Testing Testing
The sweat patch was supposed to solve the problems of urinalysis, but it created a host of its own.
By Lara A. Bazelon
SHERYL WOODHALL, A CALIFORNIA WOMAN in her late 30's, first lost custody of her four children in 1995, when her youngest tested positive for methamphetamine at birth. The state's Child Protective Services intervened and sent her two older children to live with her parents and placed the younger two in foster care. Woodhall, who acknowledges that she struggled with drug addiction for years, went into a residential treatment center, then to an outpatient program. In October 1998, she got her kids back three days a week under the condition that she continue regular drug testing.
CPS suggested that Woodhall wear a sweat patch, a new drug-testing technology, to show Judge Lois Haight, who was presiding over the custody case, that she was staying clean. "The judge believes in this patch," a social worker from CPS told her. Woodhall's first patch fell off, and she lost it. Her second patch came back positive for methamphetamine. So did her third one.
Woodhall was ordered back to court. Although random urine tests performed by her outpatient program during the same period of time all came back negative, her three days a week with her kids were reduced to one a month. The judge ordered her to continue wearing the patch, and she repeatedly tested positive. In March 1999, the court held a hearing to determine whether to grant a request by the state that Woodhall's four children be placed in foster care because of her positive drug tests. At the proceedings, the prosecutor called Neil Fortner as an expert witness to testify on behalf of the patch's reliability. Fortner was an employee of PharmChem, the Texas-based commercial laboratory that markets the sweat patch.
Based on the positive tests and Fortner's testimony, Judge Haight granted the state's request and terminated Woodhall's parental rights. "The patches were positive and that's what she was going with," Woodhall said recently. "I could have taken a hundred urine tests and it wouldn't have mattered."
DRUG TESTS HAVE BEEN A REGULAR EVENT in the lives of many Americans since the Reagan Administration launched its War on Drugs in 1986. All told, about 25 million people are tested in the United States each year—as a prerequisite for serving in the military, piloting a commercial airliner, operating a forklift, selling kitchenware at Wal-Mart, or playing on the high school football team.
Until the late 1990s, substance-abuse screening for people convicted of federal offenses had been done almost exclusively through urinalysis, considered the gold standard because of the refinement and standardization of its testing protocol after decades of scientific scrutiny and government oversight. In use since the Army began screening personnel with it in 1972, urinalysis is the only method used to test the millions of federal workers in "safety-sensitive positions"—mainly Department of Transportation employees, law enforcement agents, and officials involved in national security matters—for whom substance-abuse screening is mandatory.
But urine testing isn't perfect. It's time consuming and invasive: In the federal justice system, a probation officer is often stationed in the bathroom when the sample is given to prevent tampering. Urine testing provides only a "snapshot in time," detecting drug use during a period that may cover as few as 72 hours, depending on how quickly the drug moves through the offender's system; stimulants like cocaine and methamphetamine may disappear from the body in less than two days.
Urine samples are also highly susceptible to tampering and contamination. Anyone with access to the Internet can visit dozens of websites offering everything from dehydrated drug-free urine samples to advice on which over-the-counter and prescription medications are most likely to mask the effects of substance abuse.
PharmChem's sweat patch has undeniable advantages over urine testing. Worn on the upper arm, the patch looks like a big Band-Aid and works by trapping sweat excreted by the body in an absorbent pad. Once the patch is removed, the adhesive is thrown away and the pad is taken to a laboratory, where the collected sweat is analyzed for drug content. Wearing a patch is less embarrassing than peeing into a cup under supervision. Since the patch can stay on the skin for up to two weeks, it provides continuous monitoring by retaining all drugs used during the time it is worn, catching users who would "beat" a urine test by getting high sporadically or by flushing the drugs out of their systems.
These advantages have made the sweat patch a popular alternative to urine testing. Since the mid-'90s, state and federal governments have used the patch to test thousands of people, mainly criminal defendants and parents like Sheryl Woodhall who are battling a state for custody of their children.
Nowhere is drug screening more common, or the consequences more severe, than in the federal criminal justice system, where most offenders must submit to drug tests on a regular basis as part of their probation or supervised release (the equivalent of parole), with the understanding that they will face jail time if the results come back positive. The Administrative Office of the U.S. Courts, the agency that manages the federal judiciary and criminal justice system, authorized its probation departments to use the patch to test federal offenders in the fall of 1998. Two years later, the agency was still wholeheartedly endorsing the new drug-testing technology. Administrative Office director Leonidas Mecham stated in a September 2000 memo that the patch had been "deemed superior" to urine testing for the detection of cocaine and amphetamines.
