Have they calculated the Odds Raito incorrectly? It appears that it's incorrect.
Table 3, row 1:
There were 71 cases, where head injury was judged to be a significant of cause of death. 58/71 did not wear a helmet.
There were 58 controls, where cyclists died as a result of other injuries. 37/58 did not wear a helmet.
Odds ratio is then:
|Head Injury*||Other causes!|
OR = (58/(58+37)) / (13/(13+21) ))
OR = 1.60
Yet the paper states 2.5