It has been known for a long time that purpura occurs in the last stages of tuberculosis, and "one of the earliest references to a possible association between these two conditions is found in Willan's Treatise on Cutaneous Diseases (1801). He says 'Purple spots occur in the last stage of pulmonary consumption as the immediate forerunners of death.'"1 Cruice2 in 1912 said that in the last 1,000 cases at the Adirondack Cottage Sanitarium Brown had only three with purpura and that among 1,626 cases of tuberculosis in the wards at the Henry Phipps Institute of the University of Pennsylvania in Philadelphia only eight instances of purpura were observed.
However, it should be remembered that purpura is more likely to occur in cases of miliary and generalized tuberculosis, that this class of case is rare in sanatoriums and that therefore the statistics just cited may be misleading.
WRIGHT HP, BACAL HL. ALLERGIC TUBERCULOUS PURPURA. Am J Dis Child. 1937;53(5):1276–1281. doi:10.1001/archpedi.1937.04140120100009