Federal probation offices were convinced that the sweat patch was more convenient and reliable than urinalysis. At the height of its popularity, more than 80 percent of federal probation offices used the patch in some capacity. Recently, however, the federal system's use of the patch has come under attack by defense attorneys who have challenged the patch's reliability. They have questioned whether PharmChem's financial stake in the patch's success may have led it to make claims about the product that have been contradicted by the company's own research.
PHARMCHEM'S CEO, JOE HALLIGAN, declined to answer questions about the sweat patch, referring all inquiries to the company's lawyer, who despite requests made over the course of several weeks, did not provide any answers. But PharmChem is a publicly traded company, and the battles over the sweat patch have taken place in courtrooms nationwide, so the story of the technology that was supposed to replace urinalysis can be traced through public documents and the experiences of lawyers on both sides of the sweat-patch cases.
PharmChem began marketing the sweat patch as a drug-testing device in 1992, funding studies to establish that it was capable of isolating five commonly abused drugs—cocaine, amphetamines, opiates, phencyclidine (PCP), and marijuana. The studies were submitted to the FDA, which cleared the patch as a screening device for all five drugs in July 1995. Before acquiring the rights to the patch, PharmChem had mainly provided urinalysis testing services; by 2000, it had grown into a company with more than 300 employees that generated nearly $40 million in annual revenue and maintained laboratories in Menlo Park, Calif., and London.
Between 1992 and 1996, the company invested heavily in the sweat-patch technology, spending over $4 million on marketing, research, and development. Its expectation was that the sweat patch would become a popular tool in the company's drug-testing arsenal as its use spread from the criminal justice system to the federal workplace and the private sector, where the demand for drug-testing services was steadily increasing, along with the potential for profit.
Although PharmChem has long provided drug-testing services to the private sector, nearly half of its business comes from contracts with criminal justice agencies. Its most lucrative contract remains the one it struck with the Administrative Office. Criminal defendants in the federal system may be tested by urinalysis or any method of "equivalent accuracy." Aaron Lucas, a probation administrator in the Administrative Office, noted that his office did not approve the sweat patch for testing on federal offenders until some three years after the FDA had cleared it for the market.
In the interim, the Administrative Office reviewed the documentation submitted to the FDA and the results of two separate clinical trials involving the sweat patch that were undertaken by the office's Federal Corrections and Supervision Department and its Office of Pretrial and Probation Services. While confirming that the sweat patch was highly effective at detecting drug use—more effective, in most cases, than traditional urine testing—both studies were field tests and therefore not subject to rigid scientific controls.
Still, the Administrative Office advised its probation officers that the sweat patch was "a credible illicit drug detecting tool." Michael Severance, a drug-treatment supervisor for the federal probation office in Las Vegas, estimates that by 1999 he was purchasing approximately 800 patches annually at a cost of about $25 per patch (including the cost of testing services). "We were told that the sweat patch was the Second Coming," he said. "I pushed hard for it to be used in Nevada because I believed in it totally." Sales of the patch jumped by nearly 50 percent between 2000 and 2001.
But the patch's strengths were also its weaknesses. The amount of a drug that must be present in a sweat patch for that patch to be reliably judged positive (known as a cutoff level) was quite low. PharmChem's cutoff levels, measured in billionths of a gram, were approximately one-thirtieth of those set by the federal government for urine testing. The patch's opponents believe that low threshold raised the risk of false positives.
FRANNY FORSMAN HAS BEEN THE HEAD of the Federal Public Defender's Office in Las Vegas since 1989. Between 1998 and 2000, she litigated seven sweat-patch cases on behalf of clients accused of violating their supervised release by using drugs. In the process, she learned that the patch was perhaps too good at detecting drugs.
In 1999, Forsman began to notice similarities among her clients that led her to question the patch's reliability. Having run a drug-treatment program before becoming an attorney, she had extensive experience working with addicts. Her sweat-patch clients weren't fitting the addicts' profile.
Unlike clients who had positive urine tests, many sweat-patch clients steadfastly denied using drugs, even when those denials made them ineligible for treatment programs, resulted in the revocation of their supervised release, and exposed them to the possibility of going to jail. They had every incentive to admit they were users because there were benefits to them if they did and costs to them if they didn't.
