Yep, but if death was due to the combined effects of both bacterial infection and fever, we'd expect to see a reduction in mortality when fever was treated compared to when it was untreated, which the linked study didn't find.