Haemophilus influenzae-like bacilli in the human vagina

The factor which initiated the present investigation was the routine use of the blood agar medium developed by Professor W.M. Levinthal and used in this Unit for many years. The details of its production are given in the Appendix, but the important features in so far as HILB are concerned are:(1) the use of human blood, which is much superior to horse, ox, sheep or rabbit blood, (2) the addition of horse pancreatic digest made with the minimum possible amount of heating, (3) the use of 2% Evans peptone (with no Lab Lemco) which has also received the minimum amount of heating (4) the use of {the minimum amount of agar for adequate solidity (0.8-0.9% New Zealand, according to batch), (5) the use of thick plates (25 ml. in a 4" Petri dish) with the minimum amount of drying and (6) the addition of it small amount of penicillinase to destroy traces of penicillin which the donor may have been given before the blood was collected.Our finding concerning the superiority of Transfusion Service citrated blood over the blood of other species, contrasts with the use of sheep 271 blood by Lutz et al, (1956) and by Gardner and Dukes (1955) in their studies on similar organisms in France and the U.S.A. respectively. Gardner and Dukes went so far as to state that citrated human blood was unsuitable for solid medium. It may be that their organism, which they call Haemophilus va inalis is different from HILB, or the medium may differ in other respects, e.g., pH, ordinary laboratory media being near the upper limit for growth of HILB of around pH 7.5, and relatively small increases therefore jeopardising the growth- promoting quality of the medium. It seems likely at this stage of our experience of the organism, that some unexplained failures of growth may have been due to undetected variations in pH, although some cases were adequately explained by the demonstration of traces of penicillin in the medium.Clinically, the outstanding feature of this organism is, in our experience, its apparent ability to cause puerperal pyrexia. Attempts to arrange for inoculation of volunteers failed, owing to the natural reluctance of clinicians in charge to put at risk patients in the puerperium, so that the evidence for its human pathogenicity is limited to the marked association with pyrexia revealed by comparison with control cases, 71% being the incidence in cases of puerperal pyrexia of 99° F. 272 or over, as compared with 31 ¡, in a control group collected at the same time. The high incidence of positive cultures throughout the series, even in control groups, naturally throws some doubt on the pathogenicity of the organism and it is clear that it does not rank with,for example, Strep. pyogenes as a pathogen of the female genital tract. However, one may draw a comparison here with H. influenzae, which is a commonly occurring organism capable of both commensal and pathogenic properties.The effects of puerperal HILB infection are not confined entirely to mild degrees of pyrexia, as is seen in Table 3, where, out of 12 cases with a temperature of 100 °F. or more, 11 (92%) yielded HILB, of which 9 were in almost pure + ++ growth. None of these cases showed evidence of any other possible cause of pyrexia apart from HILB. This power to produce quite severe reactions in a minority of cases has been confirmed by experience since gained, and the unwillingness of the clinicians in charge of the patients to subject them to inoculation with HILB, indicates that they have a considerable respect for its pathogenic powers. In one case, a temperature of 103°F. has been recorded, without any other cause found than the isolation of HILB in +++ growth