More damaging information to the prosecution’s case came concerning the insulin-encrusted needle. The new evidence team gathered all the testimony and lab tests concerning the needle it could find and presented it to “several of the world’s leading forensic experts,” including the chief of serology for the New York State Medical Examiner. About the needle, Dr. Robert Shaler wrote: “The needle alleged by the State of Rhode Island to have been injected into Martha [Sunny] Bülow ... was not, and could not have been, injected.”

First, the experts said, if the needle had been injected into Sunny, there would have been traces of human tissue and blood elements on it in addition to insulin. There were none. Second, amobarbital was found on the needle, and that drug always leaves bruising and welts There were none on Sunny; physicians looked everywhere for indications of injections and found none. Third, Valium was found on the needle, but no Valium was found in Sunny’s body. Finally, the encrustations on the needle were found at the tip, which experts say is inconsistent with injection. The skin acts as a sponge and, when the needle is withdrawn, wipes the serum from the tip. The only residue would be located “at the lever fitting of the needle, that is the point of attachment of the needle to the syringe.”

Since the lab tests had shown insulin on the syringe, where had the insulin come from? The answer to that question came from across the Atlantic. The British version of the FBI had been doing research on needle washings and had found that washings containing a combination of amobarbital and Valium but no insulin can result in false positive results for insulin tests. The defense team submitted in a blind experiment four samples to the same lab that conducted the original tests. The results came back indicating two false positives for two barbiturate and Valium-laced needles and one accurate positive for a needle with Valium.

The defense team’s experts, who included the chairman of the department of medicine at the University of Chicago and the head of diabetes research for the Downstate Medical Center in New York, claimed that the lab’s speculation that the “murder weapon” contained insulin to be “not a valid result.”

“The experts thus decided that they would never rely on the ‘correct’ reading given by the lab for their own practice, yet the prosecution had relied on it in convicting Claus von Bülow and sentencing him to prison for 30 years,” Dershowitz wrote.

The last bit of science that had helped convict Claus was the expert testimony of Dr. George Cahill of Harvard Medical School, who had been asked a hypothetical question about the cause of Sunny’s two comas. Since Cahill had never seen Sunny, he could not answer questions directly about her coma, but as a recognized expert in the area of endocrinology, he could speculate about a hypothetical case similar to Sunny’s. He was asked by Famiglietti to address whether exogenous insulin could have caused coma in a woman with reactive hypoglycemia who had received two glucose pushes at 8 and 9 p.m. Cahill said yes.

The discrepancy was that Sunny von Bülow was not given glucose pushes at 8 and 9 p.m. She was given one push at 9 p.m. “The 8 p.m. glucose push had been a product of the prosecutor’s imagination. Thus, the hypothetical patient’s hypothetical coma might have been caused by an insulin injection, but Sunny von Bülow’s very real coma might well have been caused by other factors,” Dershowitz said.