Document J3VL3nOGz7yerZJ7myam22E2e

Uj (1 .. 0(K i \ O - IN 75 A MISTAKE received F3 9 `5 v `"".Y JAN i u Apparently Dramatic Trend| in First 7 Months Laid to the Vagaries of Data l By HAROLD M. SCHMECK JR. ' Sprntu) to Hi* New YnflrtlDK* WASHINGTON, Jani 28--The .' apparent dramatic increase in i the nation's cancer death rate ` is the first seven months of . 1975 resulted from vagaries of ; data and does not: reflect a ` real disease trend. Government ; experts said today. . An expert from the National f Cancer Institute described it tot. day as ``the increase that wasn't ' there." S1""statistician of the i TTational Center for Health Sta tistics said a flu epidemic last ; January and February may have been a key factor in the seemf, ing increase in the cancer death rate. A. rise in flu customarily ' brings increases in deaths from " chronic diseases such as heart disease and cancer presumably r because flu weakenj persona * nearing death from their under- lying illnesses. Provisional data for the first * seven months of list year ' showed a 5.2 percent rise in . cancer death rate over the com parable period of 1974. For .'' years the rate has risin 1 per cent a year. The 5.2 percent increase was published, without comment, last November in a routine vital statistics report of the National Center for Health Statistics. The startling size of the ap parent increase immediately brought much publicity to the figure amid speculation as to its causes. The National Cancer Insti. tute, main Federal agency for cancer research, received more than 50 inquiries from Congress in a week. Some persons sug gested that the increase might be a result of some unknown man-made pollution of the en vironment several decades ago. Experts at the cancer institute began a painstaking search of the available data in coopera tion with the statistics center. Imperfections and Vagaries Today, at a lecture for cancer institute staff members, Dr. Leonard Chiazze Jr., chief of the institute's biometry branch, 'said the apparent rise seems . to have resulted from imperfec' tions and vagaries of the data. Dr. Chiazze, a bio-statisti cian, said the provisional data for the seven months had in deed shown an over-ail cancer death rate of 176.3 deaths for each 100,000 persons in the population and that this was a 5.2 percent increase,. over the previous year. '. Provisional data -for thk foist nine months of the .year, however, brought the rise over 1974 down to 2.8 percent, he said, and compensating for some vagaries of data would lower this still furthr to 2.3 percent 1 What the real rate will be will not be known until the final data for 1975 become available late this year. .The scientists said provisional 'data were far less reliable than final data and the over-all cancer death rate was far less mea ningful than rates subdivided according to age, sex, race and particular type of cancer. The trend of the last seven years has been a roughly 1 percent a year increase in can cer deaths, with the growing number of older people in the United States evidently a major contributing factor. Factor of Flu Outbreak At the meeting today, Robert Armstrong, chief of the Morta lity Statistics Branch of the National Center for Health Sta tistics, said he believed the flu outbreak last winter might have contributed greatly to the illusory cancer trend. He said there was no comparable flu in 1974 and that this would enhance the seeming difference between the two years. , As an aside, he said fins over-all United States mortality figures for 1974 would be made public next week and would show that year to have had the lowest over-all death ratq in the 75 years of record keep* ing. Deaths from automobile ac cidents were down sharply from 55,000 in 1973 to about 45,000 in 1974, perhaps because of the gasoline shortage. Furth ermore, there was little flu in 1974. Mr. Armstrong and Dr< Chiazze noted that the monthly provisional figures, which caused so much excitement concerning cancer, were neithei suitable as, nor intended a$: guides to trends in cancer or! any other chronic disease, Tfc^ figures are published to giv$ early warning of major acuta disease outbreaks. Indeed, the irony of the pro*visional cancer data is that the figures for the first three months of 1975 were in even more startling contrast to 197/f than were the seven-month data. Provisional figures for thi January-Febniary-March period showed a 6.2 percent rise over 1974, Mr. Armstrong said. When April was added, the disparity rose to 6.7 percent By thfe time nonstatisticians first no*) ticed the trend--after seven months--it was already well Oft the way down, Ycrk T\ wi e-S Tiny rs ^*7 ~Sojni4(^ry f> 17 BOR 009977