Information Technology Reference
In-Depth Information
FIRST AECL INVESTIGATION, JULY-SEPTEMBER 1985
After the Ontario overdose, AECL sent out an engineer to investigate. While the engineer
was unable to reproduce the overdose, he did uncover design problems related to a
microswitch. AECL introduced hardware and software changes to fix the microswitch
problem.
YAKIMA, WASHINGTON, DECEMBER 1985
The next documented overdose accident occurred at Yakima Valley Memorial Hospital.
A woman receiving a series of radiation treatments developed a strange reddening on
her hip after one of the treatments. The inflammation took the form of several paral-
lel stripes. The hospital staff tried to determine the cause of the unusual stripes. They
suspected the pattern could have been caused by the slots in the accelerator's blocking
trays, but these trays had already been discarded by the time the staff began their in-
vestigation. After ruling out other possible causes for the reaction, the staff suspected a
radiation overdose and contacted AECL by letter and by phone.
AECL replied in a letter that neither the Therac-25 nor operator error could have
produced the described damage. Two pages of the letter explained why it was technically
impossible for the Therac-25 to produce an overdose. The letter also claimed that no
similar accidents had been reported.
The patient survived, although the overdose scarred her and left her with a mild
disability.
TYLER, TEXAS, MARCH 1986
A male patient came to the East Texas Cancer Center (ETCC) for the ninth in a series
of radiation treatments for a cancerous tumor on his back. The operator entered the
treatment data into the computer. She noticed that she had typed “X” (for X-ray) instead
of “E” (for electron beam). This was a common mistake, because X-ray treatments are
much more common. Being an experienced operator, she quickly fixed her mistake by
using the up arrow key to move the cursor back to the appropriate field, changing the
“X” to an “E” and moving the cursor back to the bottom of the screen. When the system
displayed “beam ready,” she typed “B” (for beam on). After a few seconds, the Therac-25
shut down. The console screen contained the message “Malfunction 54” and indicated
a “treatment pause,” a low-priority problem. The dose monitor showed that the patient
had received only 6 units of treatment rather than the desired 202 units. The operator
hit the “P” (proceed) key to continue the treatment.
The cancer patient and the operator were in adjoining rooms. Normally a video
camera and intercom would enable the operator to monitor her patients. However, at
the time of the accident neither system was operational.
The patient had received eight prior treatments, so he knew something was wrong as
soon as the ninth treatment began. He was instantly aware of the overdose—he felt as if
someone had poured hot coffee on his back or given him an electric shock. As he tried to
get up from the table, the accelerator delivered its second dose, which hit him in the arm.
The operator became aware of the problem when the patient began pounding on the
door. He had received between 80 and 125 times the prescribed amount of radiation. He