There was also conflicting evidence—many had taken urine tests that had come back drug free, contradicting the results of the sweat patches they were wearing at the same time. Finally, many lived in homes that may have been contaminated with drug residue. Forsman began to suspect that the patch was so sensitive that its wearers didn't need to have taken drugs to test positive, they only needed to have been near them.
The environmental contamination theory that Forsman's work suggested anecdotally turned out to have scientific support as well. Dr. David Kidwell, a chemist at the Naval Research Laboratory in Washington, D.C., and Dr. Frederick Smith, the director of the Criminalistics Laboratory at the University of Alabama at Birmingham, decided to study the patch after reviewing PharmChem's FDA submissions and after listening to the complaints of Forsman and other criminal defense lawyers. "There had really been no research to specifically address these questions," Smith said. He believed there were "significant gaps in the work that had been done."
Kidwell and Smith thought that environmental contamination was very much a possibility, and they set out in the summer of 1999 to test the sweat patch's vulnerability to such contamination. In their experiment, they placed the equivalent of one-tenth of one percent of an average methamphetamine dose on their skin, showered, and then applied a sweat patch the following day, after cleaning the contaminated area with two alcohol wipes, as recommended by PharmChem in its sweat-patch application manual. During the rest of the week, they engaged in normal hygiene and activities, and added an additional patch each day.
In a proceeding in federal court in the fall of 1999, Smith testified that the patches he and Kidwell applied on the first day and wore throughout the week contained approximately 59 times the PharmChem cutoff level for methamphetamine. The patch applied on the final day and worn for 24 hours came back with seven times the cutoff.
Kidwell and Smith believed that some of the methamphetamine they used as a contaminant formed a chemical bond with their skin that survived normal hygiene and alcohol swabbing. That portion remained fixed on the skin until loosened by sweat, at which point it soaked into the absorbent pad and created a positive test result. They asserted that there was a distinct possibility of drugs settling on the bodies of non-drug users as they went about their normal business. They cited experiments in which items like money, furniture, and even the foreheads of children had been shown to retain trace amounts of drugs. According to Kidwell and Smith, because many patch wearers lived in drug-contaminated environments—for example, houses or apartments that had previously been inhabited by drug users—"the very people most likely to be tested by the sweat patch are also the most likely to be externally contaminated."
CHALLENGES TO THE SWEAT PATCH relying on Smith and Kidwell's research first arose in federal court in November 1999. PharmChem insisted the claims were baseless. The company emphasized that the FDA had cleared the patch for drug testing after the agency had reviewed numerous studies. Experts for the government (often employed by PharmChem), which was using the patch as a basis to revoke the supervised release of federal offenders, pointed out that the FDA studies didn't consider the possibility that drugs in the environment could bond to the skin and cause a positive test result.
But PharmChem gave no ground, continuing to insist that "thoroughly and aggressively" scrubbing the skin with two alcohol wipes was adequate to cleanse it of contaminants.
In the opinion of PharmChem toxicologist James Meeker, the amount of drug contaminant used in the Kidwell and Smith study was far greater than the amount likely to be found in the environment. He also criticized Kidwell and Smith for mixing the drug contaminant in an alcohol-based solution to facilitate its application to their skin, arguing that this type of drug transfer would never happen in "real-world conditions."
The company's response in these legal proceedings was consistent with its practice of aggressively defending its testing methods. According to company documents, PharmChem representatives testify as experts in an average of 100 adversarial proceedings each year. In the first six months of 1999, PharmChem representatives testified to the reliability of the sweat patch in court proceedings in Cleveland, Las Vegas, Los Angeles, and Cedar Rapids, Iowa. In each of the cases, judges found the sweat patch reliable.
BUT WHILE PHARMCHEM'S DEFENSE of the patch was vehement, the company also relied on questionable tactics and testimony. In United States v. Stumpf, a sweat-patch case that went to a hearing in January 1999, Joseph Sullivan, an assistant U.S. attorney, called Neil Fortner, the PharmChem toxicologist who would later appear in the Woodhall case, as an expert witness. Fortner's performance, Sullivan said recently, "was a nightmare."
During cross-examination, under questioning by civil rights attorney Julian Gross, Fortner acknowledged that his description of his Ph.D. as nearly completed and based on "doctorate work in the area of neurochemistry at Cleveland State and San Francisco State," was "not a correct statement." Fortner admitted that he had testified falsely in several earlier sweat-patch cases about his academic credentials and had misrepresented the results of several FDA studies as well.
