The aim of this study was to perform a systematic review and meta-analysis to evaluate the possible benefits of platform-switching (PSW) implants when compared to regular platform (RP) implants in the categories of bone preservation and longevity. This systematic review and meta-analysis was performed in accordance with the PRISMA statement, PICO question, and Jadad scale. The relative risk (RR) of failure and the mean difference for marginal bone loss were calculated considering a confidence interval (CI) of 95%. Heterogeneity and subgroup analyses were performed, and funnel plots drawn. Twenty-five studies (17 randomized controlled trials (RCTs) and eight prospective studies) involving 1098 patients and 2310 implants were analysed. The meta-analysis revealed a significant reduction in crestal bone loss for PSW implants compared with RP implants (−0.41 mm, 95% CI −0.52 to −0.29, P < 0.00001). However, there was no statistically significant difference in implant failure (RR 1.10, 95% CI 0.6–2.02, P = 0.75). A reduction in bone loss with PSW implants was observed for the following subgroups: RCTs only, implants in the maxilla, and implants in the mandible. PSW implants presented lower bone resorption compared with RP implants. RCTs should be done to explain the possible biases.
The introduction of larger-diameter implants during a period when compatible prosthetic components were not accessible allowed for standard prosthetic components (4.1 mm) to be used with large-diameter implants (5 mm and 6 mm). This concept became known as ‘platform switching’. The first clinical case studies and retrospective studies on platform switching indicated a lower rate of bone loss around these dental implants when compared with implants that received prosthetic abutments of the same diameter platform ( Fig. 1 ).
Several theories have emerged to explain the lower bone loss with this platform-switching treatment modality. It has been suggested that positioning the implant/abutment interface away from the bone crest allows the biological width to be determined horizontally, enabling the creation of an additional horizontal surface area for the attachment of soft tissue. The peri-implant microbiota is another relevant factor, since the design of these implants can increase the distance between the inflammatory cell infiltrate and the bone crest, thereby minimizing the effects of inflammation on peri-implant marginal bone with platform-switching (PSW) implants. Finally, there is a biomechanical theory that relates the possibility of centralization stress on the long axis of these implants, thus reducing tension in the peri-implant cortical bone.
After the phenomenon of bone preservation was confirmed, clinical studies evaluating the platform-switching concept began to appear. However, several case reports presented a sample of 10 patients or fewer. Randomized controlled trials (RCTs) with the aim of comparing the effects of PSW implants and regular platform (RP) implants in patients have emerged in the last 5 years, allowing the preparation of literature reviews addressing the topic of bone preservation around these implants. However, there remains a need to clarify the effects of PSW implants in relation to marginal bone loss, as indicated by previous systematic reviews.
Recently, RCTs have been published addressing the issue of PSW implants, leading to the need for an updated analysis of published studies. Moreover, biomechanical studies have been published that may provide further insight into the proposed subject. Thus, the aim of this study was to conduct a systematic review and meta-analysis of the proposed topic. The null hypothesis was that PSW implants show a rate of bone remodelling similar to RP implants.
Materials and methods
This study was conducted in accordance with the criteria put forward in the PRISMA-2009 guidelines. The PICO question was formulated. This study was also performed with reference to other previous systematic reviews and meta-analyses.
Protocol and registration
This systematic review was registered in the PROSPERO database, an international prospective register of systematic reviews in health and social care (National Institute for Health Research, UK; pre-protocol CRD 42013005728).
The studies selected for this analysis met the criteria established by the index PICO: (1) population: patients undergoing dental implant surgery; (2) intervention: patients receiving implants with a platform-switching geometry; (3) comparison: patients receiving implants with a regular abutment; (4) outcome: the main outcomes were the comparison of bone loss and implant survival rates (platform-switching and regular platform).
Inclusion criteria were the following: articles published in the English language; studies with at least 12 months of follow-up (clinical studies in humans); RCTs and prospective studies with at least five implants (titanium implants) placed in the control group (RP) and in the study group (PSW).
Sources of information
The MEDLINE/PubMed, Cochrane Central Register of Controlled Trials, and EMBASE databases were searched. These searches were conducted for articles published up until 1 July 2015. All studies identified by the inclusion criteria were analysed. Authors were contacted when necessary to obtain possible additional information.
