Immediate and Delayed Restoration of Dental Implants in Patients with a History of Periodontitis: A Prospective Evaluation up to 5 Years

Transcription

1 Immediate and Delayed Restoration of Dental Implants in Patients with a History of Periodontitis: A Prospective Evaluation up to 5 Years Jacob Horwitz, DMD 1 /Eli E. Machtei, DMD 2 Purpose: To evaluate the radiographic crestal bone level changes around immediately restored dental implants up to 5 years after insertion in patients with a history of periodontitis. Materials and Methods: Patients previously treated for chronic periodontitis who required a fixed full-arch restoration for the maxilla or mandible or a fixed partial restoration in the esthetic zone were treated. Implant surgery included extraction of hopeless teeth, debridement around remaining adjacent teeth, and implant insertion guided by a surgical stent. A prefabricated screw-retained provisional restoration was immediately placed on selected implants. Periapical radiographs were taken at implant placement, 6 and 12 months postsurgery, and annually thereafter. The distance between the alveolar crest and the implant shoulder was measured at the mesial and distal aspects of each implant. Bone changes (BC) and annual rate of bone change (Rate) were calculated for the first year and the following 4 years. Results: Total BC (mean ± standard deviation) from baseline to 5 years (BC t0 t5 ) was 1.41 ± 0.67 mm. First-year BC (BC t0 t1 ) was 1.14 ± 0.86 mm, and BC in years 2 to 5 (BC t1 t5 ) was 0.27 ± 0.69 mm. Rate t0 t1 was 0.98 ± 0.79 mm/year and rate t1 t5 was 0.06 ± 0.17 mm/year. Fifty-seven of 61 available implants met the criterion of cumulative bone loss of no more than 1.5 mm for the first year and 0.2 mm/year for the following years. Four implants (7%; 95% confidence interval: 0.4% to 13.6%) failed the criterion. Conclusions: Dental implants in patients with a history of periodontitis showed radiographic bone changes similar to previous reports in the literature. After the first year, immediately restored implants exhibited crestal bone loss rates similar to those seen for conventionally restored implants. Int J Oral Maxillofac Implants 2012;27: Key words: bone loss, dental implants, immediate loading, periodontal disease Implant therapy is currently considered a successful and acceptable means to restore missing teeth. During the decades that have passed since the widespread acceptance of implant dentistry, placement protocols have evolved to vary the timing of implant placement from late (in completely healed sites) through delayed and, finally, immediate placement following extraction. These procedures were developed to better meet patients expectations Assistant Professor, Department of Periodontology, School of Graduate Dentistry, Rambam Health Care Campus and Faculty of Medicine, Technion, Haifa, Israel. 2 Professor, Department of Periodontology, School of Graduate Dentistry, Rambam Health Care Campus and Faculty of Medicine, Technion, Haifa, Israel. Correspondence to: Dr Jacob Horwitz, Department of Periodontology, School of Graduate Dentistry, Rambam Health Care Campus and Faculty of Medicine, PO Box 9602, Technion, Haifa, Israel. Fax: j_horwitz@ rambam.health.gov.il The criteria for implant success, as proposed by Albrektsson et al, 5 include crestal bone loss as a major outcome variable, with success defined as < 0.2 mm of bone loss annually following the implant s first year of service. During the first year, significantly more bone loss may be observed, 6 often up to 1.5 mm, although some implants may exhibit greater bone loss. The definition of periodontally compromised/periodontally susceptible patients has been used when evaluating survival and success rates of implants, because periodontal disease has been considered a risk factor for implant therapy. Greater long-term periimplant marginal bone loss versus periodontally healthy subjects, sometimes statistically significant, has been observed in this category of patients. 7,8 Still, there are relatively few long-term prospective studies of immediately restored implants and even fewer in periodontally susceptible patients. The present authors previously reported on the 1-year outcome of immediate placement and restoration of dental implants in periodontally susceptible patients. 9,10 Therefore, the aim of the present study was to evaluate the The International Journal of Oral & Maxillofacial Implants 1137

