IN THE FIGHT to keep drugs out of prisons, wardens resort to all kinds of arcane and invasive tactics. They turn cameras on visitors, let dogs sniff bags, and require extensive pat-down searches. Before being allowed into a visiting area, people must shake out their hair, pull out their pockets, strip off their belts, remove their shoes, open their mouths, and even remove their false teeth.

Now wardens are turning to what they call a "non-intrusive" method of discovering drug smuggling among prison visitors: the Itemiser Contraband Detector. The electronic drug-detection device — manufactured by Ion Track Instruments (ITI), in Wilmington, Massachusetts — seems simple to use. Prison visitors receive a piece of paper that looks like a coffee filter. They rub it on their hands, between their fingers, over their pockets, and on their shoes. The paper is inserted into the machine, which uses ion-mobility spectrometry — a scientific technique that measures charged atoms, or ions, as they move through an electromagnetic field — to scan for 40 different illegal substances, including cocaine, LSD, heroin, and PCP. Machines are currently in place at United States embassies, along state highways, and at customs checkpoints. The US Federal Aviation Administration (FAA) plans to install the Itemiser in 400 airports nationwide. And the device can be found at nearly 125 prisons in 28 states, including Florida, Pennsylvania, New York, and, as of January, Massachusetts. Under a pilot program, prison officials installed an Itemiser at the Souza-Baranowski Correctional Center (SBCC) in Shirley, a "maximum" facility featuring the most restrictive, state-of-the-art security measures in the Massachusetts prison system. Just six months into the program, however, prison visitors have joined a class-action lawsuit charging that use of the machine violates constitutional protections against unreasonable searches, as well as due-process guarantees. The suit also alleges that use of the machine violates Massachusetts Department of Correction (DOC) regulations detailing what kinds of searches prison visitors may be subjected to. No DOC rule authorizes searching people with an electronic scanner.

The desktop device looks innocent enough. Suzanne Glacken, 57, a retired Milford educator who first underwent the drill last January during a trip to visit her 30-year-old son Derek, compares it to an electrocardiograph. And she believed she had as little to fear from the Itemiser as she did from a heart machine. She says she’s never ingested an illegal substance in her life. She quit smoking cigarettes 40 years ago. She doesn’t even like using Tylenol. "The only drugs I touch are prescribed by my doctor," says Glacken, who takes medication for arthritis, asthma, and anxiety. "I figured, I don’t use illegal drugs. Why worry?"

But just a few weeks later, Glacken returned for another visit with her son, who’s serving a life sentence for murder. She routinely swiped her clothing and skin as required, and the machine’s computer panel lit up with a flashing red light. She was told she had tested positive for not one, but two street drugs: heroin and marijuana. "I nearly fell over," she recalls. "I thought, ‘You have got to be kidding me. This machine is wrong.’"

Glacken pleaded with SBCC guards for a retest, but they refused. She demanded to see a copy of her results, to no avail. Insulted and reduced to tears, she left that day and hasn’t returned since. Glacken is one of 237 visitors among 4442 — just over five percent — who have registered at least one positive scan by the Itemiser since January. Although she misses her son "terribly," she is convinced that she could touch off the Itemiser again — something that she does not want to risk. According to DOC policy, visitors who test positive once must submit to a pat-down search; if no contraband is found, they can enter the facility. People who receive a second positive result lose the right to "contact visits" for six months. Instead, they must be separated from prisoners by a plexiglass wall while they talk over a phone. Visitors with a third positive are barred from the prison for as long as DOC officials see fit. Glacken can’t shake her fear of being banned from visiting her son, she says. Nor can she shake the sense of shame that comes from "being flagged a heroin user by a machine."

"It’s like I’m Hester Prynne," Glacken adds. "I’ve got this big scarlet letter on my chest. Except I haven’t done anything wrong."

MASSACHUSETTS ISN’T the only place where the Itemiser — and its competitor, the Ionscan, manufactured by Barringer Technologies, Inc. in Warren, New Jersey — has come under fire. In October 1999, criticism prompted Missouri legislators to institute a temporary moratorium on using the Itemiser. And after a lawsuit was filed in that state similar to the one just filed in Massachusetts, Missouri lawmakers voted to maintain the ban indefinitely. The Itemiser also faces challenges in Colorado, California, and Louisiana.