Fortner's concessions came after Gross produced two letters, one from an administrator at San Francisco State confirming that Fortner had never been a student in any of its graduate programs, and the other from a Cleveland State vice provost who noted that "Cleveland State University does not offer, nor has it ever offered, a Ph.D. program in neurochemistry." According to the second letter, Fortner had enrolled in Cleveland State's master's program in biology in 1987, but he never completed his degree and "is no longer considered to be an active student." "Technically, Fortner is not even a scientist," Gross said recently, "and yet in a lot of cases he was the only expert who testified."
Still, Sullivan believed at the time that Fortner's statements were more of a personal failing than an indictment of PharmChem or the sweat patch. Sullivan based his assessment in part on the results of the Stumpf case itself: Although Fortner's credibility was badly damaged, the judge concluded that the patch was a reliable drug-testing tool based on the opinion of Sullivan's second expert, Dr. Edward Cone, the former chief of clinical pharmacology at the National Institute on Drug Abuse, who had impeccable credentials and no ties to PharmChem.
Cone believed the Kidwell and Smith study was seriously flawed. He testified that scientists doubling as subjects in their own experiment "is not a good way to do science" because "you cannot divorce your own bias from the outcome of the study," and he characterized their findings of environmental contamination as based on unrealistic conditions.
Cone argued that the amount of drug contaminant used by Smith and Kidwell was "extreme" when viewed in the context of the relatively small area of skin to which it was applied. The sweat patch, in Cone's opinion, was essentially reliable. The judge agreed and found that Stumpf and his codefendant had violated the terms of their supervised release.
PHARMCHEM'S WINNING STREAK was relatively short-lived, however, thanks in large part to a case called U.S. v. McLemore.
No one disputed that Marsha McLemore, a single mother in her late 30's, had made impressive strides in the 18 months since her release from prison on federal drug-trafficking charges. Even as her probation officer asked a judge to send McLemore to jail in the fall of 2000, he described her as a woman "with lots of good things going for her" and ticked off a list of her achievements: holding down a job as an aide in a nursing home, caring for her 9-year-old son, and saving enough money to move into a one-bedroom apartment in Las Vegas. Michael Severance didn't want to see McLemore go back to prison but felt that he had no choice but to recommend it. "She had 19 positive drug tests," he said. "We had to do something."
McLemore had submitted to regular drug testing as a requirement of her supervised release. Although she was also submitting to urinalysis during the same nearly two-year period, sometimes as often as every other day, all but one of her 19 positive tests came from sweat patches; the exception was a urine sample taken over a year before the government requested that her parole be revoked. Between late February 1999 and early October 2000—when the positive sweat-patch tests occurred—McLemore had submitted 109 negative urine samples at 48-hour intervals. For part of that time, she lived in an apartment that had previously been occupied by a drug user.
Sullivan nevertheless expected the McLemore case to go much as the Stumpf case had. "Based on the witness list, it looked like it was going to be the same thing all over again," he said. But undeterred by Stumpf, federal defender's offices around the country had continued to press the environmental contamination issue, and a few sweat-patch cases had been dismissed.
The successful challenges buoyed Forsman, who was representing McLemore. She was convinced that PharmChem must have done studies to put the environmental contamination issue to rest once and for all; she believed that the company couldn't afford to do nothing about it, particularly when the necessary experiments were relatively inexpensive. Several times, she asked Sullivan to retrieve any in-house studies that PharmChem had conducted prior to the McLemore hearing and turn them over to the defense as required by law.
Shortly before the hearing, Sullivan and Severance sat down with James Meeker, the PharmChem representative whom Sullivan planned to call as a witness, to prepare him to testify. Severance recalls that Meeker "seemed very nervous." Severance was unsettled enough that he recalls saying to Meeker at one point, "You are acting like there's a problem." But Meeker denied there was anything wrong.
Following up on Forsman's request, Sullivan asked Meeker if PharmChem had done its own studies "to come up with ammunition to counter" the Kidwell and Smith article. Meeker mentioned that PharmChem had funded a study at the Center for Human Toxicology at the University of Utah, but it was not yet complete and no reports about it were available.
Meeker also told Sullivan that PharmChem had done two minor experiments of its own, one involving sweat patches whose surfaces were wetted with drugs and fixed to glass plates. The purpose of this experiment, he explained, was to determine whether drugs could penetrate the patch membrane from the outside and seep into the pad. In the second experiment he described, several PharmChem employees had worn the sweat patch under drug-contaminated T-shirts, again to test the vulnerability of the patch to outside contamination. None of the patches in either study had tested positive for drugs. Sullivan relayed this information to Forsman.
