A widely used method of detecting alcohol consumption in people prohibited from drinking is under assault from a federal agency that has declared the test too scientifically uncertain to be the sole basis for legal or disciplinary action.
The federal Substance Abuse and Mental Health Services Administration last week issued a so-called black-box warning asserting that the urine-alcohol screen called EtG doesn't offer surefire proof of drinking. Officials at the agency say the screen is so sensitive that a positive reading may reflect exposure to alcohol-based hand sanitizers or alcohol-containing foods or medicines. A carton of apple juice left a long time in the refrigerator could conceivably produce a positive EtG test, says Kenneth Hoffman, the agency physician who wrote the warning.
The warning represents a victory for the growing number of people who insist they flunked the EtG test despite having abstained from liquor. Their cases, replete with polygraph exams and other evidence of sobriety, convinced even the scientist who pioneered EtG screening in America that the test is prone to so-called innocent positives.
Whether the agency's warning will help these people reclaim the jobs that some lost after flunking EtG tests is unclear. In any case, the warning is a blow to the credibility of the $4 billion-a-year urine-testing industry, which introduced the EtG test two years ago as offering fail-safe proof of alcoholic-beverage consumption.
EtG, short for ethyl glucuronide, is a unique metabolite of alcohol that stays in urine for up to 80 hours -- four times as long as does alcohol itself. Earlier, detection of alcohol had been difficult because it dissipates so quickly. The wider window of detection made EtG an instant hit with drug courts, professional licensing boards and other agencies that monitor sobriety -- and an instant star of the urine-testing industry, which is performing tens of thousands of EtG tests per month in the U.S.
However, SAMHSA officials say the industry never conducted the large-scale clinical trials needed to prove EtG isn't prone to snare the innocent. No federal regulatory approval or rigorous trials are required for a urine-testing firm to introduce a new product or process.
Even after evidence emerged that the EtG test could detect incidental exposure to alcohol in food and the environment, many urine-testing firms continued marketing the screen as definitive proof of booze consumption. Some continue to do so. "EtG is not detectable in urine unless an alcoholic beverage has been consumed," says the Web site of a urine-testing firm called AccuDiagnostics LLC. An AccuDiagnostics spokesman attributes that claim to toxicologists at laboratories to which it outsources its samples.
At industry giant Quest Diagnostics Inc., the director of the Salt Lake City laboratory conceded during a July interview that exposure to alcohol in foods or medicines could produce a positive EtG score. It was only after The Wall Street Journal published a page-one article on Aug. 12 about EtG tests that Quest removed from its Web site a claim that "EtG is not detectable in urine unless an alcoholic beverage has been consumed."
A Quest spokeswoman says the company regrets not removing that information sooner. However, Quest says that the claim was based on an internal study of 1,500 abstinent people, none of whom tested positive for EtG. That study wasn't published. The spokeswoman says Quest is carefully studying the SAMHSA warning.
The warning makes clear that the EtG test remains useful. Increasingly, drug and drunken-driving defendants, along with recovering addicts in high-risk professions such as health care, are required to abstain entirely from alcohol and illicit drugs. But monitoring these people has been difficult because while illicit drugs can be detected for days after usage, alcohol consumption has been easy to disguise -- until the advent of EtG. About 10 percent of EtG tests have turned up positive, the vast majority of them reflecting genuine violations of sobriety requirements.
But a small percentage of those positive findings appears to involve no wrongdoing. In one case, a California pharmacist named Lorie Garlick -- whose pharmacy license has been suspended since she failed an EtG test in the spring of 2005 -- quarantined herself in an addiction-treatment center with no access to booze and flunked the test again.
Because of such cases, "legal or disciplinary action based solely on a positive EtG ... is inappropriate and scientifically unsupportable at this time," said the SAMHSA warning, recommending that a positive EtG be regarded as a possible sign of relapse that triggers a broader investigation.
A negative EtG score appears to represent persuasive evidence of sobriety. This was what Gregory Skipper, a physician who runs the Alabama monitoring and assistance program for recovering doctors, was seeking when he helped to pioneer the EtG test along with some European doctors several years ago. As a recovering addict himself, Dr. Skipper understood that malpractice insurers and state licensing boards desire documentation of abstinence. Dr. Skipper, who supports the SAMHSA warning, says he hopes it doesn't overshadow the EtG's value as a marker of sobriety.
Many urine-testing firms say that they merely provide the EtG results and that their clients -- drug courts or professional licensing boards -- bear responsibility for deciding whether a positive finding represents proof of drinking. But the urine-testing companies themselves have guided clients to interpret positive results as proof of drinking. On a laboratory report stating that Nancy Clark, a Pennsylvania nurse, had an EtG score of more than 300 nanograms per milliliter, National Medical Services included the statement that "any value above 250 ng/ml indicates ethanol consumption."
Ms. Clark has passed a polygraph test stating that she hasn't drunk, and her 12-step group awarded her a medallion in May honoring five years of abstinence from alcohol and drugs. But two positive EtG scores prompted Pennyslvania to suspend her nursing license early this year. Now, the 20-year veteran of nursing waits tables at Charlie Brown's Steak House in Reading, Pa.