Sir.—Hageman et al,1 in their study of persistent pulmonary hypertension of the newborn (PPHN), stated that improved survival in patients with this disorder was related to earlier, more frequent diagnosis and more rapid initiation of therapy. Unfortunately, the data they presented did not support this conclusion.
Among 6,384 inborn live births in 1980 there were eight cases of PPHN and four deaths. With a "higher index of suspicion" among the 6,642 live births in 1981 (4% more than in 1980) there were 27 cases of PPHN and three deaths. By changing the level of suspicion, the authors have altered their standards for admission into the study and increased the number of children at risk for development of the disease by more than threefold. They transformed a one-death difference by this statistical alchemy into a statistically significant event with a P value of less than.02. Using these at-risk values,