BRATTLEBORO -- A potentially fatal dose of insulin could be delivered quickly and in a "virtually painless" manner, an expert in diabetes treatment declared Monday.

That's just the type of injection that prosecutors claim Jodi LaClaire gave 83-year-old Nita Lowery on March 23, 2009 - an injection that allegedly plunged the Thompson House resident into a hypoglycemic coma and led to her death nine days later.

Dr. Richard Comi, an endocrinology specialist at Dartmouth-Hitchcock Medical Center in Lebanon, N.H., took the stand Monday - the fourth day of LaClaire's murder trial in Windham Superior Court Criminal Division - and said there was nothing in Lowery's medical history that should have spurred such a sudden collapse.

"I don't see how this could have happened without the administration of an agent to cause low blood sugar and that would have to be by injection," Comi said.

He added that "this was a profound, single event, which takes it out of natural causes."

But LaClaire defense attorney Dan Sedon spent about 90 minutes hammering away at Comi's conclusions, pointing to a lack of hard evidence and arguing that "speculation" is a key element in the state's case.

"Do you understand the danger of what you're doing here?" Sedon asked at one point during his cross-examination.

LaClaire, 39, of Bennington, N.H., was the sole nursing assistant on duty on Lowery's floor at Thompson House the night before she was found unconscious. Lowery was not a diagnosed diabetic, but LaClaire is and uses insulin to regulate her blood sugar.

LaClaire was arrested after a lengthy investigation and faces nine counts that are being prosecuted by the Vermont Attorney General's office - second-degree murder, abuse of a vulnerable adult and seven counts of financial exploitation of a vulnerable adult.

The latter counts stem from allegations that LaClaire, shortly after Lowery fell ill, began using the elderly woman's USAA credit card to withdraw money from her account.

The theft allegations nearly derailed the court proceeding Monday afternoon when Gail Speno, Lowery's sister-in-law, testified that there had been a total of approximately $4,700 in unauthorized withdrawals from Lowery's account.

However, LaClaire actually is charged with stealing a little more than half that amount.

After Speno stepped down and the jury left the courtroom, Sedon requested a mistrial. He argued that prosecutors had, by way of Speno's testimony, "introduced uncharged and unnoticed prior bad acts."

"We cannot unring the bell," Sedon said. "We couldn't have anticipated that was going to be (Speno's) answer, and now it's too late."

But Assistant Attorney General Ultan Doyle noted that neither prosecutors nor Speno specifically had claimed that LaClaire stole the full $4,700.

Doyle also pointed out that, on Friday during cross-examination of a Peoples United Bank fraud investigator, the defense had referenced ATM surveillance photos from a transaction in which LaClaire is not charged.

"It seems like the defense is speaking out of both sides of their mouth," Doyle said.

Assistant Attorney General Matthew Levine acknowledged asking Speno a "poorly worded or (poorly) thought-out question" but said the inquiry had not been designed to imply any additional criminal conduct.

Judge David Suntag, after leaving the bench for about 15 minutes to consider the matter, declined to declare a mistrial. He noted that jurors will be instructed that the state must prove each financial-exploitation count beyond a reasonable doubt.

"At this point, I don't think the prejudice is such to merit a mistrial," Suntag said.

That capped a day of testimony mostly consisting of Comi's appearance. He described insulin as "the calibrator and the dominant hormone in our bodies' handling of foodstuffs."

Giving insulin to non-diabetic patients will decrease their blood sugar, often leaving them uncomfortable and possibly shaky and sweaty, Comi said. But "the body's own natural processes" then kick in and correct the problem, usually within an hour or so, he said.

In Nita Lowery's case, however, that did not happen. In fact, witnesses have testified that she remained unresponsive even after an emergency injection of dextrose -- essentially, sugar.

Normal blood-sugar readings are in the 60 to 100 range, and two initial tests showed Lowery's blood sugar at critically low levels of 13 and 16. That jumped to 160 after administration of dextrose by medics, but Comi said that still remained "relatively low" for someone who had received the emergency injection.

Comi also said that, normally, "it doesn't happen quickly that one is unconscious with low blood sugar."

So he was surprised by initial reports that Lowery might have been seen awake and alert an hour or two before she was found unconscious. He believes the only culprit in such a quick hypoglycemic downturn is "intravenous injection of insulin."

"There's really nothing else that could do that quite that quickly," Comi testified.

Comi said a somewhat longer time frame might point to a subcutaneous insulin injection -- an injection under the skin rather than in a vein. Such an injection, while requiring more time to take effect, can be administered "very fast," he said.

Comi added that modern insulin needles - called a pen -- have a "very tiny needle."

"They're virtually painless to give an injection," he said. "And they can use them more than once."

Comi also said insulin needles are available without a prescription in Vermont, while insulin requires a prescription. In New Hampshire, he said, the opposite is true.

"You can get your supplies if you cross the border a couple of times ... without a prescription," Comi said.

Summing up his position, Comi was willing to take a step that Vermont's deputy chief medical examiner had not been: He tied Lowery's death not only to hypoglycemia, but also to an injection of insulin.

"Can you think of any other reasonable medical explanation for Nita Lowery's coma and death?" Levine asked.

"No," Comi replied.

Sedon then delved into now-familiar themes, pointing to alcohol, malnutrition and stroke as potential factors in Lowery's demise. When comparing hypoglycemic brain damage and a certain type of stroke, he argued, "these injuries, as you know, are so close that they're called mimics and chameleons of each other."

The defense attorney also pointed out that the initial emergency tests performed at Thompson House are the only samples showing that Lowery had low blood sugar.

"You can't say the evidence of hypoglycemia here is overwhelming. You just can't," Sedon said.

Sedon also invoked themes as various as French medical research, a hypoglycemic-diagnostic tool called Whipple's triad and the 1930s-era work of Austrian-American psychiatrist Manfred Sakel, who introduced insulin-shock therapy as a treatment for schizophrenics.

None of that convinced Comi, who was retained by investigators to review Lowery's case, to change his opinion.

Prosecutors have not produced -- and, if Sedon is correct, they won't be producing -- any physical evidence such as needles or drugs that would directly tie LaClaire to a fatal injection.

But Comi may have been summing up the state's case so far when he told Sedon that an insulin injection "is the most consistent story for all of the facts."

Mike Faher can be reached at mfaher@reformer.com or 802-254-2311, ext. 275.