A study of tuberculous disease in infancy and childhood, with particular reference to the primary sites of infection

Abstract

The present investigation is based on a survey of 2,500 consecutive autopsies on children, in whom a thorough search for evidence of tuberculous disease was made. The relative incidence of tuberculous lesions to the number of autopsies was less in children in the first three years of life than in children between three and thirteen, though it was in the third year that the peak incidence occurred. Girls at all ages were more often affected than boys. The yearly incidence of positive tuberculin tests in a large number of children who came to autopsy practically corresponded with the incidence of tuberculous lesions found. The tuberculous lesions were the cause of death in 240 (95·2 per cent.) of 252 subjects under three and in 93 (74·4 per cent.) of 125 over this age. In 5 cases there were double primary sites of infection, 1 in thorax (responsible for death) and 1 in the abdomen. The primary site of infection was most commonly thoracic, and then, in order of frequency, abdominal, cervical, or unknown. Primary thoracic tuberculosis occurred in 63·4 per cent. of the tuberculous cases, and relative to the number of autopsies had a higher incidence in subjects over three years, but was more fatal under this age. Only 0·6 per cent. of the cases under three and 9·0 per cent. over this age with primary thoracic tuberculous lesions did not die as a result of the tuberculous infection. Support for this finding is derived from (1) observations on adult autopsies, (2) tuberculin tests in children, (3) clinical findings and results of stomach washings. Primary lung lesions of more acute progressive type were found in children under three years than in those over, while evidence of arrested lesions was commoner in the older children. The primary lung lesion was always associated with tuberculous tracheo-bronchial glands and was very rarely the sole tuberculous lesion in the lungs. Most of the children with primary intrathoracic tuberculosis died of generalised infection, and this was slightly commoner in children under three than in those over. In 149 children who died, the infection was due to human strains in 143 and bovine in 6 (4 per cent.). There was a slightly higher percentage of bovine strains in children in the first three years of life than in those between three and thirteen. From children not dying as a result of the tuberculous lesion, 5 human strains were isolated. The prognosis is grave in cases of primary intrathoracic tuberculosis, and for a diagnosis it is necessary to correlate the clinical and radiological findings, the family history, result of tuberculin tests, and examination of sputum or stomach washings. A wider general application of the tuberculin test in young children is recommended in order to detect any positive reactors. In 123 (32·4 per cent.) investigated, the primary site of the tuberculous infection was abdominal. The relative incidence of primary abdominal lesions to the total necropsies was greater in children over than in those under three years, but more deaths resulted from the abdominal infection in children under three. Intestinal ulceration was present in 22 abdominal cases; such lesions were always fatal and were commoner under three years. In 101 instances there was no tuberculous bowel lesion, but the mesenteric glands were caseous. Evidence of healing in these glands was commoner in children over three years. In 74·4 per cent. of children under three and 29·3 per cent. of those over this age generalisation of the infection caused death. Tuberculous disease localised to the abdomen caused more deaths than similar disease confined to the lungs. Of the bacilli isolated from the abdominal cases in which death resulted from tuberculosis, 11 were human strains and 47 (81·0 per cent.) bovine. From children not dying as a result of the abdominal disease, 2 human and 13 bovine (86·7 per cent.) strains were obtained. The prognosis in primary abdominal tuberculosis is better than that in primary thoracic tuberculosis, but the diagnosis of the abdominal disease in the early stages is often attended with much difficulty. Primary tuberculosis of the cervical glands was not often found in this autopsy series, occurring in only 8 (2·1 per cent.) cases, of which 2 did not die as a result of the lesions, the remainder dying of generalisation of the disease. From this group 3 human and 3 bovine strains were isolated. In 7 (2·1 per cent.) subjects no primary source was found, and from these 3 human strains were obtained.