IS DRUG TESTING GOOD OR BAD? Businesses will spend $200 million this year on urine tests that are not foolproof. Still, for some companies, particularly those involving public safety, testing makes sense.
By Andrew Kupfer REPORTER ASSOCIATE Karen Nickel

(FORTUNE Magazine) – AS THE AGE of Aquarius turns to the age of abstinence, more and more companies are requiring job applicants and employees to take urine tests to prove they do not use drugs. Managers believe that drug-abusing workers cost millions in lost productivity, absenteeism, and medical expenses. Companies responsible for public safety live in fear of tragedies like the 1987 train wreck near Baltimore involving a Conrail engineer and brakeman who smoked marijuana before a fatal collision with an Amtrak train. The Department of Transportation recently announced a plan to require random tests of four million railroad, trucking, and airline workers -- and has already been challenged by the unions. The tests are becoming so common throughout industry that classified ads in the drug-culture magazine High Times offer products such as Uri-Clean that promise to beat the system. For all this popularity, there is great confusion over what drug tests prove. Some companies report that absenteeism, medical costs, and productivity have improved since they started testing job applicants and workers suspected of drug use. But these same companies say the tests were accompanied by a raft of managerial initiatives that may be equally responsible. The Navy, which performs more drug tests than any other American institution, has found only a low correlation between smoking marijuana, by far the most popular drug, and performance. As yet there is no solid evidence that candidates who flunk preemployment tests are more likely to perform poorly than those who pass. Part of the problem is that the drug tests reject the casual weekend marijuana smoker as readily as the daily injector of heroin. Another part of the problem is less-than-perfect testing. No wonder managers are confused. When do drug tests make sense and when don't they? One place they clearly make sense: corporations whose operations involve public safety. Airlines, railroads, bus companies, and electric utilities can't be too careful. They have good reason to test job applicants for the presence of drugs. What about preemployment drug testing in less ''safety sensitive'' ! industries? Nearly half of all major U.S. companies, including IBM, Kodak, AT& T, Lockheed, 3M, and Westinghouse, require some or all job applicants to provide urine specimens. These companies say they have a responsibility to provide a safe, healthy, and productive environment for their workers and use preemployment drug tests in the hope of avoiding potential problems later on. If they lose some excellent potential employees, they have many more eager, qualified applicants. Companies in low-paying service industries cannot be so selective. Drug tests could eliminate so many candidates that these employers cannot fill all their slots. ONCE ON THE JOB, which employees should be tested for drugs? Again, the case is strongest for industries involving public safety. That may mean subjecting employees to drug tests, even by surprise and at random, to protect themselves, their co-workers, and the general welfare. The Nuclear Regulatory Commission is drawing up guidelines for the random testing of power plant operators. But random testing stirs employee resentment, is not likely to yield many offenders, and may be an invasion of privacy. Though the courts are leaning toward that interpretation, they have not ruled conclusively.

