The statistics in the story smell funny. Looking at some government stats, accidents account for well over twice as many deaths as homicides for the 15-45 age group. The largest subcategory of these is motor vehicle accidents, which are more than 10 per 100,000 population more than all homicides.
Its hogwash, both studies eliminate any cause of death that is not deemed "pregnancy related" including accidents and illnesses not associated with prganancy but arbitrarily decide to classify murder as "pregnancy related". Easy to come to this kind of conclusion if you care to follow the authors' taste for data manipulation.