Methods: We performed a computerized literature search for all studies using PubMed database (in English) and Japana Centra Revuo Medicina (in Japanese). The main search terms were "smoking", "smoke" and "tobacco" in combination with "implant" and studies that published between 1993 and 2003 were selected. All statistical analyses for a meta-analysis were conducted using Excel software. The odds ratio (OR) were considered statistically significant when the lower 95% confidence interval (95% CI) exceeded 1.0.
Results: Among 143 studies identified and chosen for detailed review, 22 were appropriate for inclusion in our meta-analysis. For smokers compared with non-smokers, OR for osseointegrated implant failure was significantly elevated (OR 2.22, 95% CI, 1.50-3.27). Six studies were appropriate to examine the association between smoking and implant failure placed in the maxillary arch or the mandible. The OR for implant failure placed in the maxillary arch was significantly elevated (OR 2.22, 95% CI, 1.64-3.01), however, the OR for implant failure placed in the mandible did not demonstrated an increased risk by smoking (OR 1.34, 95% CI, 0.96-1.88).
Conclusion: Our meta-analysis revealed that a significant association between smoking and the risk of osseointegrated implant failure, with a higher failure rate in maxillary arch accounting for the majority of the difference.