As for less safety-sensitive companies, managers should try to spot drug- related problems without testing the entire work force. However, if a worker's performance or absenteeism makes a manager suspect that he is using drugs, a test is an appropriate way to confirm those suspicions and lead him to treatment. After that, further tests can monitor an employee's progress as part of a broad program of counseling and rehabilitation -- with firing the last resort for those who continue to fail. AMERICAN laboratories will process 15 million to 20 million drug tests this year. About half will be for companies, the rest for prisons, police, and public drug-treatment programs. Of the corporate tests, 85% will be for preemployment screening. Businesses will pay about $200 million in 1988 for drug testing, a figure some experts estimate will reach $500 million a year by 1991. Besides trying to generally improve worker productivity, companies might begin preemployment testing for several reasons. Other employers in their area may have done so and they could fear being swamped by drug rejects. They may worry about being sued for negligent hiring if an employee using drugs causes an accident that harms the company, the public, or fellow employees. And they may simply be responding to a fundamental change in the zeitgeist. In the 1960s many people looked upon drugs as a way of enhancing the imagination without serious side effects; now people tend to see them as detracting from performance and downright dangerous. While drug tests are reliable, they are far from foolproof. First, testers have to make sure that the urine sample belongs to the person who fills the bottle. The only way of knowing, as any baseball fan can attest, is to see the ball enter the glove. But because watching someone urinate is taboo, observers often avert their eyes and may not see a specimen being tampered with. The new guidelines for testing federal employees say that the donor may produce a urine sample ''in the privacy of a stall.'' Besides the possibility of the old switcheroo, adding common table salt to a sample may sneak it past some initial screenings. Urinalysis is better than blood tests at detecting the presence of drugs. Impurities are more highly concentrated in urine by a factor of 10. The telltale signs of drugs remain in urine much longer than they do in the bloodstream. (However, a blood test is more efficient in determining drug use within the past few hours.) The urine test is a high-tech, two-part affair, increasingly performed by such companies as PharmChem Laboratories in Menlo Park, California, whose only business is drug testing. Samples arrive by overnight courier in sealed bottles with a label signed by both the donor and the person who collected the specimen. First the lab runs each sample through a screening test for an array of substances, including marijuana, cocaine, opiates, amphetamines, barbiturates, and hallucinogens. Screening exams are simple to run and relatively cheap, usually under $10 each. The most common screen is the enzyme multiplied immunoassay technique, or EMIT, which is made by Syva Co. EMIT uses ultraviolet light to tell if a target drug in the urine binds with a drug antibody in the test mixture. Screening exams are extremely sensitive, but not very specific; that is, many legal substances -- such as poppy seeds and cold medicines -- can set off alarm bells. The International Olympic Committee in Seoul prematurely announced that British runner Linford Christie had failed a drug test only to recant and let him keep his silver medal when it found that the banned substance in his urine came from ginseng tea. Up to 10% of samples will register as false positives. So all positive results should be presumed innocent until the lab runs a more accurate confirmatory exam. Still, according to Northwestern University's Lindquist-Endicott Report, one-third of companies do not retest if the initial test is positive. Companies should ask applicants to list any medications that might cause a positive test result. But as Washington management consultant Theodore Rosen says, ''If you don't remember that two weeks ago the doctor gave you Tylenol Plus 3 -- the 3 being codeine -- you're on the hot seat.'' Employers do not usually tell applicants why they failed to get a job unless they ask, so they rarely have the chance to explain an innocent cross-reaction. For that reason such states as Connecticut, Minnesota, and Vermont require that test results be sent directly to the applicant. The best confirmatory test is called gas chromotagraphy/mass spectrometry, or GC/MS, which uses $100,000 machines made by Hewlett-Packard or Perkin- Elmer. GC/MS makes a positive identification by breaking down drugs into their constituent molecules. The test is intricate and expensive, costing $40 to $70 each. But if the complex GC/MS test is not administered expertly, false negatives and false positives can still pop up. The National Institute on Drug Abuse is compiling a list of laboratories that conform to its standards; companies should use one that does. MISTAKES and misrepresentations aside, what can a urine sample tell potential employers? A test can show only past exposure to drugs. It cannot tell how much of the drug was consumed or when. It cannot prove drug addiction or the degree to which the drug impairs the faculties of the user, if at all. A true positive result does not mean that a person was under the influence of the drug at the time of the test, for the metabolic byproducts may remain in the urine for up to a month in the case of marijuana. At the same time, a person in the worst stages of cocaine withdrawal may show no trace of the drug because it leaves the system in about three days. Some companies avoid preemployment drug testing because it could wipe out their worker pool. Drug use is highest among men 18 to 35 years old without college degrees, who include most