Key words available in medical subject headings (MeSH, PubMed) related to PSW implants and RP implants were selected. The Boolean search operators used were ‘Dental Implant Platform Switching’ and ‘Platform Switching, Dental Implant,’ and the key words were ‘dental implant–abutment design’ [MeSH Terms] OR (‘dental’ [All Fields] AND ‘implant–abutment’ [All Fields] AND ‘design’ [All Fields]) OR ‘dental implant–abutment design’ [All Fields] OR (‘dental’ [All Fields] AND ‘implant’ [All Fields] AND ‘platform’ [All Fields] AND ‘switching’ [All Fields]).
A manual search of journals published over the last 6 months was also done: Clinical Implant Dentistry and Related Research , Clinical Oral Implant Research , Implant Dentistry , International Journal of Oral and Maxillofacial Surgery , Journal of Clinical Periodontology , Journal of Dental Research , Journal of Maxillofacial and Oral Surgery , Journal of Oral Implantology , Journal of Periodontology , Journal of Prosthetic Dentistry , Journal of Prosthodontics , Journal of Oral Rehabilitation , Oral Medicine, Oral Pathology, Oral Radiology, and Endodontics , Periodontology 2000 , International Journal of Oral and Maxillofacial Implants , and International Journal of Periodontics and Restorative Dentistry .
Process of data collection
Study selection was organized independently by two calibrated examiners (J.F.S. and V.E.S.B.) and by a third reviewer (E.P.P). Inter-examiner (kappa) tests were conducted to evaluate the selection of titles and abstracts, and complete reading with interpretation of the article, resulting in concordance test values of κ = 0.88, 1, 1 for MEDLINE/PubMed, κ = 1, 1, 1 for Cochrane Central Register of Controlled Trials, and κ = 1, 1, 1 for EMBASE. For the MEDLINE/PubMed database search, a meeting was required to reach consensus, in which all the discrepancies were discussed and resolved by the third reviewer (E.P.P.). All titles and abstracts evaluated as eligible were separated and analysed completely. A manual search of the journals was conducted by one reviewer (J.F.S.) and independently by another reviewer (V.E.S.B.), adding six articles to the original sample.
The selection of studies for the systematic review and meta-analysis is shown in detail in Fig. 2 , as recommended in the literature.
The following data were extracted from each study: (1) author; (2) publication year; (3) number of patients; (4) number of implants and sites; (5) implant system; (6) length and diameter of implants; (7) timing of the installation of prostheses; (8) mismatching of the platform and implant; (9) bone remodelling rate for implants – platform switching and control group; (10) survival rate of implants for each situation analysed; (11) follow-up time after implant placement; and (12) study design analysed.
Quality analysis of the studies included in the systematic review
The quality of each study included in this review was assessed using the Jadad scale. They were thus classified on a scale of 0–5, with a score above 3 indicating an appropriate study.
For the purpose of comparing the success rate of implants using the concept of platform switching with the regular platform (a dichotomous outcome), the risk ratio (RR) with 95% confidence interval (CI) was used.
In order to analyse the rate of peri-implant bone loss (continuous outcome), the average bone loss for the RP implants and PSW implants was identified, and the overall standard deviation of each group was analysed. The weight contribution was also calculated in the analyses.
A P -value of <0.05 was considered to indicate statistical significance. The software program Review Manager was used for the meta-analysis, as well as to construct forest and funnel plots (RevMan version 5.3; The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark, 2014).
Risk of bias in the studies
The fixed-effects model was used when there was no statistically significant difference, and the random-effects model was adopted when there was a statistically significant difference, i.e., a high level of heterogeneity between trials was considered significant for P < 0.1. Heterogeneity was assessed using the Q method ( χ 2 ), and the value of I 2 was calculated. The statistical value of I 2 was used to analyse the variations in heterogeneity: I 2 above 75 (0–100) was found to indicate relevant heterogeneity. Funnel plots were used to assess heterogeneity.
