DIAGNOSIS OF TUBERCULOSIS 489 
Cutaneous reaction, 6571 cases: 
2192 tuberoiiloiis patients (clinical) 1851 positive reaction 
865 suspected cases (clinical) . . 563 positive reaction 
3514 clinically non-tuberculous 1047 positive reaction 
Conjunctival reaction, 6788 cases: 
2834 tuberculous patients (clinical) 2370 positive reaction 
1188 suspected cases (clinical) 685 positive reaction 
2766 clinically non-tuberculous 407 positive reaction 
Summary. 
Tuberculosis 
Suspicious 
Non-tuberculosis 
per cent. 
per cent. 
per cent. 
Subcutaneous reaction . . . . 89. 2 -j- 
65.5 + 
51.2 + 
Conjunctival reaction 83 . 6 -|- 
56.6 + 
14.7 + 
Cutaneous reaction 84 . 4 -|- 
65.0 + 
29.8 + 
Cutaneous reaction (excluding chil- 
dren) 
33.9 + 
The tuberculin test, therefore, when positive gives no absohitely 
definite distinction between healed, latent, or active foci of infection 
with the tubercle bacillus. Furthermore, no quantitative evidence 
of the nature of the reaction or extent of the lesions is elicited by a 
])ositive reaction. A negative reaction properly performed, however, 
in non-cachectic subjects or those who have not had progressive 
treatment with tuberculin is fairly conclusive. The subcutaneous 
test of Koch is moderately reliable in adults provided the dosage is 
correctly selected ~a matter requiring unusual skill and experience. 
The von Pirquet reaction is quite unreliable as an index of an active 
focus in adults, because from 30 to 60 per cent of all persons over 
five years of age react positively in varying degrees. It is a fairly con- 
clusive test in children under five years of age. Similar fallacies are 
to be expected in the ophthahiio and Moro reactions.^ 
B. Serological Diagnosis. — 1 . Opsonic Index. —Wright and his pupils 
have followed the opsonic index in tuberculous patients and they 
believe that the changes in opsonic index furnish a reliable index 
for treatment with tuberculin or other products of the tubercle bacillus. 
The inherent and una^'oidable errors of the opsonic index determina- 
tion make it unreliable for general use.^ 
2. Agglutination.— Agglutinins occur in tuberculous patients,'* but 
the agglutinating reaction is unreliable, partly because of the diffi- 
culty in obtaining a proper suspension of tubercle bacilli. It is 
practically never used in practice.^ 
3. Complement-fixation in Tuberculosis. — Complement-fixation has 
been studied in tuberculosis for its value as a diagnostic procedure, to 
' For a detailed summary of the Koch, Wassermann, von Pirquet and Wolff-Eisner 
theories of the tuberculin reaction and its importance for therapy, see Kuthy and Wolff- 
Eisner (Die Prognosenstellung bei den Lungentuberkulose, Berlin and Vienna, 1914, 
393-399). Also, Wolff-Eisner: Die Tuberkulinbehandlung, 1913. 
2 Trudeau: Am. Jour. Med. Sci., 1907, 133, 813. Baldwin: New York Med. Jour., 
1908, 87, 1227. 
' Romberg: Deutsch. med. Wchnschr., 1901, 27, 273, 292. 
* Eisenberg and Keller: Centralbl. f. Bakteriol., orig., 1903, 33, 549. 