applicants for low-paying jobs. Attorney Jerry Glassman, a partner in a New York City management-labor law firm, reports that a government agency in the Northeast, which he won't identify, recently gave drug tests to 1,000 applicants for security guards; 980 failed. As it is, a general shortage of unskilled labor will hamstring business in the 1990s. Companies that need to fill a lot of these jobs face tough choices. For all the problems with preemployment testing, some companies do it right. IBM requires a urine test of all job applicants. Unlike many corporations, IBM tells applicants if they failed the test and gives them a chance to come up with a solid medical reason. If they cannot, they can reapply after six months. COMPANIES that test job applicants often mistakenly assume that most employees with drug problems started using the substances before they were hired. But Thomas E. Backer, president of the Human Interaction Research Institute in Los Angeles, surveyed 1,238 employee assistance programs -- confidential services that help workers with personal problems -- and found that on-the-job pressure led to a large number of drug-abuse referrals to the programs. Says Backer: ''Self-medication for stress is the single most unrecognized and unexplored reason for abuse of drugs in the workplace, and one that companies can't screen out with preemployment tests.'' Most controversial is random testing of all employees. While it is a strong deterrent to drug use, random testing is not necessarily good management practice. An official with a nuclear power company that tests employees only when suspicions warrant does not look forward to the random testing required by the proposed Nuclear Regulatory Commission regulations. He says: ''Our supervisors have been brought along to be able to spot employees to test for cause. If random testing occurs, they might take a step backward and say, 'Well, they'll get caught by the random test.' '' IBM does not perform random tests and tells managers to give employees every benefit of the doubt before recommending a drug test. If workers test positive, Big Blue offers treatment and gives them periodic drug tests during a one-year probationary period. If they fail during the year, or anytime thereafter, they may be fired. Many companies say testing has helped lower costs and improve working conditions. The injury rate at Southern Pacific Transportation Co. is down over 60% since 1984, when it began preemployment and post-accident testing. Commonwealth Edison reports 25% to 30% lower absenteeism after six years with a drug program and a reduced rate of increase in medical costs. At Georgia Power sick days have fallen 23%, and serious accidents decreased from 39 in 1981, the year before drug testing was adopted, to nine in 1987. Georgia Power's program is unusual. The company encourages workers to go to its employee assistance program if they have a drug problem; if they do not come forward and managers suspect them of drug use, the company will order a drug test and dismiss them with no second chance if they fail. ''You might say it's a nonrehabilitative approach, but we say the opposite,'' says labor relations coordinator Howard Winkler. ''Since the price of detection is so high, it encourages employees to seek assistance before they're caught.'' The company has fired 75 to 100 employees who resisted the encouragement; 527 came forward for help. Over union objections, Georgia Power is about to start random testing of all employees. When it does, it will offer a second chance to those who fail. SOUTHERN PACIFIC gives its drug tests total credit for the impressive improvement in its injury rate. At Commonwealth Edison and Georgia Power the drug program was part of wide-ranging attempts to curtail injuries and health costs and increase productivity. To determine the relationship between drug testing and productivity, the U.S. Postal Service is conducting a novel study. Last year the post office told 5,400 job applicants in 20 cities that they had to take a urine test, which would be used for research only. The personnel managers doing the hiring had no access to the tests; they judged the applicants on other criteria. Only the Washington headquarters staff saw the results. The post office hired 4,156 of the people who took the test; 340 of them, or 8.2%, tested positive for drug use. The breakdown: 236 for marijuana, 71 for cocaine, and 33 for other illegal substances. The post office will monitor the careers of the drug-positive hirees for two years to gauge their performance against that of other new workers. Lou Eberhardt, a post office spokesman, says that preliminary results will be available early in 1989, about halfway through the surveillance period. When news of the study broke recently, union leaders complained about using workers as guinea pigs, and drug enforcement proponents criticized the government for hiring people known to take drugs. Both groups should cool down and wait for the results. On the one hand, the tests cost no potential union member a job. On the other, without the research there would have been no drug test, so the post office would have hired the drug-using applicants anyway. AN EXECUTIVE of a utility that routinely tests potential and current employees worries that companies are rushing to adopt drug testing without giving it much thought. ''It's almost a case of managerial macho,'' he says. ''If all you do is drug testing, if you don't train your supervisors to spot problems, or promote your employee assistance program, or communicate clearly and sympathetically with your employees, then you don't have a drug program.'' Before joining the burgeoning ranks of willy-nilly drug testers, managers might want to consider the two companies that make the fancy mass spectrometers -- Perkin-Elmer and Hewlett-Packard. Neither uses drug tests. They say that they know their employees, so they don't need to.