Ernest J. Sternglass - Three Mile Island

Three Mile Island

In April 1979, Sternglass was invited to testify to Congressional hearings on the Three Mile Island accident. Two days later, when the hearings were moved from the House to the Senate, he was told his testimony was no longer desired. Sternglass believed that an effort was being made to suppress any evidence about possible deaths as a result of the accident. In a paper presented at an engineering and architecture congress, Sternglass argued that an excess of 430 infant deaths in the U.S. Northeast that summer could largely be attributed to Three Mile Island radiation releases. This led some writers on environmental issues to claim that he had proven that figure as a minimum.

Sternglass's methodology was criticized—including by the medical researcher who provided him with the statistics (Gordon MacLeod), and by an otherwise-sympathetic researcher with the Natural Resources Defense Council (Arthur Tamplin) -- on several counts:

• for not attaining statistical significance (Frank Greenberg, Centers for Disease Control and Prevention, CDC);
• for lacking a sufficient baseline, since screening for hypothyroidism hadn't started until 1978 (Greenberg)
• for not looking at the number of babies who didn't die (Gary Stein, CDC);
• for not noticing that the sex ratio of newborns hadn't changed—males being more susceptible to fetal injury than females (Stein; George Tokuhata, Pennsylvania Health Department, director of epidemiology),
• for "ignoring close to the reactor, where the infant mortality was very low" (Tokuhata);
• for simply being incomplete (Tamplin).

As well, he had relied on figures that had incorrectly compounded fetal deaths with infant mortality (Tokuhata).