In order to analyse the sensitivity of the tests employed, a subgroup analysis was performed to identify any potential causes of heterogeneity. Specifically, the subgroups considered were (1) RCTs only, (2) installation in the maxilla, and (3) installation in the mandible. Furthermore, funnel plots (effect size vs. standard error) were used to evaluate bias, with asymmetry indicating a possible bias in the studies evaluated.
The database search identified 1715 articles; Fig. 2 shows the article selection process. A total of 25 studies were eligible, reporting 1098 patients with 2310 implants placed. The average age of patients included in all of the studies, except for three clinical studies, was 50.73 years. The main results are summarized in Table 1 .
|Canullo et al.||Canullo et al.||Canullo et al.||Cappiello et al.||Crespi et al.||Dursun et al.|
|Patients ( n )||31||22||9||45||45||19|
|Age, years (mean)||52.1||50||59||NR||48.73||42.93|
|Implants ( n )||61
(G1, 17; G2, 13; G3, 14; CG, 17)
(G1, 11; CG, 11)
(G1, 6; G2, 5; G3, 6; CG, 5)
(G1, 75; CG, 56) b
(G1, 30; CG, 34)
(G1, 16; CG, 16)
|Implant location||Posterior maxilla||Maxilla||Posterior maxilla||NR||Maxilla and mandible||Maxilla and mandible|
|Implant system a||Global||Global||Global||3i||Seven and Ankylos||PS system and SP system|
|Connection||NR||Internal||NR||Internal||CG, EH; PSW, MT||NR|
|Length (mm)||NR||NR||13||10, 11.5, 13||14||11|
|Diameter (mm)||3.8, 4.3, 4.8, 5.5||5.5||3.8–5.5||4||3.8–5.5||CG, 3.75; PSW, 3.8|
|Loading protocol (months)||Delayed||Immediate||Delayed||2||Immediate||Delayed|
|Definitive rehabilitation (months)||3||2||2–3||12||6||3|
|Implant–abutment diameter difference on each side (mm)||CG, 0; G1, 0.25; G2, 0.5; G3, 0.85||CG, 0; G1, 0.85||CG, 0; G1, 0.25; G2, 0.5; G3, 0.85||CG, 0; G1, 0.4||NC||G, 0.37|
|Marginal bone level changes (mean (SD), mm)||G1, 0.99 (0.42)
G2, 0.82 (0.36)
G3, 0.56 (0.31) c
CG, 1.49 (0.54)
|G1, 0.3 (0.16)
CG, 1.19 (0.35)
|G1, 0.832 (0.3939)
G2, 0.486 (0.2242)
G3, 0.375 (0.1234) c
CG, 1.358 (0.3939)
|G1, 0.95 (0.32)
CG, 1.67 (0.37) d
|G1, 0.78 (0.45)
CG, 0.73 (0.52)
|G1, 0.84 (0.36)
CG, 0.76 (0.41)
|Survival rate (%)||100||100||100||G1, 98.3; CG, 100||100||100|
|Failures ( n )||0||0||0||G1, 1||0||0|
|Mean follow-up period (months)||33||25||36||12||24||12|
|Enkling et al.||Fernández-Formoso et al.||Hurzeler et al.||Kielbassa et al.||Pieri et al.||Pozzi et al.||Prosper et al.|
|Patients ( n )||25||51||15||177||38||34||60|
|Age, years (mean)||51||43.3||55.3||48.7||46.6||52.2||53.9|
|Implants ( n )||50
(G1, 25; CG, 25)
(G1, 58; CG, 56)
(G1, 14; CG, 8)
(G1, 199; CG, 126)
(G1, 20; CG, 20)
(G1, 44; CG, 44)
(G1, 180; CG, 180)
|Implant location||Posterior mandible||Maxilla and mandible||Maxilla and mandible||Maxilla and mandible||Posterior maxilla||Posterior mandible||Maxilla and mandible|
|Implant system a||SICace Invent||Straumann||3i||Nobel Biocare||Samo Smiler||Nobel Biocare||Bioactive|