2 radiographic bone loss of immediately placed and restored dental implants up to 5 years after insertion in the same patient group. Materials and Methods The methodology was described in detail in previous publications. 9 Briefly, 19 patients were treated at the Department of Periodontology, School of Graduate Dentistry, at the Rambam Health Care Campus. All were between the ages of 35 and 72 years and had been diagnosed with chronic periodontitis based on assessments of clinical attachment loss, alveolar bone loss, periodontal pocket measurements, and gingival inflammation. 11 They had no complicating systemic conditions that contraindicated surgical periodontal and implant treatment and/or radiographic evaluation (eg, pregnancy, uncontrolled diabetes), and required either (1) a fixed full-arch restoration for the maxilla or mandible or (2) a fixed partial implant-supported restoration in the esthetic zone. Four patients were smokers (5 to 10 cigarettes/day with 3.75 to 10 pack-years). Patients received periodontal treatment from various therapists, including oral hygiene instructions, scaling and root planing, and periodontal surgery, as necessary prior to recruitment into the study. Casts, periapical and panoramic radiographs, and computed tomography were used for evaluation and treatment planning. The treatment options were presented to the patient, and final eligibility was ascertained when patients expressed their preference for an implant-supported fixed restoration. The study protocol was approved by the health care campus medical ethics committee. Implant surgery included extraction of teeth with inadequate osseous support. Implants (MIS Implant Technologies) were inserted with the aid of a surgical guide that had been prepared with planning software (Med3D). One or two implants were used for the support of partial prostheses. In edentulous arches, three or four implants were used for the support of a provisional restoration. Implant locations were selected strategically to provide an adequately esthetic provisional solution. Conical abutments were connected to the implants, and a prefabricated screw-retained restoration was adapted and delivered. The rest of the implants either received healing abutments and were left to heal (nonsubmerged single-stage implants) or they were submerged for a period of 6 months. Thus, implants could be categorized as (1) immediately restored (R) (n = 42), (2) nonsubmerged and not restored (NR) (n = 9), or (3) submerged (S) (n = 23). The opposing dentition consisted of natural teeth and/or a fixed tooth-supported restoration. Postoperatively, patients were prescribed a 0.2% chlorhexidine mouth rinse, amoxicillin (500 mg three times a day for 7 days), and analgesic therapy as necessary. Patients were examined 7 to 10 days after surgery for suture removal and again after 2, 4, and 8 weeks and 3 months. At 6 months the provisional restoration was removed, implant integration was evaluated, stage-two surgery was performed for the submerged implants (group S), and patients were referred to a prosthodontist for their definitive prosthetic restorations. Any failed implants were replaced by additional implants, which were not included in the analysis. After they had completed the active phase of therapy, patients were enrolled in a maintenance program; visits occurred two to four times per year and included plaque examination, probing depths and bleeding, hygiene control and motivation, scaling and root planing as necessary, and, annually, periapical radiographs of the implants. Radiographs were taken with a paralleling appliance (XCP film holder, Rinn/Dentsply International), and the technician attempted to align the film/ sensor parallel to the implants axes; however, standardization of the position of each implant in each evaluation was not attempted. Initially, conventional radiographs were used, digitized, and stored electronically using a scanner (DiMAGE Scan Elite II, Konica Minolta Holdings). Later, when