Therac-25 Case Study

Surname 1
Student’s Name
Professor’s Name
Course
Date
Therac-25 Case Study
The widely cited Human-Computer Interface accidents that relates to safety-critical
systems are the recorded misfortunes of the massive radiation overdoses of six patients by the
Therac-25 machine between 1985 and 1987. The computerized radiotherapy machine
administered a hundred times the standard therapeutic dose. The device was an updated version
of Therac-6 and Therac-20, but its incorrect system architecture resulted in software failure as
the engineers removed both the manual controls and hardware interlocks (Holzmann 16).
Patients were reporting burns that ultimately resulted in deaths and injuries. However, the
operators are not to blame for these accidents are they acted ethically.
BACKGROUND INFORMATION THE OPERATOR
Machine operators were the least in control of the Therac-25 being used at the facility. The CMC
manufactured the machine; the FDA approved it; the hospital accepted to use it in their facility
(Rae 3). An operator works for the hospital and does not play a part in the decision for the
machine to be used at the hospital. The hospital employs an operator to complete whatever tasks
the administration asks them to do; they don’t have the liberty to do otherwise. In case of a
lawsuit of anything legal, they are not at fault because they have an authority that tells them to
complete the scheduled treatments within specific days.

Surname 2
For the Therac-25 case, the operator controlled the radiotherapy instrument using the
DEC VT100 terminal (Holzmann 18). During treatment, the operator positioned patients on the
therapy table, set the gantry rotation and electric field sizes manually, and attached all the
required accessories to the machine. The operator then left for the VT100 console to enter the
patient’s identity, gantry rotation, electronic field, and the treatment prescription, which included
the dose rate, time and the energy mode (Leveson 9). A patient’s treatment was permitted if the
manually set data compares with that at the console. Equally, the therapy is aborted if a
mismatch occurs, and only proceeds if the error is corrected.
INCIDENT INVOLVING THE OPERATOR
THowever, the operators complained about how long it took to feed in the treatment plan. In
response, the CMC modified the console interface to allow carriage returns to complete the
treatment data entry (Griffith Angela par.2). This modified interface would make the operator
tolerant of error messages, and readily allowing them to resume treatment. The interface did not
distinguish between life-threatening and minor errors but only when significant errors did not
prompt the “proceed” command (Rae 6). Despite the frequent machine errors and frequent
shutdowns, there was no information regarding the likeliness of an overdose.
REMEDIES FOR THE PROBLEM
It is essential that such erroneous malfunctions must be eliminated to avoid injury or deaths.
Nations ought to establish rigorous software quality assurance standards, which will ensure that
manufacturers perform tests to ascertain that their machines are safe. Stress tests and fault
injection can help detect any software’s bug or malfunction during the various testing levels
(Spichkova 309). Besides, the manufactures should report any failure on their devices to an
oversight authority and indicate the possibility of damage or injury to users. Equally important,