OPRRMS wrote:

I'm not sure how it relates to Chatsworth, since there's really no similarity between the two incidents.

Au contraire. There’s a lot of similarities between the wrecks at Secaucus, NJ and Chatsworth, CA:

(1.) Both trains were running with lonesome cab. i.e., just one person in the operating cab of the train.

(2.) Both engineers had diabetes. The one in NJ was suffering from color blindness because of his illness. While it’s unproven that the Metrolink engineer was suffering from colorblindness, it is on the record that he had diabetes.

(3.) Both engineers worked under screwbally split shifts, where adequate rest was probably not part of the equation.

(4.) In both wrecks, the ONLY thing preventing a catastrophic collision with other rail traffic was ONE man, the engineer, diligently interpreting signal lights correctly and running his train in accordance with those signal indications. Both collisions took place in territory that did not have safety back up systems such as Automatic Cab Signals or Automatic Train Stop.

Here are a couple of reference articles:

NJ Transit wreck: was it engineer fatigue?
February 9, 1996 New Jersey Transit accident
Railway Age , March, 1996 by William C. Vantuono

The wreck of a New Jersey Transit train on Feb. 9 that killed three and injured 162 has raised questions about the practice of split shifts for engineers and about whether the 89-year-old Hours of Service Act should be modified.
At 8:41 a.m. on Feb. 9, an eastbound NJ Transit train traveling toward Hoboken Terminal ran past a stop signal, fouling the interlocking where NJ Transit's Bergen and Main lines converge, and struck a westbound train. The impact killed both engineers, and one passenger on the eastbound train. It was the first fatal accident on a New Jersey commuter train since 1958.
Early speculation focused on human error, with federal investigators reporting that John DeCurtis, the eastbound's engineer (a 40-year veteran), had been suspended a total of 115 days on four occasions in the past 13 years (twice for running a stop signal, once for overshooting a station, and once for derailing a work train). It was also revealed that, according to an eyesight test taken in December 1995, DeCurtis was classified as having a minor (but normal for his age, 59) color deficiency. And, there were charges of signal equipment malfunction from NJ Transit union officials.
The National Transportation Safety Board said that signals on the line were found to be in proper working order. It did point out, however, that NJ Transit does not employ cab signals and automatic train stop--which may have prevented the accident--on that portion of its system. Such protection devices are employed only on lines (those shared with Amtrak) that were so-equipped prior to NJ Transit assuming operations in 1983. NJ Transit Executive Director Shirley DeLibero said that NJ Transit planned to equip all its lines with cab signals and automatic train stop by the end of the decade.
Soon after the wreck, investigators said they were looking more at the possibility of engineer fatigue than at DeCurtis' safety record. DeCurtis was on the last leg of a 14.5-hour overnight split shift during which he had gotten, according to investigators, four and 3/4 hours of rest.
NJ Transit assigns split shifts on a voluntary basis to 62 of its 279 engineers. Split shifts, which normally encompass more than 12 hours, are popular with some veteran engineers, since they are paid for rest time as well as active duty. Split shifts also allow the railroad to save on labor expenses by hiring fewer engineers and paying less overtime.
On the night before the wreck, DeCurtis, who normally worked a 67.5-hour week with 22.5 hours of paid rest time, reported for duty at 6:11 p.m. According to records, he worked until 12:58 a.m., then slept in a railroad coach until 5:44 a.m.
DeCurtis was scheduled to finish his shift at 7:28 a.m., but instead was asked to make an extra eastbound run. NJ Transit officials said that the practice of requiring engineers to make extra runs is routine, and is known by all engineers prior to their employment.
Investigators have raised questions that could, in the long run, affect railroad work scheduling practices nationwide. The Hours of Service Act, instituted in 1907 and last updated in 1968, limits the amount of consecutive hours (12) an engineer can work and sets the minimum amount of rest (at least four hours between shifts) for split shifts. It does not, however, specify the type of facilities railroads must provide for rest periods. Although NJ Transit, in compliance With federal regulations, provides facilities where engineers working split shifts can relax, it does not provide sleeping quarters.
"It looks like everything NJ Transit was doing complied with federal regulations, but these were written in the early part of the century before we understood how the body operates," Steve Mardon, editor of Shiftwork Alert, a Boston-based newsletter that covers workplace scheduling, told the Newark Star Ledger on Feb. 13. "It's particularly difficult to be awake during late night and early morning hours, when your body is just crying out for sleep."
The Federal Railroad Administration and the NTSB also questioned whether NJ Transit engineers working split shifts are able to get adequate rest under the conditions available to them. In response, NJ Transit said that the split shift that DeCurtis worked would be eliminated; it also said it would consider eliminating split shifts entirely.
Engineer fatigue is something that FRA has been looking into. In the wake of the NJ Transit wreck and a worse crash on Feb. 16 involving a MARC commuter train and Amtrak's Capitol Limited in which 11 died (and in which engineer fatigue is being considered as a possible factor), FRA has ordered all U.S. passenger railroads to submit detailed safety plans.

~~~~~~~~~~~~~~~~~~~~~~~~

March 26, 1997
EYE PROBLEM CITED IN '96 TRAIN CRASH

By MATTHEW L. WALD

A New Jersey Transit train engineer who ran a red signal in Secaucus a year ago and caused a collision that killed him and two other people was going blind from diabetes but had kept his condition secret for nine years, the National Transportation Safety Board said today.
A doctor hired by the railroad to conduct the required annual physical examination had given his approval for the engineer to continue working, even though he failed a color vision test, the investigators said.
The doctor, who is an occupational medicine specialist and not an ophthalmologist, did not know that the engineer, John J. DeCurtis, was a diabetic, according to the board, because Mr. DeCurtis always answered no to the annual questions about whether he had diabetes, was taking any prescription medication or was under another doctor's care. Mr. DeCurtis had had eye surgery twice, but apparently paid for it out of pocket rather than filing insurance claims with New Jersey Transit, investigators said.
