriological
examination). The post-mortem examination of the heart's blood has not as much diagnostic value as the examination during life. I must, therefore, take issue with those writers who align under the category of sepsis all cases in which bacteria (especially bacillus eoli connnunis) have been found in the heart's blood after a long death agony. Even less ground exists for the opinion that all cases of mela‘na are due to sepsis.
Although the presence of the above-mentioned maladies is not pre cluded by the absence of evidence for their presence, still search must then be directed elsewhere for the cause of the lu.umorrhage. The local diseases of the abdomen and its viscera must be taken into considera tion, as must also the duration of the labor and the condition of the child post. partum. A prolonged labor might have given rise to a cere bral luemorrhage. At times much may be learned from the condition of the fontanelles; a bulging of the fontanelles with convulsions sug gests meningeal or cerebral haemorrhage. Lumbar puncture, in case there be a justifiable suspicion, may occasionally confirm the diagnosis. Asphyxia has occurred in only a small percentage of the observed cases (9 per cent. Silbermann). A careful history in reference to asphyxia. neonatorum must be elicited, since this can undoubtedly give rise to haemorrhages. A careful examination of the chest and observation of the respiration will determine atelectasis or congenital cardiac defects.
The age of the child at the time of the onset is of diagnostic value. It is highly improbable that sepsis is responsible for the liimorrhage case it begins immediately or shortly after parturition and no evidence is at hand for the existence of a congenital infection or an infection acquired during birth. On the other hand, the diagnosis of sepsis is. highly probable when the Needing starts after the fourth day of life. Haemorrhages occurring in the second week of life or later (Ritter saw such cases occurring as late as the second month) may well be regularly laid at the door of sepsis.
The presence, in the stool, of large quantities of only slightly al tered blood, is of diagnostic import, since in such cases spurious mehena is less probable than in cases in which comparatively small amounts of blood are vomited.
There still remain some cases, however, which in spite of the most thorough investigation are explained by neither the clinical examina tion nor the post-mortem findings (at the necropsy careful attention must be paid to the nose and brain and to blood infection).
The prognosis of gastro-intestinal haemorrhage varies with that of underlying disease. Silbermann estimates the average mortality to be 44 per cent.; in individual cases, where organic disease can be excluded, the prognosis becomes more favorable the less blood the child loses.