Antall József szerk.: Orvostörténeti közlemények 55-56. (Budapest, 1970)

TANULMÁNYOK - Mádai Lajos: Semmelweis és a statisztikai tudomány (angol nyelvű közlemény)

condemned those who applied "concealing diagnoses" in order to maintain the good reputation of their hospitals by glossing over the real number of illnesses and deaths due to puerperal fever. The validity of Spaeth's criticism must be objectively considered, and it can be rightly applied to such physicians as Seyfert, professor of obstetrics in Prague, one of the opponents of Semmelweis, whose maternity statistics on the years 1853—1860 were indeed unscholarly as proved by the facts disclosed by the Czech medical historian Boucek, and referred to in details in the work of Benedek. (15) But Spaeth's general criticism cannot be applied to the data of the maternity hospitals in England and Ireland and to claim that they were unreliable and palliated. Why? 1. The English and Irish maternity hospitals were not in need of palliation; due to reasons set forth by Semmelweis in detail their mortality rate was considerably more favourable through many decades than that of the maternity departments in most European hospitals. 2. The series of decades treated by Semmelweis and the great majority of the English and Irish statistical figures, referred to the period preceding the discovery of the cause of puerperal fever (1750—1847), remote from the years of the later debate when the possibility of subjective distortions cannot be excluded. 3. The hint of unreliability cannot be ascribed to the statistics of the English and Irish maternity hospitals because it would have meant the consecutive and consistent palliation of the figures in various hospitals by several generations of physicians. The linear trendline which excludes the consequences of possible glossing and occasional fluctuations also shows the reliability of the mortality rates provided by the London and Dublin hospitals. Semmelweis selected and evaluated the international hospital statistical figures with a sound critical sense, for instance he learned English to be able to examine more profoundly the causes of the more favourable mortality rates during a planned study-tour in England and Ireland. Semmelweis's critical approach is well shown by the fact that he exposed the elemental errors of the Würzburg professor of obstetrics, Kivisch, the unscientific comparison of ratios based on an insufficient number of figures, and similarly he revealed the statistical mistakes committed by professors Lampe, Scanzoni, and Károly Braun (10). Finally mention must be made of Semmelweis's methods of calculation. He worked very accurately with the vast collection of raw figures, computed the mortality rate to two places of decimals, and used intensity ratios as well. The precision of the latters, the accuracy of rates expressed in mixed fractions (e.g. in 1848 one childbed death fell to 79—; women in child-bed) was pointed out by Regöly-Mérei in his article (17). The few flaws — correctly pointed out by Benedek in his book mentioned — do not affect the validity of the essential regularities, and further, I found that they were not mistakes in calculation but misprints, something even today's modern typography is not free of.