Keywords
systematic review - meta-analysis - digital dentistry - implant-supported dentures
- intraoral scanners - impression materials
Introduction
The accuracy of dental impression is essential for the predictability of oral rehabilitation
treatment.[1]
[2] That is influenced by several factors, type of tray (stock vs. custom),[3]
[4] quality of the impression material, compatibility with the impression material,
and gypsum pouring technique.[5] Additionally, the accuracy of implant impression techniques can be influenced by
the angulation of the implants, the type of implant-abutment connection, and splinting
material.[6]
[7] However, some of these variables can be eliminated with the advent of digital impressions.[8]
[9]
There are two different types of digital scanners for dental impressions: intraoral
scanners (IOSs) and benchtop scanners.[10]
[11] IOS exhibits advantages such as fewer materials (no need for trays, adhesives, and
dispensers), better control of the aseptic chain (tip can be sterilized), ease of
duplication, and easy storage (there is no expiration date), and recoverability of
files.[5]
[12] Several studies have been published assessing the accuracy of digital impressions
for single unit restorations[13]
[14] and quadrants.[15]
[16]
For instance, Flügge et al evaluated the marginal accuracy of crowns employing digital
impression techniques. As a result, the digital impression showed comparable results
to the conventional impression. Likewise, studies comparing conventional and digital
implants impression have demonstrated similar results between the techniques.[17]
[18]
On the other hand, the use of IOS has shown limitations for full-arch cases. These
limitations are related to the accuracy of the scanning strategy.[19]
[20] Several in vitro studies are comparing the accuracy of IOS for full-arch implant impressions; however,
there are no systematic reviews of these studies to identify if IOS can perform similar
to conventional impressions in full-arch implant impressions. A much-debated question
is whether the current scientific evidence of in vitro studies can determine if IOS is more accurate than impression materials for the impression
of complete implant-supported prostheses. Therefore, the aim of this study was to
perform a systematic review and meta-analysis to answer the following question: What
is the accuracy of IOS compared with conventional full-arch impressions used in the
fabrication of implant-supported prostheses? The research hypothesis of this study
was that there would be no significant differences in the accuracy of impressions
made with IOS or conventional impression materials.
Methods
This systematic review followed the Preferred Reporting Items for Systematic Reviews
and Meta-Analyses,[21] and was registered at the Open Science Framework under the DOI 10.17605/OSF.IO/SH3J2.
The focused question was structured using the PICO acronym. Population (P) were models
with fully edentulous arches, intervention (I) was impressions with IOSs, comparison
(C) was conventional impressions with impression materials, outcomes (O) were linear
and angular deviations, and the type of study (S) was comparative in vitro experimental studies. To answer the focused question, inclusion criteria consisted
in vitro study that compared conventional and IOS impressions in full arches for implant-supported
prostheses. Papers written in any Roman Latin language were included. Conversely,
observational studies, studies that evaluated partial arches, that did not compare
digital impressions (intraoral) and conventional impressions, as well as reviews,
letters, abstracts, and case reports were excluded. The main search strategy was formulated
and applied in the PubMed (MEDLINE) database. Furthermore, the main search strategy
was used as a reference and applied to the following databases: Scopus, Web of Science,
LILACS, Cochrane Library, and EMBASE in April 2022. Additionally, a gray literature
search was performed on Google Scholar, Clinical Trials, Open Gray, and ProQuest.
Besides, a manual search on the Internet and in references within the articles was
performed ([Supplementary Table S1]). Two independent reviewers participated from the first phase of the study in the
article's selection based on the title and abstracts' information using online software
for systematic reviews (Rayyan, Qatar Computing Research Institute). After that, the
intra- and inter-examiners calibration level was performed with the first 10% of the
references. The acceptable level of agreement (> 7.0) among reviewers was achieved,
and the second phase of the study (full-text) was performed ([Fig. 1]).
Fig. 1 Flowchart of the identification, screening, and inclusion of the studies.
Data extraction was performed independently by two reviewers for the primary outcomes
and compared collected data. The following data were extracted from the eligible studies:
the purpose of the study, study details, sample features, scanning methods, measurement
details, findings, and conclusion ([Table 1]). Two independent and blinded reviewers performed the risk of bias using an adapted
Checklist for Reporting In vitro Studies tool.[22] The tool comprised questions related to five domains: sample size calculation, sample
preparation and handling, blinding, statistical analysis, and limitations and potential
bias of the study. Each question was scored with “yes,” “no,” or “unclear.” Studies
that fulfilled all questions were determined to be high-quality studies (low risk
of bias), whereas studies containing 3 or 4 “yes” scores were considered medium-quality
articles (moderate risk of bias) ([Table 2]). Furthermore, studies with fewer than 3 “yes” scores presented low methodological
quality (high risk of bias). The analysis was performed using a software program (RevMan
5.4; The Nordic Cochrane Center) ([Fig. 2]).