digital radiography became available, radiographs were acquired with a digital sensor and stored digitally (Planmeca Dimaxis Pro, Planmeca). Radiographic measurements were performed digitally with the assistance of software (Planmeca Dimaxis Pro). Radiographic images were calibrated using the implant shoulder diameter of 4.1 mm. The distance between the alveolar crest and the implant shoulder was measured at the mesial (CBM) and distal (CBD) aspect of each implant. The mean implant bone level (BL) was calculated as the average between CBM and CBD. Bone level changes (BC ta tb ) were calculated as BL ta BL tb, where t indicated the time of examination (eg, t 0 = baseline; t 1 = 1 year; t 5 = 5 years) recorded in months from baseline, and were expressed in millimeters. Baseline (t 0 ) for groups R and NR was set at implant placement, while for group S it was set at stage-two surgery. An annual rate of bone change was calculated to compensate for the differences in periods between patients, as follows: Rate ta tb = (BC ta tb ) 12/(t b t a ); results were expressed in millimeters/year. Positive values represent radiographic bone gain and negative values represent bone loss. Data analysis was performed using a statistical software program (SPSS version 18, SPSS, IBM). Descriptive statistics were used. Analysis of variance with the Bonferroni adjustment was performed for the three categories of implants. The level of significance was set at 5%. The Wilcoxon signed ranks test was used to compare rates Volume 27, Number 5, 2012

3 Table 1 No. of Implants with Available Radiographic Data Time Implants with available data Missing data (dropouts) Landmark identification impossible Total T T T T Table 2 Bone Level Changes (mm) and Annual Rate of Change (mm/y) Bone level change Annual rate of change Group/time n Mean ± SD Median Range n Mean ± SD Median Range R T 0 T ± to ± to 0.03 T 0 T ± to ± to T 1 T ± to ± to T 1 T ± to ± to 0.22 NR T 0 T ± to ± to 0.16 T 0 T ± to ± to 0.33 T 1 T ± to ± to 0.25 T 1 T ± to ± to 0.16 S T 0 T ± to ± to 0.06 T 0 T ± to ± to 0.02 T 1 T 3 T 1 T ± to ± to 0.11 Total T 0 T ± to ± to 0.03 T 0 T ± to ± to 1.72 T 1 T ± to ± to 0.60 T 1 T ± to ± to 0.33 T 0 = baseline; T 1 = 1 year; T 3 = 3 years; T 5 = 5 years. Results Nineteen patients were initially treated and received a total of 74 implants. As described in previous publications, 12 implants in four patients failed during the initial 6 months and were replaced (replacement implants were not included in the present study). One patient who received one implant did not show up after the first-year appointment. Therefore, the present report includes the 61 implants for which data were available. During subsequent years, no further implants were removed. Patients received fixed restorations and were invited to participate in a maintenance program every 3 to 6 months. Until the end of the study, no implant showed evidence of peri-implant radiolucency; persistent and/ or irreversible signs and symptoms such as pain, infections, neuropathies, paresthesia; or violation of the mandibular canal. Implants were restored with fixed cemented restorations; therefore, after the first year they were not tested individually for signs of mobility. Some patients did not attend their examinations or it was impossible to identify landmarks on the radiographs for bone loss measurements (Table 1). At 3 years, data were missing for six implants (two R and four NR implants), and at 5 years, 19 implants could not be analyzed (10 R and 9 S implants). No statistically significant differences were observed in R (the annual rate of bone loss) between the three implant groups and between 3 and 5 years (statistical data on file). Thus, 61 of the original 74 implants were available for the present analysis. Table 2 summarizes the radiographic findings. BC t0 t5 (mean ± standard deviation [SD]) was 1.41 ± 0.67 mm. BC t0 t1 was 1.14 ± 0.86 mm, and BC t1 t5 was 0.27 ± 0.69 mm. The International Journal of Oral & Maxillofacial Implants 1139

