Fatal train wreck shows risk of hidden illness, dangerous deception

Engineer hid diabetes, serious vision problems from railroad’s doctor

A year-long investigation of a fatal train wreck in New Jersey has revealed a surprising cause that has nothing to do with mechanical failures, fatigued engineers working long hours, or faulty track signals. All of those suspected causes went out the window when it was discovered that one of the train engineers was colorblind and could not discern a red track signal that told him to stop.

The man’s colorblindness was caused by diabetes he had kept secret for nine years, even though he underwent regular examinations by an occupational health physician. The engineer eventually failed a color-perception test administered by the physician, but he was allowed to continue working after passing an alternate test. Because of that oversight, some critics suggest that occupational health physicians are not qualified to measure vision deficiencies.

The 59-year-old train engineer, John J. DeCurtis, was killed at 8:40 a.m., Feb. 9, 1996, when his eastbound New Jersey Transit (NJT) commuter train ran a red stop signal and collided nearly head-on with a westbound NJT commuter train near Secaucus, NJ. The engineer on the other train also died. One of the 400 passengers from both trains died, and 158 were injured.

A few seconds’ difference in the timing of the accident could have resulted in even greater tragedy. DeCurtis’ train hit only the locomotive of the other train at a track-switching point, but a few seconds later it would have slammed into the passenger cars broadside.

The National Transportation Safety Board (NTSB) in Washington, DC, investigated the accident and determined that the track signal properly showed three red vertical lights to the train driven by DeCurtis, meaning he had to stop. Investigators determined DeCurtis most likely saw the lights but misinterpreted their colors. In fact, the NTSB concluded that the engineer’s colorblindness led him to interpret the signal to mean he should accelerate, which he did just before the collision. Unlike an automobile traffic signal, the location of the lights on a train signal gives almost no clue to their colors, shifting only slightly when the colors change.

The engineer’s possible fatigue was considered but ruled out as a cause of the accident. He was working overtime after a 12-hour split shift and had rested only a few hours the night before. He had 40 years experience as a railroad employee, including 34 years as a locomotive engineer.

Clinicians must understand job demands

Mitchell Garber, MD, MPH, MS, ME, is the NTSB’s first full-time medical director, having joined the board in November 1996. Garber has extensive experience as an occupational medicine specialist, including two years as a flight surgeon with the U.S. Air Force. He tells Occupational Health Management that the accident underscores the need for occupational health professionals to look beyond the daily examinations they provide workers.

"With any occupational medicine role, it is important that the person performing the occupational exam have a clear understanding of what the job entails and what it means when the person fails an exam," Garber says. "It is crucial that the physician understand what the person’s job is and why the test is important."

Because the NTSB’s report is a formal version of the board’s findings, Garber and other members of the NTSB are reluctant to criticize the health care provided to the engineer by the occupational medicine physician beyond referring to the board’s overall conclusion that the doctor’s error in providing the vision exam was a significant contributor to the accident. Garber notes, however, that there are unanswered questions concerning how the health care system failed to screen out a worker who was unfit for duty in such a safety-sensitive job.

A troubling aspect of the accident, especially for occupational health professionals, is that the engineer knew he was going blind from diabetes but was able to hide his problem from the doctor hired by the railroad. He had been treated for diabetes for nine years, but he saw a personal physician and paid her out of pocket rather than filing insurance claims. DeCurtis even had eye surgery twice and avoided insurance claims.

The personal physician treating his diabetes knew he worked for the railroad but did not notify the employer of his vision problems. The occupational health physician hired by the railroad, on the other hand, did not even know that DeCurtis was diabetic. In the annual physical examinations required by federal regulations, the engineer apparently lied to the doctor and said he was not diabetic, not taking any prescription medication, and not under the care of another physician.

Part of the annual examinations for NJT engineers include testing eyesight with the Dvorine pseudoisochromatic plates, known as the PIP test, a series of 14 circular plates of polka dots with numbers in slightly contrasting colors. When DeCurtis was tested in February 1994, he missed two of the 14. When he was last tested in February 1995, he missed six of the plates. That is considered a sign of rapid vision loss, but the occupational health physician allowed DeCurtis to take another type of vision test, the Dvorine nomenclature test.