Table 1
Characteristics of each included study
|
|
|
Groups
|
|
|
Control group
(convention impression)
|
Experimental Group
(digital impression)
|
|
Study, year
Country
|
N total
(master model material)
|
Jaw status
|
Material used Technique used Pouring material
|
Measurement
(n)
|
Scanner used
(n = total scannings)
Scanning technique
|
Measurement
|
Main conclusions
|
|
Abdel-Azim et al, 2014
USA
|
1
(NR)
|
Edentulous mandible and maxilla
|
Polyether (Impregum, 3M ESPE, USA)
ClosedTray
Type IV Stone (Resin Rock, Whip Mix)
|
Image Pro version 6.2.1.491 (Media Cybernetics, USA)
(n = 2)
|
iTero (Cadent)
(n = 6)
Unclear
|
Infrastructure level (Marginal discrepancy was assessed with the stereomicroscope
at two points for each retainer on each model, buccally and lingually)
|
For complete-arch frameworks, the digital impression/fabrication technique resulted
in an overall more accurate fit when compared with the conventional impression/fabrication
method
|
|
Albaryak et al, 2021
USA
|
1
(Polyurethane)
|
Edentulous mandible
|
Polyvinyl siloxane, Elite HD (Zhermack SpA, Italy)
OpenTray
Type IV Stone (Fujirock, Japan)
|
Unclear
|
Carestream 3500,
Cerec Omnicam;
3Shape Trios
(n = 30)
Right posterior region and continued toward left posterior region on the opposite
side of the arch and the scan paths were
determined according to the instructions
|
STL level
(The reference points were determined on the scan bodies. Two circles were created
at 0.7 and 3.4 mm from the triangular pyramid base to center the upper and lower parts
of the scan bodies. The center points of these circles where the distance measurements)
|
Complete arch implant case with high angulations and asymmetric distribution, digital
impression methods achieved superior results in both distance and angular parameter
than conventional method using nonsplinted open tray impression technique. Besides
different acquisition methods and working principles of IOS can affect the accuracy
|
|
Alikhasi et al, 2018
Iran
|
2
(Acrylic resin)
|
Edentulous mandible and maxilla
|
Vinyl siloxanether (Zhermack Elite HD + Regular Body, Italy)
Open tray/closed tray
Type IV Stone (Herostone, Vigodent Inc., Brazil)
|
Coordinate-Measuring Machine (Mistral, DEA Brown&Shape, Italy)
(n = 3)
Start to scan in the palatal of the right tuberosity and lingual surfaces of all scan
bodies. Next, the buccal and then the occlusal
|
Trios (3Shape)
(n = 60)
Unclear
|
STL level
(Superposition of STL data sets [best-fit alignment tool])
|
Digital impression is better than the direct technique in the edentulous arch with
straight and tilted implants, and both are more accurate than the indirect technique
|
|
Amin et al,
2017
USA
|
1
(Type IV Stone)
|
Edentulous mandible and maxilla
|
Impregum, 3M ESPE, USA)
Open tray
Type IV Stone
(Resin Rock, Whip Mix)
|
Activity 880 scanner (Smart Optics, Germany)
(n = 1)
|
CEREC Omnicam
True Definition
(n = 20)
Started at the right retromolar. Continuous stroke was completed along the occlusal
surface until the left retromolar and lastly buccal scan
|
STL level
Superposition of STL data sets (best-fit alignment tool)
|
Omnicam was more accurate than conventional impressions (splinted open-tray technique).
