586 PROCEEDINGS OF SECTION I. 



Principal Considerations in the Investigation of Typhoid 

 Outbreaks. 



The first point to be determined by the investigator of any 

 outbreak of disease is whether or not the epidemic is true tpyhoid 

 fever or some disease such as paratyphoid, resembling it. This, 

 thanks to modern bacteriological methods, is now a fairly easy 

 matter to definitely settle. In this diagnostic part of the work 

 the bacteriologist and clinician sliould collaborate whenever pos- 

 sible. Blood cultures should be made, and always they should be 

 considered the essential thing to be done, as by this means an 

 earlier diagnosis can be arrived at, that by the Widal (6). One 

 case which came under my notice showed the value of the method 

 in a striking manner. A man came into hospital with a typical 

 history of acute food poisoning. He had not been ailing before 

 and went to work, which was some type of manual labour, at 5 

 o'clock a.m. He felt well till after breakfast, when he became 

 violently ill, with purging and vomiting. Next day he was com- 

 paratively well, but being interested in the question of possible 

 bacterisemia in such cases, I made a blood culture and recovered 

 J!, typhosus, mixed with staphylococci ( ? from the skin). The 

 man had then no symptoms (except a slight temperature) even 

 suggestive of typhoid, but some three or four days later developed 

 a typical mild attack. The case, full details of which I am, I 

 regret, unable to obtain, presents a pretty problem in several ways, 

 but strongly upholds the value of blood culture. 

 • 



Besides the blood culture the pathologist has two other valu- 

 able tests — the Widal reaction and Uie leucocyte count. The 

 Widal reaction gives results inferior to the blood culture, because 

 it is often late in being developed, whereas blood culture should 

 succeed during the fii"st week. The Widal reaction is also open 

 to more errors of technique than is usually believed. When a 

 few tests are done by a person fully conversant with the method 

 and its fallacies, the test is more valuable, but when a large num- 

 ber o? tests are required, and when necessarily some of the work 

 must be done by less skilled assistants, the percentage of erroneous 

 results rapidly increases. The principal sources of error are an 

 uneven emulsion, too old culture, and erroneous dilution. The 

 often-vised method of dilution by loopfuls is, in my experience, 

 most erroneous — a dilution supposed to be 1-60 may certainly vary 

 between 1-120 or 1-30, probably even more widely. The best 

 ii'ethod of dilution is with the ordinary Wright's capillary pipette. 

 A mark is made with a blue pencil and one volume of serum is 

 sucked up, then nine volumes of saline — each volume is separated 

 by a bubble. These are mixed on a watch glass or well slide, and 

 from this 1-10 dilution in a similar way a 1-30 dilution is made. 



