Regarding implants' survival, three implants were lost in the US-while only one implant was lost in the SL-group (P = .6085; Fisher's exact test). Nevertheless, the ultrashort implants were associated with a tripling of the failure rate and uncertainty where the true failure rate is uncertain (relative risk 3.0; confidence interval 0.3-26.8).
Such a phrasing, seem like two independent contradict statements at first but they are just the same thing twice after reading the main text. The RR numbers are not even showed in the main text.