4 % of implants n = 10 n = 14 n = 13 n = 15 Fig 1 Implant distribution according to bone change during the first year. Green columns represent implants with bone loss of more than 1.5 mm during the first year n = 4 n = 2 0 < to < to < to < to < 0 Bone change during year 1 (mm) 0 to < n = 19 % of implants n = 11 n = n = 2 < to < to < 0.3 n = to < to 0.1 to < 0.1 < 0 Bone change rate (mm) 0 to < 0.1 n = to < 0.2 n = to < to < 0.4 Fig 2 Implant distribution according to bone change rate for years 2 to 5 (rate t1 t5 ). Green columns represent implants with more than 0.2 mm of bone loss per year during years 2 to 5. Rate t0 t1 was 0.98 ± 0.79 mm/year and rate t1 t5 was 0.06 ± 0.17 mm/year. While some 23% of the implants lost more than 1.5 mm of bone during their first year, four implants gained bone during this period (Fig 1). Eight of the implants (14% of the sample) lost bone at an annual rate greater than 0.2 mm during the 4 subsequent years, and 18% of the implants exhibited bone gain during this period (Fig 2). Total bone change was compared to a criterion derived from the criteria of Albrektsson et al for evaluating the success rate of implants. 5 The criterion was a cumulative bone loss of 1.5 mm for the first year and 0.2 mm/year for the following years. For example, an implant would meet the criterion after 5 years if it exhibited a cumulative bone loss of up to 2.3 mm. Using this criterion, 57 of the 61 implants that were available at the end of the study met or exceeded this criterion. Four implants failed the criterion (7% of the implants; 95% confidence interval, 0.4% to 13.6%). No statistically significant difference was found between the different implant groups (R, NR, and S) for BC t0 t5, BC t0 t1, and BC t1 t5 (analysis of variance) (Table 3). No statistically significant difference was found between rate t1 t3 and rate t1 t5 (Wilcoxon signed rank test, data on file) Volume 27, Number 5, 2012

5 Table 3 Multiple Comparisons Between Implant Groups (Analysis of Variance) Dependent variable/comparison Mean difference ± SE (mm) Significance 95% CI BC T0 T1 R NR 0.29 ± 0.31 > to 1.06 R S 0.30 ± to 0.28 NR S 0.58 ± to 0.23 BC T1 T5 R NR 0.14 ± 0.22 > to 0.40 R S 0.20 ± to 0.61 NR S 0.34 ± to 0.92 BC T0 T5 R NR 0.15 ± 0.27 > to 0.82 R S 0.10 ± 0.20 > to 0.41 NR S 0.24 ± 0.29 > to 0.47 CI = confidence interval. Discussion This prospective study describes bone changes occurring up to 5 years after implant placement/stage-two surgery in periodontally susceptible patients. No additional implants were lost following the initial healing period and 6-month integration period. Similar to previous studies, the rate of bone loss was higher during the first year and decreased markedly thereafter. These results can be compared to other studies of patients with and without a history of periodontal disease. For example, in a prospective 5-year multicenter study involving patients without a history of periodontitis, 12 clinically significant remodeling of the marginal bone occurred during the first 6 months after implant placement, with a mean (± SD) marginal bone loss of 2.44 ± 1.20 mm. This is much higher than the 1.14 ± 0.86 mm of first-year bone loss observed in the present study. In the first year postloading in the cited study, 12 a mean of 0.22 ± 0.42 mm of bone was lost, and between 1 year postloading and the last 5-year recall, a mean bone loss of 0.18 ± 0.88 mm occurred, similar to the 0.21 ± 0.59 mm result in the present study. However, the aforementioned study did not take into account the additional 0.22 mm of bone loss that took place during the first year postloading. Overall, 2.84 ± 1.63 mm of bone loss occurred between implant placement and the 5-year postloading follow-up, 12 versus only 1.41 ± 0.67 mm in the present study. Observations in periodontal patients have varied. Mengel et al found a 5-year postloading mean bone loss of 0.19 mm in osseointegrated implants in patients treated for generalized chronic periodontitis. 