The impact of smoking on failure rates, postoperative infection and marginal bone loss of dental implants.

Abstract

Electronic searches were undertaken in PubMed/Medline, Web of Science, Cochrane Oral Health Group Trials Register as well as hand searching. Clinical human studies, either randomised or not, providing outcome data for dental implant failure in smokers and non-smokers in any group of patients with no follow-up restrictions. Patients smoking a minimum of one cigarette a day were classified as smokers and implant failure was considered as the complete loss of the implant.Data extraction and synthesisThree authors independently participated in the inclusion criteria and disagreements were resolved by discussion. Quality assessment of the studies was performed using a scale to appraise observational studies. From the selected studies, data were extracted when available which included year of publication, study design, country, setting, number of patients, type of smokers, age, follow-up, days of antibiotic prophylaxis, failure, postoperative infection, marginal bone loss, implant surface, dental arch receiving implants, type of prosthetic rehabilitation and opposing dentition. For the meta-analysis, implant failure and postoperative infection were evaluated as dichotomous outcomes, while weighted mean differences were used for marginal bone loss as a continuous outcome. The statistical unit was the implant. Heterogeneity was calculated using I(2) test. From 1432 records identified, 107 publications were included in the review for the qualitative and quantitative synthesis. One hundred and four studies compared implant failure rates. From the total studies, the authors judged 85 to be of high quality and 22 of moderate quality. The meta-analysis for failure rate resulted in a statistically significant overall result of risk ratio (RR) of 2.23 (95% CI 1.96-2.53) heterogeneity of I(2): 51%. The meta-analysis of four studies evaluating the risk of postoperative infections in smokers presented a statistically significant result with an RR: 2.01, 95% CI 1.09-3.72, heterogeneity I(2): 0%, and the marginal bone loss had an overall statistically significant difference of MD 0.32, 95% CI 0,21-043; heterogeneity in this case was I(2): 95%. The insertion of implants in smokers affected the failure rates, the risk of postoperative infections, as well as the marginal bone loss. The results should be interpreted with caution due to the presence of uncontrolled confounding factors in the included studies.