|Connection||IH||MT||NC||IH, 117; EH, 82; CG, 126||G1, MT; CG, IH||G1, IH; CG, EH||NC|
|Length (mm)||9.5||CG, 8–12; PSW, 8–14||NC||10, 11.5, 13, 15, 16||NC||10–13||11, 13, 15|
|Diameter (mm)||4||3.3, 4.1, 4.8||5.0||3.5 or 4.3||NC||CG, 4.0; G1, 4.3||3.3, 3.8, 4.5|
|Loading protocol (months)||3||NC||NC||Immediate||Immediate||2||Delayed|
|Definitive rehabilitation (months)||12||NC||NC||12||4||4||3 (mandible); 6 (maxilla)|
|Implant–abutment diameter difference on each side (mm)||0.35||NC||0.45||NC||0.35||NC||0.25/0.35|
|Marginal bone level changes (mean (SD), mm)||G1, 0.53 (0.35)
CG, 0.58 (0.55)
|G1, 0.68 (0.88)
CG, 2.23 (0.22)
|G1, 0.22 (0.53)
CG, 2.02 (0.49)
|IH, 0.95 (1.37)
EH, 0.64 (0.97)
CG, 0.63 (1.18)
|CG, 0.51 (0.24)
G1, 0.2 (0.17)
|CG, 1.15 (0.34)
G1, 0.68 (0.34)
|CG, 0.193 (0.474)
G1, 0.055 (0.234)
|Survival rate (%)||100||100||100||IH, 96.6; EH, 96.3; CG, 97.6||CG, 100; G1, 94.7||100||CG, 96.7; G1, 100|
|Failures ( n )||0||0||0||10 (IH, 4; EH, 3; CG, 3)||1||0||6|
|Mean follow-up period (months)||12||12||12||12||12||12||24|
|Telleman et al.||Telleman et al.||Trammell et al.||Vandeweghe and De Bruyn||Vigolo and Givani|
|Patients ( n )||17||78||10||15||144|
|Age, years (mean)||53.7||49.8||NC||57||37|
|Implants ( n )||62
(G1, 31; CG, 31)
(G1, 55; CG, 58)
(G1, 13; CG, 12)
(G1,15; CG, 15) e
(G1, 97; CG, 85)
|Implant location||Maxilla and mandible||Maxilla and mandible||Mandible||Maxilla and mandible (posterior)||Maxilla and mandible (posterior)|
|Implant system a||3i||3i||3i||Southern Implants||3i|
|Length (mm)||8.5||8.5||8.5, 10, 11.5, 13||7||NR|
|Diameter (mm)||4.1 or 5.0||4.1 or 5.0||4, 5, or 6||7||5|
|Loading protocol (months)||Delayed||Delayed||Delayed||Delayed||Delayed|
|Definitive rehabilitation (months)||3||3||6||6||4|
|Implant–abutment diameter difference on each side (mm)||0.35 or 0.4||0.35 or 0.4||0.45||1||0.5|
|Marginal bone level changes (mean (SD), mm)||G1, 0.53 (0.54)
CG, 0.85 (0.65)
|G1, 0.51 (0.51)
CG, 0.73 (0.48)
|G1, 0.99 (0.53)
CG, 1.19 (0.58)
|G1, 0.66 (0.47) e
CG, 0.94 (0.42)
|G1, 0.6 (0.2)
CG, 1.1 (0.3)
|Survival rate (%)||93.6||CG, 93.1; CG, 94.5||100||100||100|
|Failures ( n )||4 (G1, 2; CG, 2)||CG, 4; G1, 3||0||0||0|
|Mean follow-up period (months)||12||12||24||12||60|
|Enkling et al.||Del Fabbro et al.||Glibert et al.||Meloni et al.||Pozzi et al.||Wang et al.||Guerra et al.|
|Patients ( n )||25||51||48||18||34||19||68|
|Age, years (mean)||51||55.4||>18||48||52.2||55.4||52.84|
|Implants ( n )||50
(G1, 25; CG, 25)
(G1, 55; CG, 62)
(G1, 45; CG, 70)
(G1, 18; CG, 18)
(G1, 44; CG, 44)
(G1, 15; CG, 15)
(G1, 74; CG, 72)
|Implant location||Posterior mandible||Maxilla and mandible||Maxilla and mandible||NR||Posterior mandible||Maxilla and mandible||Posterior mandible|
|Implant system a||SICace Invent||Dental Tech||3i||Nobel Biocare||Nobel Biocare||Superline||Camlog|
|Connection||IH||IH||IH||NR||G1, IH; CG, EH||IH||IH|