More recently, Iowa erupted in controversy last summer when its Department of Corrections installed the Itemiser at nine facilities. Dozens of visitors — from six-year-old kids to 60-year-old grandmothers — appealed to the Iowa Civil Liberties Union (ICLU) after facing sanctions for what they claimed were false positives. Randall Wilson, who heads the ICLU legal division, says prison officials not only rescinded the visitation rights of those who had tested positive, but also released their names to law-enforcement agencies. In several cases, police seized the results as reason to investigate people on suspicion of drug-related crimes. When a toddler scanned positive for marijuana, his mother became the subject of a child-welfare probe by the state. At one facility, a warden grew so fed up with complaints that he volunteered to test the device for accuracy, Wilson says — only to test positive himself.

"The test is too sensitive," Wilson charges. He has since negotiated a "strict" Itemiser policy with the Iowa DOC. Positive scans can be retaken, followed by searches, and appealed. The DOC has also ceased handing results over to police. "This device discovers contact with illegal substances," he says, "but it cannot discriminate between innocence and guilt. In real life, the science is guesswork."

Despite these pockets of debate, however, ITI officials have evaded public scrutiny. The scientific community accepts ion-mobility spectrometry as sound. Independent analysts have verified the machine’s accuracy; according to the company, the FAA performed a battery of evaluations that showed the Itemiser to be effective at detecting explosives. ITI officials, in short, stand by their product — so much so that they declined to answer the Phoenix’s questions about the recent problems at SBCC. Explains ITI marketing director Jim Bergen, "We do not comment on how clients use the equipment." The company has trained SBCC staff to operate it, he says. "We provide the equipment, and it does what we say it does."

But not every scientist trusts ion-mobility spectrometry to detect drugs accurately. James Woodford, an independent scientist who has studied drug-detection devices for 25 years and who’s worked at forensic laboratories for the US Army and the US Customs Service, speaks for many chemists when he describes the Itemiser as "a watered-down, cheapened version of the real thing." By Woodford’s standard — and by those of government crime labs and the courts — the only truly reliable devices use gas chromatography and mass spectrometry (GC/MS) to detect drugs.

In a GC/MS system, substances taken from the filter paper are put in a chamber where they are converted into gas. The vapors are heated and broken down into ions, which are shot through an electromagnetic field. To identify the substances, the GC/MS machine is calibrated to measure various characteristics of the ions, including their mass, their chemical composition, and the time it takes them to move through the field.

The Itemiser, on the other hand, does not purify a substance by converting its components into gases. Instead, the machine bakes material to reduce it to ions. Unlike a GC/MS machine, the Itemiser simply measures the time it takes these ions to move through the electromagnetic field. By focusing on just one of many traits, Woodford says, the Itemiser comes to a premature conclusion. "Any chemist knows the length of time an ion flies doesn’t tell you much about a substance," he explains. "It’s like trying to construct a face but ending up with only an eyelid or a nose."

Simply put, it does not require much to trigger a positive result on these machines. Sometimes, the Itemiser correctly identifies a substance but can’t determine whether it’s coming from illegal drugs or prescription medication; for example, the Itemiser has no way of distinguishing the illegal compound methamphetamine from the kind of stimulants found in numerous prescription drugs. More often, the machine detects such tiny traces of contraband that, as Graham Boyd of the ACLU’s Drug Policy Litigation Project notes, "people can show up positive just because drug residue is so common in our society." If someone rents an apartment in which the previous renter had smoked pot, for example, the new tenant could produce a positive scan for marijuana. Someone who sorts significant amounts of money while working at a bank could test positive for cocaine (Woodford and other chemists have found that as much as 90 percent of American currency carries cocaine traces).

At a May meeting between the DOC and the class-action plaintiffs, even an ITI chemist admitted that it is possible for people to set off the Itemiser if they’ve had unknowing or innocuous exposure to drugs. In an account of the session obtained by the Phoenix, the ITI chemist recognized that a woman could test positive for marijuana if her son had smoked pot in the family car, and she later touched its steering wheel or dashboard. Likewise, a prison visitor who snorts cocaine before entering the facility could contaminate portions of the lobby. Those who touched the same doorknob or locker as that visitor could then test positive for cocaine.