In reply

We would like to thank Dr Üçok for his careful reading of our article. The omission of the article he cites in the letter demonstrates one of the limitations of systematic reviews. Computer database searches are used to identify articles for inclusion in these reviews. Because we were evaluating lingual nerve and third molars, these words, as well as the other combinations outlined in our review, were linked together for the search. If the individual indexing an article for the computer database does not include both terms, the article will not be found when searching for the combination of words. In fact, we repeated the search using the combinations of words and did not find the article. The article is cited when morbidity and third molars are used in combination. Because the combination of morbidity and third molars produced so many citations of unrelated articles, we decided not to use the combination. It is not surprising that the article was not found through use of our search methods. The title of the manuscript does not include comments related to sensory impairment, and the abstract does not report the incidence of sensory impairment associated with the 2 surgical techniques used to remove the third molars. In the article, the only mention of sensory function is in the last part of the last sentence in “Results,” which states, “… and no one experienced sensory impairment of the inferior alveolar or lingual nerves.” Although the article was missed, the results from the 20 patients who participated in the study do not alter the outcome of our systematic review. Temporary numbness of the lingual nerve caused by using the lingual split technique would be 9.54% instead of 9.63%, and permanent numbness would remain the same, 0.14%. Temporary numbness of the lingual nerve caused by using the buccal approach with a retractor would be 6.30% instead of 6.35%, and permanent numbness would be 0.62% instead of 0.63%. We certainly apologize for missing Dr Üçok's article and thank him for carefully reading our review and letting us know about our omission. Even though computer database searches sometimes miss an article, it is the best method currently available to systematically review the literature and find evidence to support clinical decisions. Dr Üçok's letter also gave us an opportunity to find 2 recent articles that were published after the submission of our systematic review. Gargallo-Albiol et al reported the results of a randomized controlled trial comparing protection of the lingual nerve with and without a lingual retractor. This study meets our inclusion criteria for review and included the extraction of 300 third molars. These investigators did not find a significant difference in temporary numbness to the lingual nerve with and without the lingual retractor and no patients had permanent numbness. Adding these data to our results did not change the results of our review. The investigators found that 3 patients had numbness with the retractor and 1 patient had numbness without the retractor. A second study reported in this Journal found greater lingual nerve damage after third molar removal with a lingual retractor than without the retractor. This was a prospective clinical study involving consecutive patients having third molars removed by various surgeons. The decision to use the lingual retractor was based on the surgeon's preference. This study does not meet our inclusion criteria and would not have been considered for detailed analysis. These 2 articles demonstrate that systematic reviews must not be based on a particular outcome but on the validity of the study. The systematic review establishes inclusion and exclusion criteria to identify the “best” evidence available. Sometimes the studies available for review are inadequate to make a definite conclusion. The current evidence for the use of a lingual retractor when removing lower third molars indicates that there is greater tendency for temporary numbness of the lingual nerve. It is unclear whether the retractor protects the lingual nerve from permanent numbness, because permanent numbness is very low for extraction done with and without the lingual retractor.