The purpose of this study was to systematically review and perform a comprehensive meta-analysis of the current literature to answer the following question: among patients receiving dental implants, does the use of antibiotics, when compared with a control group, reduce the frequency of implant failure and postoperative infection? A manual and electronic PubMed search of the literature was made to identify randomized controlled trials (RCTs) on the efficacy of antibiotics compared with a control group (not receiving antibiotics or receiving placebo). Four RCTs were included in the final review. These four RCTs grouped a total of 2063 implants and a total of 1002 patients. Antibiotic use significantly lowered the implant failure rate ( P = 0.003), with an odds ratio of 0.331, implying that antibiotic treatment reduced the odds of failure by 66.9%. The number needed to treat (NNT) to prevent one patient from having an implant failure was 48 (95% confidence interval 31–109). In contrast, antibiotic use did not significantly reduce the incidence of postoperative infection ( P = 0.754). Based on the results of this meta-analysis, and pending further research in the field, it can be concluded that there is evidence in favour of systematic antibiotic use in patients receiving dental implants, since such treatment significantly reduces implant failure. In contrast, antibiotic use does not exert a significant preventive effect against postoperative infection. Our recommendations for future research focus on the performance of large-scale RCTs to identify the best choice of antibiotic, timing of administration, and dose. Increased effort is also required to reach consensus and define the most effective antibiotic treatment protocol for patients who are allergic to beta-lactams and for those who are not.
Dental implants are an effective, safe and predictable solution for those who have lost a tooth or teeth due to dental caries, periodontal disease, injury, or other reasons. The biological complications of dental implants are classified as early or late. Early failure is defined as failure occurring up until connection of the implant abutment, and proves easier to diagnose due to the lack of osseointegration. It is believed that a certain number of early dental implant losses are due to bacterial contamination at implant insertion.
It is still debated whether early dental implant failure and postoperative infection can be reduced by antibiotic prophylaxis. While it is important to minimize implant failure, there are concerns associated with the widespread use of antibiotics, since adverse events may occur. A number of authors have reported that systemic antibiotic use does not reduce the implant failure rate. In contrast, other investigators have found such treatment to reduce implant failure.
The role of antibiotic use in the setting of implant placement and its effects on implant failure and postoperative infection are unclear. Antibiotic over-prescription has a negative impact on general health and the economy. Proper selection of the antibiotic regimen in clinical practice is therefore very important. The use of antibiotics in implant surgery in healthy patients and its correlation to failure and success rates remain poorly documented in the literature. It is widely agreed by health authorities that overall antibiotic use should be reduced, and that antibiotics should be prescribed for life-threatening infections to reduce the emergence of resistant bacterial strains. The decision to administer prophylactic antibiotics is based on various factors, including the general condition of the patient, the surgical site and extension, the preoperative diagnosis, and even surgeon preferences. Penicillin is considered to be active against aerobic and anaerobic microorganisms commonly associated with oral infections, and the results of different studies support the use of short-term penicillin prophylaxis in oral procedures.
The purpose of the present study was to systematically review the current literature and to perform a comprehensive meta-analysis of this literature to answer the following question: among patients receiving dental implants, does the use of antibiotics, when compared with a control group, reduce the frequency of implant failure and postoperative infection?
Materials and methods
This meta-analysis complies with the QUOROM (Quality of Reporting of Meta-analyses) statement.
Search strategy for the identification of studies
The PubMed (MEDLINE) database of the United States National Library of Medicine was used for a literature search of articles published up until December 2012. The following search terms were used in different combinations: ‘dental implants’, ‘antimicrobial agents’, ‘antibiotics’, ‘prophylactic antibiotics’, and ‘infection’. The search was completed with a review of the references of the selected articles in order to identify additional studies not found in the initial literature search.