When Meeker testified at the McLemore hearing, Sullivan asked him to discuss PharmChem's internal studies once again. Meeker described the glass plate experiment and then made a passing reference to the T-shirt study. Sullivan said, "Okay. So you're talking about there were two separate tests done?"
Meeker responded, "We've done numerous tests. One—I described one test just now, and we've done two additional types of tests, as well."
Forsman's ears immediately pricked up at the reference to a third test that she had never heard about. On cross-examination, she challenged Meeker's opinion that alcohol wipes can effectively remove all drug residue from the skin. Meeker replied that he had "done studies at the lab" on that issue. "You've done studies in the lab on placing drugs on the skin?" Forsman asked.
Meeker described an experiment that mimicked the study performed by Kidwell and Smith. At the highest concentration of drug applied to the skin of the subjects in the PharmChem study—a fraction of the dose applied by Kidwell and Smith—two of the five sweat patches tested positive for drug amounts above the cutoff level.
"It was like bombs going off," Sullivan said. "[The study] totally undermined the reliability of the sweat patch." Shortly after Meeker's testimony, there was a break in the proceedings, and Sullivan huddled with Severance and McLemore's probation officer. Neither of them knew anything about the third study. "I was furious," Severance said. "Meeker had every opportunity to tell us about those tests and he didn't."
Sullivan says that had he known about the study, he would not have brought the case. Unlike Fortner's statements in Stumpf, which did more to damage his personal credibility than bolster substantive claims of environmental contamination, Meeker's surprise disclosure about PharmChem's internal study went to the heart of the case against the sweat patch. Looking back on the meeting with Meeker before the hearing and Meeker's failure to give him the information he asked for, Sullivan said, "My perception was that he was willing to omit the truth, but not to commit perjury."
Sullivan asked that the case be dismissed, and the judge granted his request. PharmChem made no public statement in response to subsequent coverage of the case in the Nevada press at the time. Meeker left the company shortly after the case and now works as a toxicologist at a drug-testing laboratory in northern California.
AFTER MCLEMORE, SEVERANCE'S OFFICE no longer relied on positive sweat-patch tests in court unless there was substantial evidence corroborating a defendant's drug use. The outcome in McLemore and in U.S. v. Snyder, a subsequent case in which the results of PharmChem's internal studies were introduced, seemed like very bad news for PharmChem and its shareholders.
In the Snyder case, the presiding judge ruled that the company's internal tests "support the conclusion that the sweat patch is susceptible to outside contamination." The court's decision in Snyder is frequently cited in other sweat-patch cases as evidence that the patch is not a reliable indicator of drug use, particularly when the defendant's living conditions suggest the possibility that his body picked up traces of drugs from his environment.
It's difficult to say how much of PharmChem's financial woes have come as a result of the legal battles over the sweat patch, but in 2001, the year the Snyder case was argued, sales of the sweat patch dropped by almost 35 percent. Many factors may be at work, including the struggling economy and competition from other testing providers, but McLemore and Snyder, and the negative press coverage of those cases, suggest that PharmChem, in its zeal to promote the sweat patch and defend its reputation, served the interests of neither.
In February 2003, the company received a letter from NASDAQ stating that the value of its publicly traded stock had fallen below $1 million, the minimum amount necessary for its shares to be traded on that exchange. Unless the company's finances improved, the exchange warned, PharmChem would be delisted at the end of May.
WHILE THE MARKET HAS PUNISHED PHARMCHEM, the federal system has been slow to let it go. With the blessing of the Administrative Office, most probation offices continue to use the sweat patch as an independent basis for prosecuting offenders for suspected violations of their supervised release. The office has not changed its position regarding the sweat patch's validity, and the patch is still relied on in criminal prosecutions in 40 federal districts, including probation offices in Boston, San Francisco, Los Angeles, and Kansas City, Mo.
PharmChem's contract with the Administrative Office expired at the end of 2002 but was extended through August 2003 while the agency goes through its usual practice of soliciting contract proposals for drug testing from interested companies.
Kimberly Golder, a probation administrator at the Administrative Office, declined to comment directly on whether PharmChem's finances and the ongoing controversy surrounding the sweat patch would affect the company's chances of securing a new contract with her office. "They have the right to bid," she said, "just like everyone else."
Lara A. Bazelon is a federal public defender in Los Angeles.
http://www.legalaffairs.org/issues/July-August-2003/review_bazelon_julaug03.msp