13 Another 5-year prospective study of implant-supported fixed partial dentures in periodontitis-susceptible subjects found that during the first year in function there was an average of 0.31 ± 0.81 mm of bone loss. 14 During the subsequent 4 years, the calculated annual change in peri-implant bone level was 0.03 ± 0.20 mm, half the rate of 0.07 ± 0.19 mm measured in the present study. Since the SDs in the cited studies were all relatively large, the results of the present study may be interpreted to fall within the same range. Comparison of bone loss rates between studies is difficult because, as previously described, 15 there are considerable discrepancies among different studies, including differences in methods for measuring bone loss. Also, various parameters, including the accuracy of the measurements themselves, influence radiographic interpretation of bone gain and bone loss. 16 In the present study, bone gain was observed radiographically around 4% of the implants during the first year and around 21% during the next 4 years. This may represent true bone apposition but may also be attributed to errors arising from incorrect alignment or inaccurate landmark identification. 17 The issue of periodontal susceptibility has been addressed in many studies and reviews In spite of considerable variations in study design, length of follow-up, definition of patient populations with respect to periodontal status, outcome measures, and supportive periodontal therapy regimens, the systematic reviews indicate that subjects with a history of periodontitis are at a greater risk for peri-implant disease. 24 Therefore it may be expected that patients in the present study would demonstrate more bone The International Journal of Oral & Maxillofacial Implants 1141

6 loss than periodontally healthy subjects. Indeed, 25% of the implants exhibited more than 1.5 mm of bone loss during the first year, and 14% exhibited more than 0.2 mm of bone loss annually during subsequent years. This can be compared to the findings of Eliasson et al, 25 who, in a retrospective analysis of early loading of full-arch mandibular prostheses, found that only 6.3% of the implants lost more than 1.2 mm during the first year and 13.2% lost more than 0.6 mm during the following 4 years. This smaller bone loss was found in a group of patients who were not periodontally susceptible. It should be kept in mind, however, that the magnitude of the difference in bone loss around implants between periodontally healthy and susceptible patients as presented in the literature cannot yet be ascertained. Therefore, caution should be exerted when interpreting the results of the present study in implicating periodontal susceptibility/previous periodontal therapy as a risk factor for bone loss. In a recent study reporting on the 3-year results of immediate versus early loading of dental implants, the mean 3-year bone loss for immediately loaded implants was reported to be 1.51 ± 0.79 mm. 26 The patients in the study were not reported to be periodontally susceptible. The mean follow-up time after implant placement was 39.8 months (range, 36.7 to 53.1 months). The survival rate for immediately loaded implants was 85%, compared with 100% for implants loaded early, after an observation period of 3 years. The present study results are within the same range for all the aforementioned parameters, except that the total survival rate in the present study was 84%, while that of immediately restored implants was 74% (keeping in mind that 9 of 10 failures were implants that had been inserted into fresh extraction sites and loaded immediately). Exclusion of smokers or comparison of smokers to nonsmokers was not performed in the present study for two reasons: (1) the small number of smokers and (2) these smokers, with a habit of up to 10 cigarettes/ day and up to 10 pack-years of smoking, were considered light smokers. Although smoking is considered a risk factor for implant survival, a significant association has been found only for heavy smokers (> 20 cigarettes/ day 27 ). Smoking is also a risk factor for periodontal disease progression in a dose-dependent manner. However, the odds ratio for light smokers is low. An odds ratio of 1.32 was found for smokers with 0.1 to 9.9 pack-years. 