There has been 1 passenger fatality in the last 13.5 years on the railway, during which time there were an
estimated 20 billion passenger journeys.
That seems to depend on what exact you count as a "passenger", if you assume that someone counts as a passenger from when they enter railway property intending to catch a train to when they leave railway property again afterwards then there seem to be two. The recent stonehaven deailment and an incident at liverpool james street where a girl leant on the train after alighting. I also wonder if your time window was cherry picked to exclude greyrigg.
And with such small numbers good statistics are hard, Out of the 7 people on the stonehaven train 3 died 3 died. The only reason it wasn't a disaster was that the train was virtually empty.
Still noone is disputing in terms of direct accident risk trains are much safer than cars and that the risk of dying in a rail accident is negligable.
The question which is much harder to answer is how the risk (both to the traveller and to their friends/family/colleagues) of catching covid on a train compares to the risk of a car accident. With the constantly changing prevalence of covid in the community, the great uncertainty about the relative importance of different spreading mechanisms and the huge individual factors (how much contact does a person have with vulnerable people) i'm not sure it is one that can be answered in general.