Reports and notes of the Public Health Laboratories, Cairo. No. 1, 1917.

Date:
1918
Catalogue details

Licence: In copyright

Credit: Reports and notes of the Public Health Laboratories, Cairo. No. 1, 1917. Source: Wellcome Collection.

    27/144 (page 15)
    a reliable account in cases which had been moved to Cairo from other areas, but in 817 cases a fairly reliable report could be given ; the results so obtained are analysed in Table II. TABLE II. Blood-Cultures. Total. Positi ¢. . Spee PTE CASCS ers teins hii. eee tse) Ly LOO 226 19 irataweakae tis J1re7 ie ee 199 65 33 SeCOlGEWEe Kune woe haart 305 60 la Third week : 5 eee seccenial 145 13 9 Later than third week Ret sae 3532 118 8 7 Winknowne- ae Riscal iret cahedis ns 371 80 — The results show that the probability of isolating the causative organism from the blood in the second week is-only half that in the first week, and in the third week again only half that in the second week. It also shows that even in the fourth week in a certain proportion of cases the bacilli can still be obtained from the blood. In view of the comparative simplicity of taking a blood-culture and of dealing with it, a case should probably never be regarded as of too long standing to make this form of diagnosis worth a trial. There remains for consideration the question of the actual value of a blood-culture in the diagnosis of a case of suspected enterica. The figures obtained and set forth in Table II cannot as a matter of fact be regarded as furnishing a definite elucidation of this point, as they themselves only show the percentage of positive results obtained in a number of suspected cases, while the datum actually required to answer the question is the proportion of positive results to be obtained from a number of cases not of suspected enterica but of definitely, 2.e. bacteriologically-established typhoid or paratyphoid fever. The cases in this series were undoubtedly not all cases of enterica, but it is impossible to ascertain the number to be excluded on this basis. At first, however, it appeared that the point might be readily elucidated by selecting from the series those cases in which a blood-culture had been performed and which had been proved positive on other grounds, viz. by the isolation of the causal organism from the feeces or urine ; the proportion of positive and negative blood-cultures from these cases appeared capable of giving a definite indication of the probability of obtaining a positive result in a case of actual enterica. The cases, amounting to 116, were accordingly separated and analysed, but with a result which showed that they were not capable of employment for this purpose ; in fact, the proportion of positive blood-cultures among these cases, proved positive on other grounds, was less than that among all the cases taken together. The detrimental factor at work is undoubtedly that referred to above, viz. that the majority of the cases were examined with a view to diagnosis only and that in many instances a positive blood-culture in itself inhibited the further investigation of the case ; hence, by selecting any series of cases in which the urine or feeces were also examined, we are thereby exercising a bias in favour of negative blood-cultures. Evidently, therefore, this method of determining the proportion of positive blood-cultures in positive cases is fallacious, and it is necessary to evolve some other process if we are to make use of the investigation for this purpose. Now, we only wish to find the proportion of positive blood-cultures in positive cases, and therefore we may be permitted to make use of those methods of approximation which the ordinary laws of chance provide us with in the evolution of such a process. The question is, assuming that when we select a series of cases in which specimens of blood, feeces, and urine have been examined we leave behind a certain number of cases in which the blood: only has been examined and found positive,