Bacteriological diagnosis of tuberculous endometritis.

Ever since chemotherapy for tuberculosis became available endometrial tuberculosis has aroused increasing interest amongst gynaecologists. The diagnosis of this condition, and also the assessment of progress during therapy, depends largely on the pathological examination of curettings or biopsy specimens. The isolation and identification of the infecting organism provides definite confirmation of the diagnosis. Histology may provide strong presumptive evidence, but it cannot be as definite as positive bacteriological findings. Haines (1952) has reviewed the literature on the subject and has pointed out some of the difficulties of pathological diagnosis. The continual renewal of the endometrium allows little time for extensive lesions to develop, and the amount of tuberculous tissue in a specimen obtained by biopsy or curettage may be limited and probably unevenly distributed in the material available for examination. This report deals with the examination of 600 specimens taken by biopsy or curettage between August, 1950, and June, 1953. Of the 600 specimens, 436 were from patients in whom the diagnosis was not confirmed by the laboratory. One hundred and sixty-four were taken at various times from 42 patients in whom a diagnosis of tuberculous endometritis was established. Sixty-nine of these 164 specimens were found to be positive by one or more of the methods employed, and 95 were negative by all methods. The number of specimens examined from the cases of endometrial tuberculosis varied from one to nine. Of the 42 patients, 27 each produced one positive specimen, and in 10 of these cases it was the only specimen examined. Two positive specimens were received from each of eight patients; three positive specimens from each of four patients; four positive specimens from one patient; and five positive specimens from each of two patients. The intervals between specimens were usually about three months. Technique