SERUM TREATMENT OF LOBAR PNEUMONIA 611 

 Figure 7 



type i and type ii cases with and without bac- 

 teremia treated with serum. number 

 and percentage of deaths 



Type I Type II 



Cultures 

 negative 



Cultures 

 positive 



II «% 11111 i<w* 



191 cases, 10 deaths 54 cases, 9 deaths 



lllilllll 26.2% lllllllPllIillllll S*-n 



61 cases, 16 deaths 24 cases, 13 deaths 



Note: Cases were taken from the Massachusetts Pneumonia Study. 



untreated groups. The inauguration of treatment within the first 

 four days makes it improbable that any important number of cases 

 were excluded because in extremis from pneumococcus infection. 



Certain avoidable causes of failure in the application of specific 

 therapy may be briefly reviewed. Chief among these is the delay be- 

 fore beginning treatment. Second in importance is the lapse of 

 time, previously often unavoidable, required for the determination 

 of the type of infection and now largely obviated by the use of the 

 Neufeld method. A third factor is the acceptance of other than 

 Type I and Type II pneumococci in the sputum as the etiologic 

 agent without repetition of the typing and a possible consequent 

 error of about 10 per cent. Other causes may be the use of insuffi- 

 cient antiserum, low potency, or too long intervals between doses 

 and failure to recognize the importance of administering more 

 than the usual amount of antibody in severe cases and with preg- 

 nancy, during the post-partum period, and with bacteriemia. 



Among unavoidable causes of failure, the presence of complica- 

 tions, and especially empyema, is not infrequent. Mixed infection 

 with more than one type of pneumococcus, the streptococcus, or 



