Thus it appears that influenza frequently affects parts of the body other than the respiratory tract. Without doubt these 13 out of 98 serologically confirmed influenza cases give an erroneous indication, erring heavily on the side of excess, of the actual frequency of extrarespiratory complications since most cases of simple influenza are not hospitalized. We can only assert that of the cases of influenza among children who were hospitalized, 13 per cent have presented unexpected complications.

In some of these 13 cases the clinical diagnosis was influenced by the existence of an epidemic (six cases during the Asiatic flu pandemic). In 8 cases there was a previous history of influenza with or without viral pneumonopathy. But in two cases, one with serofibrinous pleurisy and one with reticulosis, there was neither the epidemic nor the history of an attack of influenza to account for these complications.

Thus, even during non-epidemic periods and even without histories of recent attacks, it is well to think of influenza in illnesses as diverse as jaundice, myocarditis, pericarditis, acute nephritis or the syndrome of acute lymphadenitis. Confirmation of the diagnosis, however, must rest on two simultaneous actions: first a search for and elimination of any possible secondary infection or associated disease (tuberculosis in the case of pleurisy, streptococcal infection in the case of nephritis, infectious mononucleosis in the case of reticulosis, etc.) and, secondly, a rigorous insistence on a positive complement fixation test for influenza with truly significant titers.