DIPHTHERIA 169 



mortality was 49*9 per cent. ; of 3,723 patients bel^ween 5 

 and 10 years it was 28*1 per cent. ; of 1,330 patients 

 between 10 and 15 years it was 10*6 per cent. ; and of 

 1,972 patients between 15 and 40 years it was 4*6 per 

 cent. ; the mortality in each quinquennium between 

 15 and 40 being, roughly speaking, the same. It is thus 

 obvious that in comparing statistics of mortality the age of 

 the patients must be stated. 



' New methods of diagnosis may also lead to errors in 

 statistics. The diagnosis of the cases treated by antitoxin 

 has been verified by a bacteriological examination, while 

 in former times this plan has usually been omitted. We 

 must consider what effect this has upon the statistics. 



' A bacteriological examination enables us now to exclude 

 from our statistics many cases of angina and croup which 

 would formerly have been included. These cases are less 

 severe than cases of true diphtheria, and on this account 

 the older statistics of mortality are lower than they should 

 be. On the other hand, a bacteriological examination 

 sometimes enables us to recognise as diphtheria mild cases 

 of angina which in former days would not have been 

 included in the diphtheria statistics. I have no doubt 

 that among hospital patients, at any rate, this class of 

 cases is decidedly less frequent than the former class, con- 

 sequently the mortality of cases in which the diagnosis has 

 been verified by bacteriological examination should, ceteris 

 paribus, be higher than that of cases in which the examina- 

 tion has been omitted. 



' Another point to consider is the varying severity of the 

 epidemic. It is not common to meet with series of mild or 

 severe cases occurring at irregular intervals. The only way 

 to avoid this fallacy is to take either a large number of 

 cases in each series, or to take a large number of series for 

 comparison. 



