The Post, in its Feb. 17 editorial chastising the Ninth Circuit Court of Appeals' decision striking down postaccident testing in the railroad industry as unconstitutional, seems to have developed a blind faith in the powers of drug testing. The editorial states that in the tragic Conrail accident, the evidence that the engineer's judgment had been impaired by marijuana was available because the railroad could give the engineer a drug test, which he flunked.

In the same edition of The Post, however, reporter Paul Valentine tells the true story about the worth of these drug tests. He states that experts reported finding traces of drugs in the blood and urine of both the engineer and the brakeman, but "said they could not pinpoint when it was ingested or determine whether it impaired the two men."

By this time Post readers must be thoroughly confused: What can these drug tests show? Well, the experts quoted by Valentine are correct: there is no available drug test that can show that an employee is impaired by drugs or that can show exactly when a drug was ingested. And this was an important basis for the Ninth Circuit's decision striking down post-accident testing as an unreasonable search under the Fourth Amendment.

In its criticism of the Ninth Circuit ruling, The Post states that now the only way a railroad engineer can be tested after an accident is if he or she is "reeling drunk or in a drug-induced stupor." Not so. The Ninth Circuit decision permits testing when there is a reasonable suspicion that the test will show evidence that an employee has violated work-place drug or alcohol use rules. Although the court acknowledged that accidents by themselves don't create reasonable grounds for believing that tests will show alcohol or drug impairment by one railroad employee (much less an entire railroad crew), it certainly did not require employees to be reeling drunk or in a stupor before they can be tested.

The Post also supports the use of postaccident tests with the claim that "one or more crew members tested positive in 37 accidents in which the regulation was invoked last year." What The Post failed to mention was that this actually amounted to 42 employees testing positive out of a total of 770 who were tested -- about a 5 1/2 percent positive rate.

But what about the other 94 percent who took the test without positive results? Courts have almost unanimously held that these intrusive tests are a search under our Fourth Amendment law, just as a police officer's coming into our homes is a search. Would we justify searches of all of our homes for illegal drugs or guns with the claim that 5 percent or 6 percent of the time illegal items would be found? The Fourth Amendment says no.

What can employers do to curb alcohol and drug abuse in the work place? They can train supervisors to detect drug and alcohol use on the job, institute employee treatment programs and invest in research that will produce nonintrusive testing methods that can show current drug impairment.

Next time I get on an airplane, I'm going to hope that my pilot has been carefully observed by co-workers and has drug and alcohol abuse education and treatment programs available and that perhaps the airline has administered a truly job-related test to the pilot just before takeoff (such as hand-eye coordination tests, which are currently available). The Post can pin its hopes on drug tests if it wishes. I'd rather try my chances with methods that have something to do with safety, not just blind faith.

-- Elizabeth Symonds

The writer is staff attorney for the American Civil Liberties Union of the National Capital Area.