|Length (mm)||9.5||8–16||8.5–15||8 or 10||8.5–13||8–12||9–13|
|Diameter (mm)||4.0||3.75 and 4.75||4.0 and 5.0||4.3 and 5.0||3.9 and 4.1||4.5||3.8, 4.3, 5.0|
|Loading protocol (months)||Delayed||2/4–6||Immediate non-occlusal/delayed||Delayed||Delayed||Delayed||Delayed|
|Definitive rehabilitation (months)||4||2/4–6||3||6||4||3||2–3|
|Implant–abutment diameter difference on each side (mm)||0.35||0.5, 0.75, 1.25||0.45||0.35||0.2||0.6||0.3, 0.35|
|Marginal bone level changes (mean (SD), mm)||G1, 0.69 (0.43)
CG, 0.74 (0.57)
CG, 0.48 (0.26)
|G1, 0.63 (0.18)
CG, 1.02 (0.14)
|G1, 0.84 (0.23)
CG, 0.93 (0.26)
|G1, 0.83 (0.27)
CG, 1.29 (0.42)
|G1, 0.04 (0.08)
CG, 0.19 (0.16)
|G1, 0.40 (0.46)
CG, 0.69 (0.68)
|Survival rate (%)||100%||G1, 90.3%; CG, 96.5%||100%||100%||100%||100||G1, 97.3%; CG, 100%|
|Failures ( n )||0||G1, 5; C, 2||0||0||0||0||G1, 2|
|Mean follow-up period (months)||36||36||12||12||36||12||12|
a Global: Global, Sweden-Martina, Padua, Italy; 3i: 3i Implant Innovations, Palm Beach Gardens, FL, USA; Seven: Seven Sweden-Martina, Padua, Italy; Ankylos: Dentsply Friadent, Mannheim, Germany; PS system: Revois, Curasan AG, Frankfurt, Germany; SP system: tapered Screw vent, Zimmer Dental, Carlsbad, CA, USA; SICace Invent: SICace, SIC Invent, Basel, Switzerland; Straumann: Straumann, Basel, Switzerland; Nobel Biocare: Nobel Biocare, Gothenburg, Sweden; Samo Smiler Implants: BioSpark, Burlington, MA, USA; Bioactive: Bioactive Covering SLA, Winsix, London, UK; Southern Implants: Max Implants – Southern Implants, Irene, South Africa. Dental Tech: Dental Tech Srl, Misinto, Milan, Italy. Superline: Superline, Dentium USA, Cypress, CA, USA; Camlog: Camlog Biotechnologies AG, Basel, Switzerland.
b This study considered the placement of 56 control implants and 75 test implants (PSW); however 73 PSW concept implants (1 failed implant) and 55 control implants were included in the follow-up (Cappiello et al. ).
e The implants (total of 15 implants PSW/non-switched) were considered PSW concept on one side and control group on the other side, therefore mean bone loss was analysed for both the switched and non-switched side (Vandeweghe and De Bruyn ).
Of the 25 studies selected, 17 were RCTs and eight were controlled prospective studies ; these were published during the period 2007–2015. Three studies considered samples at two locations, 13 studies were conducted in only one centre, and five studies were multicentre ; related data were unclear for four studies.
The studies analysed reported various inclusion criteria for patient selection. Inclusion criteria encompassed healthy subjects, age >18 years, no medical contraindication, non-smoking or smoking patients consuming ≤10 cigarettes per day or ≤20 cigarettes per day, a plaque and bleeding index ≤25% and availability of longitudinal follow-up, and a keratinized mucosa ≥4 mm and adequate thickness of soft tissue (medium/wide).
