A Review of 1,000 Cases of Asthma

rMED-A3O 965 2. But with 'the " control families " there is in the M.R.C. figures a large element of error, for the so-called i' control families " are by no -means " nio-n-rheeumatic families."They are definedl as follows (p.9): "These families were selected only in the respect that each contained a child attending the hospital for a non-rheumatic disease, and will be termed ' control families.'Whleni the figures are examined it is -found (Table 8, p. 16) that in the 200 -" control families " there were 43 cases of rheumatism, apparently spread over (p.18) pt least 36 of the families.That these were nnt unimpor- tanit instances of previous rheumatism occurring in parents is shown by the following passage (p.18):" The difference in the incidence of rheumatism among the brothers and sisters of the rheumatic and control children respec- tively is striking, the percentage of those with rheumatism in the rheumatic families in the data from St. Thomas's Hospital being almost three times greater than that of the control families....