In addition, a manual search (up until December 2012) was made of the following journals: British Journal of Oral and Maxillofacial Surgery, Clinical Implant Dentistry and Related Research, Clinical Oral Investigations, Clinical Oral Implants Research, European Journal of Oral Implantology, Implant Dentistry, International Journal of Oral and Maxillofacial Implants, International Journal of Oral and Maxillofacial Surgery, Journal of Clinical Periodontology, Journal of Oral Implantology, Journal of Oral and Maxillofacial Surgery, Journal of Periodontology, Medicina Oral, Patología Oral y Cirugía Bucal, and Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology, and Endodontology.
Study selection criteria
Before starting the study, a series of inclusion and exclusion criteria were established using a protocol. Full-text articles chosen were assessed for the following inclusion criteria: (1) patients subjected to dental implant surgery; (2) randomized controlled trials (RCTs); (3) the presence of a control group (not receiving antibiotics or receiving placebo); (4) systemic antibiotic treatment, with specification of the type of antibiotic, the administered dose, and the duration of treatment; and (5) specification of the implant failure and postoperative infection rates. Antibiotic administration was defined as presence/absence, independently of the timing of administration (preoperatively, postoperatively, or both). Studies were excluded if randomization was not performed, if growth factors were used, if no control group was included, if antibiotic administration was not carried out via a systemic route, or if bone augmentation procedures were required concomitant to implant placement. Case series, retrospective studies, and articles published as abstracts only were excluded. No restrictions were placed on the year or language of publication.
Authors were contacted for clarification of missing information when necessary. In cases of more than one publication on the same patient group for the same follow-up, only the study nearest to the objectives of this review or with the largest sample was included. All articles selected from the electronic and manual searches were independently assessed by the first and second authors of this study, in accordance with the established inclusion criteria. Any disagreements between the reviewing authors were resolved by consensus, or by consulting the last signing author of this study.
Quality and risk of bias assessment
Two reviewers independently and in duplicate evaluated the quality of the included RCTs as part of the data extraction process.
The recommendations of the Cochrane Handbook for Systematic Reviews of Interventions 5.0.0 were followed. Four main quality criteria were examined: (1) concealment of allocation; (2) assessor blinding; (3) patient blinding; and (4) compliance with follow-up (withdrawals in the case of a clear explanation for removals and dropouts in each treatment group).
After analyzing this information for each of the RCTs, the publications were grouped into the following categories: (A) low risk of bias (possible bias not seriously affecting the results) if all the criteria were met; (B) high risk of bias (possible bias, seriously weakening the reliability of the results) if one or more criteria were not met.
The primary outcome variable was implant failure and the secondary outcome variable was postoperative infection. The meta-analysis was based on the inverse variance calculation method of DerSimonian and Laird, taking the odds ratio (OR) as the measure of effect. Results were obtained for a fixed effects model.
The OR estimations are accompanied by the corresponding 95% confidence interval (95% CI), standard error, and P -value of the factor null effect contrast (OR = 1) for solution of the meta-analysis, including the QA association statistic. The Forest plot shows the OR value and corresponding confidence interval for the RCTs and the global cluster value. The relative size of each symbol reflects the weight attributed to each study, on the basis of sample size. A logarithmic scale was used to visualize symmetrical intervals.
A funnel plot was used to assess publication bias. The QH heterogeneity statistic and corresponding P -value for the χ 2 test are provided. The number needed to treat (NNT), calculated as the reciprocal of absolute risk reduction, was defined as the number of implants that must be placed with antibiotic use to prevent one implant failure. A significance level of 5% was established in the analyses ( α = 0.05).
Study selection and description
All articles included in the review were RCTs and were published in English. Four publications were excluded on applying the defined study criteria: Dent et al. failed to specify the number of included patients, mention being limited to the number of implants involved. In relation to the administered antibiotics, this author did not describe the dose or the duration of treatment. Laskin et al. conducted a study across 32 research centres, and each professional was free to chose the type of antibiotic for use (penicillins or penicillin derivatives, cephalosporins, erythromycin, and other non-specified antibiotics). These authors likewise did not describe the dose or the duration of treatment. The study published by Anitua et al. was excluded because the implants were humidified with plasma rich in growth factors before implantation, and Karaky et al. failed to include a control group in their study.
