View report page

13
small number, it would not serve any useful purpose at this stage to give a detailed
analysis of the results with regard to all the different forms of disease investigated.
Gastric lavage is now being carried out as a routine in all cases of the hilar type that
are admitted to the hospital, and figures relating to a much larger number of cases
of this class should be available by the end of next year.
The details of the technique in use at present are described in an appendix.
The children tolerate the operation well, and there seems no reason why it should
not be repeated two or three times if necessary, although this was not done in any
of the present series. The whole procedure, including guinea-pig inoculation, is now
being carried out at High Wood hospital, but in a considerable number of the earlier
cases the pathological work was carried out at the Council's Southern group
laboratory.
The following table summarises the findings ; it will be seen that the culture
was omitted in some of the cases and guinea-pig inoculation in others. At the
present time, as a special class of case is being investigated, all the material is being
inoculated into a guinea-pig except the small amount necessary to cover a slide.
In 7 cases the guinea-pigs have not yet been killed. The results of direct smear
examination in these cases are included in the table.

It will be seen that a much higher number of positive findings was obtained by
guinea-pig inoculation than either by direct examination of smears or by culture.
It must be pointed out, however, that in the combined technique a considerably
larger part of the material was injected into the guinea-pig than the amount used to
inoculate the culture medium. Despite this, in one case a positive culture was
obtained when no tuberculosis developed in the guinea-pig. Further, among the
series in which guinea-pig inoculation was omitted are to be found a relatively high
proportion of probably non-tubercular or quiescent cases. It is felt therefore that,
although the direct smear method is definitely inferior to the other two, the relative
merits of culture and guinea-pig inoculation have not yet been fully explored.
The most important question to be answered, before the value of gastric lavage
can be established as a method of diagnosis in pulmonary tuberculosis, is whether
or not the tubercle bacilli found are derived from the pulmonary and bronchial
excretions. It is, of course, impossible to exclude for certain the chance of an
accidental tubercle bacillus being swallowed and then later aspirated from the
stomach, but in the present investigation the most important source of this error,
i.e., the ingestion of tubercle bacilli in milk or milk foods, was avoided by the exclusion
of such foods from the diet for 48 hours before the lavage.* Further, organisms
derived from milk are usually of bovine type, whereas in five cases in which typing
was carried out in the present series, human type tubercle bacilli were demonstrated
in every instance. While positive findings were common among cases diagnosed
on clinical and radiological grounds as definitely suffering from active pulmonary
tuberculosis, in cases considered probably non-tubercular tubercle bacilli were not
demonstrated. This is not a result that would be expected if the accidental introduction
of tubercle bacilli from outside was a common cause of a positive gastric
lavage finding. It seems probable, therefore, that where tubercle bacilli were
obtained they originated from swallowed sputum, not from contaminated food,
tuberculous tonsils, or any of the other possible sources that adverse criticism might
suggest.
A point that is of interest in the above connection is the relationship between
the gastric lavage findings and the activity of the disease as estimated by the blood
sedimentation rate. This point requires further investigation, but, as far as the
* See Appendix.
B