The exclusion criteria included sites of acute infection, history of bisphosphonate treatment, buccal gingival recession, periodontal disease in adjacent teeth, uncontrolled diabetes or pregnant/breastfeeding patients, sites of <7 mm in width, sites with buccolingual defects, fenestrations or dehiscence, coagulation disorders, excessive consumption of alcohol or drugs and bruxism, osteoporosis, poor oral hygiene, temporomandibular disorders, previous irradiation at the implant site, psychological disorders, inability of the patient to provide informed consent, unrealistic expectations, implants installed with torque ≤35 N cm, American Society of Anesthesiologists (ASA) status ≥III, and patient was due to receive rehabilitation requiring a prosthesis with a cantilever.
Different medication protocols were recommended before the surgical stage. Oral hygiene was established before the procedure. Antibiotics and antiseptics were given 1 day before the procedure when surgery was proposed for immediate loading. Furthermore, 1 g penicillin and clavulanic acid before surgery and every 12 h for 6 days, or 1 g amoxicillin/clavulanic acid 1 h before surgery and afterwards for 6 days (2 g/day), were proposed; another option was amoxicillin 1 g given 1 h before surgery and 1 g twice a day for 1 week after surgery. As well as these options, a single dose of 2 g of prophylactic antibiotic (amoxicillin/clavulanic acid or phenoxymethylpenicillin potassium) was recommended for use 1 h before surgery, and 1 g amoxicillin/clavulanic acid was recommended 6 h after surgery or 2 g/day of antibiotic (phenoxymethylpenicillin potassium) for 10 days. Ibuprofen (600–800 mg) or ketoprofen (80 mg) was administered for pain control in some studies.
Regarding the distances between the implants and teeth, a minimum distance was observed between tooth and implant and between implant and implant of 2.5 mm, or of 3 mm between implants and 1.5–2 mm between implant and tooth. Another option was to maintain a distance between tooth and implant of 1.5 mm and of 3 mm between implants.
A soft diet and gentle brushing were recommended for a period of time postoperative. Chlorhexidine mouth rinse 0.12% was often used for 1–3 weeks. In addition, the use of chlorhexidine was suggested at 0.2% for different periods.
In some studies, the surgical procedures of immediate extraction were conducted, maintaining the integrity of the lateral walls and avoiding fenestration or dehiscences. In two studies, all surgical procedures were performed by a single surgeon, and in one study, the surgical procedures were performed by any of 12 surgeons, each with more than 10 years of experience. One study used the procedure of maxillary sinus lifting when there was 4 mm of residual bone.
Bone quality was measured in some studies : this was of type I and II ; II and III ; I–IV ; II, III, and IV ; and I, II, and III, or dense (type I), normal (types II and III), and soft bone (type IV).
Number of patients, implants, follow-up, loading type, and location
In this meta-analysis, 1098 patients received 1177 PSW implants and 1104 RP implants (total 2266 implants, as duplicate implants were excluded ). The follow-up period ranged from 12 months to 60 months. Furthermore, five studies considered the procedure of immediate loading, conducting the installation of the implants below the apex by 3 mm or 4 mm , or at 2–3 mm subcrestally. The initial torque was reported to be 32–45 N cm in one study, and ≥35 N cm in three studies. Moreover, in one study, when the distance between the implant and the buccal wall exceeded 1 mm, a mixture of bone matrix and blood was used, and in another clinical study, a mixture of autogenous bone and bovine matrix was used in all cases.
Analysis of the implants
Some studies did not report the type of connection used. Nine studies used the internal connection , two used the external hexagon, and one used the Morse taper. Two systems of connections were used in some studies: external hexagon in the control group and Morse taper in the PSW group ; internal hexagon and external hexagon in the PSW group ; Morse taper in the PSW group and internal hexagon implants in the control group ; and internal hexagon implants in the PSW group compared with external hexagon implants in the control group.
Regarding surface treatments, the following types of implant were installed: surface treatment with micro-roughness, sandblasting and etching, 0.3-mm machined neck, machined neck and micro-threads in the coronal portion, machined neck (0.8 mm) and surface titanium plasma spray, medium-rough surface, TiUnite (Nobel Biocare, Gothenburg, Sweden) surface, dual acid-etched surface, titanium oxide enriched with calcium and phosphorus, SLA surface, and NanoTite (3i Implant Innovations, Palm Beach Gardens, FL, USA).