Finally, four RCTs fulfilled the inclusion criteria and were thus selected for inclusion in the meta-analysis ( Fig. 1 ). Table 1 summarizes the reviewed articles, while Table 2 describes the implant failure and postoperative infection criteria of each included RCT.
|Author, year||Antibiotic treatment and timing of administration||Gender (M/F)||Mean age (years)||Number of patients||Number of implants|
|Esposito et al.||Amoxicillin 2 g given 1 h preoperatively||80/78||47.8||158||341|
|No antibiotic (placebo)||62/96||47.9||158||355|
|Abu Ta’a et al.||Amoxicillin 1 g at 1 h preoperatively and 2 g a day for 2 days postoperatively||23/17||60||40||128|
|No antibiotic (without placebo)||20/20||57||40||119|
|Esposito et al.||Amoxicillin 2 g given 1 h preoperatively||114/138||49.1||252||489|
|No antibiotic (placebo)||122/132||47.6||254||483|
|Caiazzo et al.||Amoxicillin 2 g at 1 h before surgery||13/12||52||25||35|
|Amoxicillin 2 g at 1 h before surgery and 2 g a day for 7 days following surgery||12/13||45||25||36|
|Amoxicillin 2 g a day started after surgery and continued for 1 week after surgery||7/18||42||25||48|
|No antibiotic (without placebo)||10/15||43||25||29|
|Author, year||Definition of implant failure||Criteria of infection|
|Esposito et al.||Implant mobility of each implant measured manually and/or any infection dictating implant removal||Any biological complications such as wound dehiscence, suppuration, fistula, abscess, and osteomyelitis, etc.|
|Abu Ta’a et al.||Presence of signs of infection and/or radiographic peri-implant radiotransparencies that could not respond to a course of antibiotics and/or judged a failure after performing explorative flap surgery by an experienced periodontist||Purulent drainage (pus) or fistula together with pain or tenderness, localized swelling, redness, and heat or fever (>38 °C) a|
|Esposito et al.||Implant mobility of each implant measured manually and/or any infection dictating implant removal||Any biological complications such as wound dehiscence, suppuration, fistula, abscess, osteomyelitis, etc.|
|Caiazzo et al.||Mechanical implant removal because of lack of osseointegration||Internal and external oedema, internal and external erythema, pain, heat, and exudate|
Due to the small number of studies included in the meta-analysis, the evaluation of publication bias was not considered relevant. Regarding the need for bone augmentation procedures concomitant to implant placement, we directly obtained further information from Abu-Ta’a et al. and Caiazzo et al., who confirmed that the need for bone augmentation constituted an exclusion factor. This information was already contained in the other two included publications.
Primary outcome variable: implant failure
The four RCTs comprised a total of 2063 implants: 1077 in the antibiotic treatment group and 986 in the control group. With these data, two-tailed χ 2 testing yielded a power of 90.3% in defining failure proportions of 0.01 and 0.03 in the groups as being significantly different, with a 95% confidence level.
In order to analyze the implant failure rate, we first used a homogeneity test to show that the four studies were suitable for obtaining a combined effect measure (QH = 3.67, P = 0.298). On examining the implant failure rate individually in each RCT, only one study was seen to show a statistically significant difference between the antibiotic group and the control group.
However, the result of the χ 2 test applied to the four included RCTs showed antibiotic treatment to reduce the risk of implant failure (OR = 0.331, 95% CI 0.157–0.696), this result being statistically significant ( P = 0.003, QA = 8.49). The NNT to prevent one patient having an implant failure is 48 (95% CI 31–109). It can be concluded that antibiotic administration lowers the odds of implant failure by 66.9%. The Forest plot is shown in Fig. 2 .