That test is not considered a reliable measure of colorblindness, especially if the patient already has failed the PIP test. But because DeCurtis passed the second test, the doctor allowed him to continue working. The engineer was due for another physical examination the same month as the accident, so it is likely that his vision had deteriorated even further. The NTSB’s investigation revealed he had diabetic retinopathy affecting the macular region. That would directly hinder the central field of vision as well as color discernment.

The NJT reports that all 250 of its engineers have since passed vision tests. The railroad also has informed its occupational health physicians that an engineer who fails "any element" of a test to distinguish colors must be sent to an ophthalmologist for further assessment.

The NTSB concluded the engineer "intentionally concealed" his illness, says Burton Simon, a senior human performance investigator with the NTSB. Simon was instrumental in tracking down the man’s medical records and piecing together the deception. The engineer’s dishonesty was a primary cause of the fatal accident, Simon says, but he points out that the investigation raises troubling questions about the occupational health system.

DeCurtis deliberately and repeatedly lied to Ralph D’Agostino, MD, the occupational health specialist in Harrison, NJ, who was one of many physicians with contracts to provide annual physical examinations to NJT employees. At each examination, the employee was required to fill out a medical history form that includes 40 yes-and-no questions. Some of these questions include the following:

• Have you been examined or treated by any physician or other practitioner?

• Are you taking any medicine?

• Have you had diabetes?

• Have you taken any medication in the past 60 days? If so, what and why?

DeCurtis answered "no" to each of those questions since 1985. (For more on DeCurtis’ medical conditions and other findings of the NTSB investigation, see related story, above.)

The NTSB’s investigation was not intended as a criticism of D’Agostino or occupational medicine, but Simon notes that the NJT medical records show glucose was discovered in the engineer’s urine nine years before the accident. This was 10 years after the engineer knew he had diabetes. The records show DeCurtis was allowed to see a personal physician for treatment and returned two weeks later with a modified diet and a clean urine test. The NJT physician cleared him for duty at that time.

Since that one suspicious urine test, there apparently was no clinical finding that should have alerted D’Agostino. Simon surmises that the engineer’s diabetes treatment and medications were working well enough to mask the symptoms and keep glucose out of his urine.

"The question is whether it is appropriate or necessary for a physician to go back so many years and see that there was a diabetes diagnosis, especially when the patient is denying it and there is no obvious reason to suspect it," Simon says. "How far back does a physician go to determine if a previous problem was not carried forward? Or is the physician’s primary concern the patient’s physical condition at the time of the exam?"

D’Agostino, the occupational medicine physician who conducted the last exam on DeCurtis, did not respond to repeated requests for an interview by OHM.

Physicians need guidance with vision tests

Members of the NTSB tell OHM that the Secaucus train wreck report is not necessarily an indictment of the policy of having occupational health professionals provide vision exams instead of ophthalmologists. Rather, they say the problem is that federal regulations and industry guidelines have not provided enough guidance to physicians concerning how to conduct the examinations. Guidelines for the rail industry, as in others, require vision testing but are not specific about exactly what tests should be used, how the testing should be conducted, and what constitutes failure. That is why the NTSB report recommends that the rail industry provide more specific guidance on vision testing, explains Robert Lauby, PE, chief of the NTSB’s railroad division.

"There is a need for specific guidelines in this area so that any physician can do this test and get consistent results that are comparable no matter who performed the test," Lauby says. "We don’t have that right now."

Garber, the NTSB medical director, points out that the NTSB was definite in determining that the occupational medicine physician who last examined DeCurtis and certified him for duty made a mistake. Even though the federal guidelines allowed D’Agostino some latitude in deciding that the engineer could distinguish colors even if he failed the appropriate vision test, Garber says he clearly was in error by basing that decision on the Dvorine nomenclature test. Proper training in the use of the vision tests might have made that clear, he says, as would a clear statement from the railroad that failing the PIP test should result in automatic disqualification from duty.

Lauby suggests that occupational health professionals can help avoid similar tragedies by remembering that any simple exam, no matter how routine and quick, can be a matter of life and death.

"The message we would like to leave occupational medicine physicians with is that, like many workers, this engineer had a very critical job. He was the only line of defense between a tragic accident and safe operation of this train," Lauby explains. "It’s important to know how critical a simple test like the Dvorine PIP test is for an individual to do a job safely."