28 There are some major drawbacks in the present study, including the high dropout rate, the lack of a control group, the inclusion of multiple implants in the same patient, and the small sample size. The difficulty of bringing all participants to every maintenance visit resulted in having 5-year data for only 70% of the implants. This serious drawback may be explained by the fact that patients were satisfied with their prostheses, felt that they no longer had problems, and did not wish to make the effort of attending the clinic for all the recommended maintenance and examination appointments. It can be speculated that the patients who were accepted into the study initially sought a rapid solution to their dental problems and were unwilling to make the repeated ongoing effort needed for regular maintenance once their main concern was resolved. This issue poses an obstacle in continuous monitoring of this patient group and diminishes the power of the present study. This should be taken into consideration in future studies. Since the present study was designed, the understanding of the relationship between periodontal disease, dental implant therapy, and immediate loading has accumulated significantly. Today, a comparison between a more focused experimental group of one immediately restored implant per patient and a control group of one delayed-loaded implant per patient would be preferred. Also, the sample size in the present study was calculated post hoc (see further discussion). The use of a criterion for bone loss served to establish the percentage of successful implants according to Albrektsson et al. 5 Clearly, different parameters influence implant integration, initial bone loss, and bone loss after integration and prosthetic rehabilitation. Therefore, an overall bone loss that includes both initial and bone changes is an important parameter for implant evaluation. In the present study, an overall failure rate (a cumulative loss of 1.5 mm for the first year and more than 0.2 mm/y during the following 4 years) of 7% (confidence interval, 0.4% to 13.6%) was calculated. Conversely, 93% of the implants exhibited less radiographic bone loss than the radiographic criterion during a 5-year period. The lower limit of the 95% confidence interval of the failure rate is 0.4% and not 0%. This indicates that indeed some implants should be expected to exhibit more bone loss, although survival may still be unaffected. The 95% confidence interval range of 13.2% is fairly large because of the small sample size and large variation in the results. To arrive at a confidence interval within a range of ± 2%, the authors calculated that a similar experiment with a sample of approximately 700 implants should be performed. 29,30 Conclusions In the present study, dental implants in periodontally susceptible patients showed radiographic bone changes similar to previous reports in the literature regarding patients with and without a history of periodontitis. After the first year, immediately restored implants exhibited bone loss rates similar to those seen around conventionally restored implants Volume 27, Number 5, 2012

7 AcknowledGMents The authors would like to express their deepest gratitude to Dr Otman Zuabi for his clinical contribution to the study. We greatly appreciate the valuable help of Ms Tanya Mashiach, Rambam Health Care Campus medical statistics unit, for the statistical analysis. This study was partly supported by a grant from MIS Implant Technologies. References 1. Brånemark PI, Adell R, Breine U, Hansson BO, Lindström J, Ohlsson A. Intra-osseous anchorage of dental prostheses. I. Experimental studies. Scand J Plast Reconstr Surg 1969;3: Lazzara RJ. Immediate implant placement into extraction sites: Surgical and restorative advantages. Int J Periodontics Restorative Dent 1989;9: Atieh MA, Payne AGT, Duncan WJ, Cullinan MP. Immediate restoration/loading of immediately placed single implants: Is it an effective bimodal approach? Clin Oral Implants Res 2009;7: Esposito M, Grusovin MG, Polyzos IP, Felice P, Worthington HV. Interventions for replacing missing teeth: Dental implants in fresh extraction sockets (immediate, immediate-delayed and delayed implants). Cochrane Database Syst Rev 2010 Sep 8;(9):CD Albrektsson T, Zarb GA, Worthington P, Eriksson AR. The long-term efficacy of currently used dental implants: A review and proposed criteria of success. Int J Oral Maxillofac Implants 1986;1: Adell R, Lekholm U, Rockler B, Brånemark PI. A 15-year study of osseointegrated implants in the treatment of the edentulous jaw. Int J Oral Surg 1981;10: Van der