Timing of loading and final prosthesis
The definitive loading time was at least 2 months and a maximum of 12 months. Fixed prostheses, singles, and full-arch prostheses were used. The occlusal materials employed included ceramic fused to metal crowns in the permanent rehabilitation, full ceramic crowns cemented in a titanium or zirconium abutment, and acrylic resin crowns in temporary rehabilitations.
A few complications were reported, such as loosening of the screw of a temporary abutment and prosthetic screw loosening mainly occurring in the molar area. Kielbassa et al. reported complications in implant-supported prostheses: loosening or loss of cementation of cemented crowns, or screw loosening (provisional). Furthermore, Pieri et al. reported two complications in the control group (RP): screw loosening and a fractured prosthesis, as it was reported that the provisional prosthesis loosened and had to be re-cemented. Furthermore, a titanium abutment was replaced with a zirconium abutment for aesthetic reasons. Other studies with follow-up of 1–5 years did not indicate any type of prosthetic complication.
The authors recommended follow-up visits once every 6 months, or at different times: 1, 3, 6, or 12 months. Vigolo and Givani suggested visits every 3 months during the first year and once every 6 months in subsequent years.
Methodological quality and risk analysis of studies
In an analysis of the quality of all studies included in this systematic review, the Jadad score ranged from 1 to 5. Seventeen RCTs and eight prospective studies were evaluated. The final scores are shown in Table 1 ; the full form of analysis is described in Appendix A .
There are some important concerns. Different models of randomization were employed for blinding of the patients (single-blind study). Authors used predefined tables, a randomization list, or software that generated a list of numbers. Furthermore, randomization in one study was done by letter method. The method of randomization was not reported in some of the RCTs, or there was a randomization of the position of the implants ; other prospective studies did not report a randomization system.
The pre-defined table or block-randomization list for randomization was used efficiently by several studies. In addition, some studies used software and some used envelopes. The system of tables and software allowed for the distribution of patients in the two study groups (PSW and control) according to the variables sex, age, gingival biotype, and tooth position ; localization ; age, sex, and bone type ; sex, age, and smoking habit ; placement area and bone quality ; and sex, age, location of the implant (maxilla or mandible), and number of implants. On the issue of randomization, Kielbassa et al. found that randomization, although adequate, was unable to cover all of the possible interferences, such as the position of the implant.
With regard to blinding, two double-blind studies were included, there were 12 single-blind studies, and 11 studies did not present the data or presented unclear information. A power calculation to determine the appropriate sample size was performed in only eight studies ; other studies included groups/subgroups of small samples of implants (≤15) or patients (≤15).
Some relevant information was missing, such as the connection type selected , the length of implants used, the diameter of the implants employed, the mean age of the patients analysed, the implant localization, the time of final rehabilitation, measures of the implant/abutment relationship, and the abutment diameter used.
The primary outcome was the average bone loss (in millimetres) around the implants. All 25 studies (prospective studies and RCTs) were included in the final sample. In this sample, two studies included subgroups of PSW with various differences between the implant and abutment diameters (Canullo et al. : group 1, 0.25 mm; group 2, 0.5 mm; group 3, 0.85 mm; Canullo et al. : group 1, 0.25 mm; group 2, 0.50 mm; group 3, 0.85 mm). In both cases, group 3 of the PSW sample was selected for comparison with the control group ( Table 1 ). Vandeweghe and De Bruyn showed a total of 15 implants in their clinical study, and the mean overall bone loss was calculated from both the switched and controlled sides. For inclusion, studies had to have a minimum follow-up of 12 months. The follow-up in the studies included in this review ranged from 12 months to 5 years, with a mean longitudinal follow-up of 19.8 months.
In this context, 18 studies showed a statistically significant difference in favour of the use of PSW implants for bone preservation. However, seven studies did not show a significant difference when comparing the control group and PSW group.
The meta-analysis conducted with 17 studies, considering only RCTs, revealed significant bone loss in the control group compared with the PSW group, with a mean difference (MD) of −0.41 mm (95% CI −0.58 to −0.24, P < 0.00001; Fig. 3 ). The χ 2 of heterogeneity was 218.51 ( P < 0.00001, I 2 = 93%).