The investigators had originally focused on the fact that Mr. DeCurtis was working overtime after a 12-hour ''split shift,'' having rested for a few hours overnight, but today they said that there was no evidence that fatigue was a factor. In fact, they said, there is evidence that he was attentive to the signal -- three red lights in a vertical row -- but unable to see what color they were.
Unlike a city traffic signal, in which the location of the light is a clue to its color, the signal that Mr. DeCurtis went past moves only slightly as it changes color, almost certainly too little to be a cue to an operator, investigators said.
New Jersey Transit had initially told the safety board that the doctor it hired to examine its engineers, Ralph S. D'Agostino of Harrison, N.J., had acted against the railroad's instructions on what to do if a worker failed the test. But on Monday, more than 13 months after the accident and a day before the conclusion of the investigation, New Jersey Transit told the board that it had not provided any such instructions.
Since the accident, on Feb. 9, 1996, the railroad has sent all 250 of its engineers for color vision testing, safety board officials said, and all 250 passed.
Investigators also found that surviving train crew members in the accident failed to use the public address system to tell the 400 passengers what was happening, leading some of the passengers to wander from the trains onto the tracks. The railroad gives adequate training to new personnel but no refresher courses, the investigators said. One railroad worker in the crash had not been trained in nine years.
One passenger and the other train's engineer were killed, and 158 passengers were injured.
New Jersey Transit, like many railroads and other agencies, uses a color-perception test that consists of 14 circular plates of polka dots, in which a subject must discern numbers formed by dots of a slightly contrasting color. In February 1994, Mr. DeCurtis missed 2 of the 14; in February 1995, he missed 6, a sign of rapid vision loss, said Dr. Mitchell A. Garber of the safety board.
Dr. Garber said that Mr. DeCurtis had had diabetic retinopathy, a progressive condition. He was due for another examination the month of the accident.
Investigators said Mr. DeCurtis had undergone laser surgery several times. Dr. Leonard J. Press, an optometric physician in Fair Lawn, N.J., who had no direct knowledge of Mr. DeCurtis's condition, said laser surgery can slow the progress of the disease but cannot reverse it.
He added that Mr. DeCurtis's perception of the signal could also have been affected by loss of visual acuity, another symptom of diabetic retinopathy of the macula. The signal was three red lights, stacked on each other, meaning stop; he evidently perceived the middle light as yellow, which would have allowed him to proceed, Dr. Press said.
The color-perception test was developed by a Baltimore ophthalmologist, Israel Dvorine, and its protocols specify that someone who misses 6 of 14 should not work in a field where color discernment is critical.
After Mr. DeCurtis failed the first test, Dr. D'Agostino gave him a second test, investigators said, which he passed. But the second test says in its introduction that it is not meant as a test of color-blindness, according to Dr. Garber. It reveals only whether the patient can look at one color at a time and identify it.
''We have a situation where an engineer apparently chooses to conceal his medical problems,'' said Robert Lauby, chief of the safety board's railroad division. The doctor, he said, was ''the last line of defense before the engineer gets back in the locomotive,'' but the doctor used the wrong follow-up test.
A woman who answered the telephone at Dr. D'Agostino's office said that he was not expected back until Thursday and that she had no way to reach him.
Mr. DeCurtis had a personal physician who was treating him for diabetes, had referred him to an ophthalmologist, and knew he was a railroad engineer. That doctor, Dorothy Wludyka did not tell the railroad about his condition.
In the hearing today, George W. Black, one of the board members, asked the staff, ''You're telling me she let him in this condition operate a train with several hundred people?''
Burt Simon, a human factors specialist at the board, said: ''She did know. We discussed it at length. The confidentiality issue was the situation.''
Dr. Wludyka would not comment today on Mr. DeCurtis's condition except to say that she had referred him to an ophthalmologist in September 1994. ''I made sure that I referred him to the doctors I felt he needed, so he could get proper care,'' she said. ''The whole thing was shocking to us. I was shocked he was in an accident.''
The surgeon who performed several laser surgeries on Mr. DeCurtis, Dr. Mark W. Koser, also knew his profession but never notified the railroad, according to the Safety Board. He declined to speak to a reporter today.
The Federal Railroad Administration requires the railroads to check that their engineers have ''the ability to recognize and distinguish between the colors of the signals,'' but it does not specify a test. One recommendation from the Safety Board today was that the Federal Railroad Administration develop a specific test.
New Jersey Transit, in a statement this afternoon, said that its annual physicals surpass the Federal requirements of physicals every three years. It has also written new instructions to physicians, the railroad said, and engineers who fail ''any element of a test'' to distinguish colors are to be sent to an ophthalmologist.
But the transit agency said it did not agree that the wrong test had been used on Mr. DeCurtis.
''We would like to review the National Transportation Safety Board's records on which they based their findings in light of their access to information N.J. Transit was unable to obtain,'' the statement said. ''Until we've had an opportunity to review their records on this test, it is N.J. Transit's position that the appropriate Federal Railroad Administration-mandated examinations were given.''
Stephen J. Klejst, the assistant general manager for rail safety and training, said that the railroad had not yet seen the medical records from Mr. DeCurtis's private physician, and wanted to study whether the second test that was given by the physician it hired was in fact inappropriate.
He said that that physician, Dr. D'Agostino, remained on a list of doctors that its engineers could use for their annual physicals.
The railroad also said it was moving ahead with automatic systems to stop a train that runs a signal.