Weijden G, Van Bemmen KM, Renvert S. Implant therapy in partially edentulous, periodontally compromised patients: A review. J Clin Periodontol 2005;32: Karoussis IK, Kotsovilis S, Fourmousis I. A comprehensive and critical review of dental implant prognosis in periodontally compromised partially edentulous patients. Clin Oral Implants Res 2007;18: Horwitz J, Zuabi O, Peled M, Machtei EE. Immediate and delayed restoration of dental implants in periodontally susceptible patients: 1-year results. Int J Oral Maxillofac Implants 2007;3: Horwitz J, Zuabi O, Machtei E. Radiographic changes around immediately restored dental implants in periodontally susceptible patients: 1-year results. Int J Oral Maxillofac Implants 2008;23: Flemmig TF. Periodontitis. Ann Periodontol 1999;4: Cochran DL, Nummikoski PV, Schoolfield JD, Jones AA, Oates TW. A prospective multicenter 5-year radiographic evaluation of crestal bone levels over time in 596 dental implants placed in 192 patients. J Periodontol 2009;5: Mengel R, Schröder T, Flores-de-Jacoby L. Osseointegrated implants in patients treated for generalized chronic periodontitis and generalized aggressive periodontitis: 3- and 5-year results of a prospective long-term study. J Periodontol 2001;8: Wennström JL, Ekestubbe A, Gröndahl K, Karlsson S, Lindhe J. Oral rehabilitation with implant-supported fixed partial dentures in periodontitis-susceptible subjects. A 5-year prospective study. J Clin Periodontol. 2004;9: Karoussis IK, Kotsovilis S, Fourmousis I. A comprehensive and critical review of dental implant prognosis in periodontally compromised partially edentulous patients. Clin Oral Implants Res 2007;18: Tidmarsh BG. Radiographic interpretation of endodontic lesions A shadow of reality. Int Dent J 1987;1: Leonardi R, Annunziata A, Caltabiano M. Landmark identification error in posteroanterior cephalometric radiography. A systematic review. Angle Orthod 2008;4: Van der Weyden GA, van Bemmel KM, Renvert S. Implant therapy in partially edentulous periodontally compromised patients. A review. J Clin Periodontol 2005;32: Schou S. Implant treatment in periodontitis-susceptible patients: A systematic review. J Oral Rehabil 2008;35(suppl 1): Karoussis IK, Kotsovilis S, Fourmousis I. A comprehensive and critical review of dental implants in periodontally compromised partially edentulous patients. Clin Oral Implants Res 2007;18: Quirynen M, De Soete M, van Steenberghe D. Infectious risks for oral implants: A review of the literature. Clin Oral Implants Res 2002:13: Al-Zahrani MS. Implant therapy in aggressive periodontitis patients: A systematic review and clinical implications. Quintessence Int 2008;39: Ong CT, Ivanovski S, Needleman IG, et al. Systematic review of implant outcomes in treated periodontitis subjects. J Clin Periodontol 2008;35: Heitz-Mayfield LJA. Peri-implant diseases: Diagnosis and risk indicators. J Clin Periodontol 2008;35(suppl 8): Eliasson A, Blomqvist F, Wennerberg A, Johansson A. A retrospective analysis of early and delayed loading of full-arch mandibular prostheses using three different implant systems: Clinical results with up to 5 years of loading. Clin Implant Dent Relat Res 2009;2: Zembić A, Glauser R, Khraisat A, Hämmerle CH. Immediate vs early loading of dental implants: 3-year results of a randomized controlled clinical trial. Clin Oral Implants Res 2010;5: Sánchez-Pérez A, Moya-Villaescusa MJ, Caffesse RG. Tobacco as a risk factor for survival of dental implants. J Periodontol 2007;78: Nishida N, Tanaka M, Hayashi N, et al. Determination of smoking and obesity as periodontitis risks using the classification and regression tree method. J Periodontol 2005;76: Gardner MJ, Altman DG. Estimation rather than hypothesis testing: Confidence intervals rather than P values. In: Gardner MJ, Altman DG (eds). Statistics with Confidence. Confidence Intervals and Statistical Guidelines. London: British Medical Journal, 1990: Gardner MJ, Altman DG. Calculating confidence intervals for proportions and their differences. In: Gardner MJ, Altman DG (eds). Statistics with Confidence. Confidence Intervals and Statistical Guidelines. London: British Medical Journal, 1990: The International Journal of Oral & Maxillofacial Implants 1143