True Definition scanner had significantly less 3D deviations when compared with the
Omnicam (precision)
|
|
Kim et al, 2019
South Korea
|
1
(Epoxy resin)
|
Edentulous maxilla
|
Vinyl siloxanether
(Aquasil XLV; Dentsply Sirona); plastic tray: high-viscosity silicone impression material
(Aquasil EasyMix Putty)
Open tray
Type IV dental stone (MG Crystal Rock, Maruishi)
|
Unclear
|
TRIOS 3 (3Shape, Denmark)
(n = 10)
Scanning started from the occlusal at the right molar, continued to the scan body
at the contralateral left first molar area, then to the palatal, and finally to the
buccal
|
STL level
(Superposition of STL data sets [best-fit alignment tool])
|
Intraoral digital scan resulted in less accurate trueness than the conventional open-tray
impression technique in terms of overall linear displacement. The conventional open-tray
impression technique resulted in more accurate precision for all the implant replica
locations, and produced significantly smaller angular deviations compared with the
intraoral digital scan
|
|
Menini et al, 2018
Italy
|
1
(Metal framework)
|
Edentulous maxilla
|
Polyether (Impregum, 3M ESPE, USA)
Open tray/Closed tray
|
|
True Definition (3M ESPE)
(n = 35)
Unclear
|
Infrastructure level
(Marginal discrepancy was assessed with the microscope)
|
The use of an intraoral digitizer might represent a viable alternative to traditional
impression materials for the fabrication of full-arch implant-supported prostheses
provided with a satisfactory passive fit
|
|
Papaspyridakos et al,
2016
USA
|
1
(Acrylic resin)
|
Edentulous mandible
|
Group I: splinted coping impression technique at the implant level
Group II: nonsplinted coping impression technique at the implant level
Group IV: from the splinted coping impression technique at the abutment level
|
|
Trios (3Shape)
Group III: from the digital impression technique at the implant level
|
One digital scan of the master cast at the implant level and one scan at the abutment
level, with the same high-resolution extraoral scanner at 6 lumens precision (IScan
D103i; Imet- ric), were used as control (golden reference)
|
The implant-level, splinted impressions were more accurate than the nonsplinted conventional
impressions for completely edentulous patients
• The accuracy of abutment-level, splinted impressions were not different than the
nonsplinted impressions for completely edentulous patients
• The accuracy of implant impressions is not affected by the implant angulation up
to 15 degrees for completely edentulous patients. The connection type seems to affect
accuracy because abutment-level impressions had no statistically significant differences
from the control, whereas differences were identified for the implant-level, nonsplinted
impressions. Seems to affect accuracy because abutment-level impressions had no statistically
significant differences from the control
|
|
Rech-Ortega et al,
2019
Spain
|
1
(Titanium)
|
Edentulous mandible
|
Polyether (Impregum Penta Soft, 3M ESPE) Type IV plaster (Vel Polyether, Impregum
Penta Soft, 3M ESPE) Type IV plaster (Vel MixStone, Kerr)/Type IV plaster (Vel- MixStone,
Kerr)
|
Dental Designer, 3Shape
(n = 120)
|
True Definition (3M ESPE)
Unclear
|
Six cylindrical scan bodies were screwed onto the analogues and the following distances
were measured: between adjacent analogues from center to center: 1–2, 2–3, 3–4, 4–5,
5–6; between intermittently positioned analogs: 1–4, 3–6; between the most distal
analogs: 1–6. This process was repeated for the 20 physical models. Five parameters
(X, Y, Z, module XY, module XYZ) were calculated for each distance 1–2, 2–3, 3–4,
4–5, 5–6, 1–4, 3–6, 1–6 measured from the 20 physical models elastomeric impression
material (EIM) and the master model
|
For adjacent analogs, the direct technique can be considered the most accurate for
the XYZ module distance 1–2 as no statistically significant differences were found
(p = 0.146) in relation to the master model. For the other distances, 2–3, 3–4, 4–5,
and 5–6, neither technique was completely accurate. True Definition (3M ESPE) (SDM)
provided accurate data, without statistically significant differences in comparison
with the master model (p = 0.255)
|
|
Tan et al, 2019
Singapore
|
2
(PMMA)
|
Edentulous maxilla
|
Polyether (Impregum Duosoft, 3M ESPE, Germany)/Type IV dental stone models (Silky
Rock, Whip Mix Co.)
|
Laboratory scanner (Straumann CARES Scan CS2 Visual 8.0 software, Institut Straumann
AG, Basel, Switzerland)
(n = 6)
|
Trios 3Shape (3Shape)
True Definition (3M ESPE)
Unclear
|
The axial portion of each scan body was measured via eight virtual probe hits at two
levels to define a cylinder (Trios, True Definition, inEos X5, and D900) or cone (Ceramill
Map400)
|
Reducing interimplant distance may decrease global linear distortions (dR) for intraoral
scanner systems but had no effect on Impregum and the dental laboratory scanner systems.
Impregum consistently exhibited the best or second-best accuracy at all implant locations,
while True Definition exhibited the poorest accuracy for all linear distortions in
both models A and B. Impression systems could not be consistently ranked for absolute
angular distortions
|
Abbreviations: 3D, three-dimensional; IOS, intraoral scanner; PMMA, polymethyl methacrylate;
STL, stereolithography.
Table 2
Quality assessment and risk of bias of eligible studies with adapted Checklist for
Reporting In vitro (CRIS guidelines)[22] tool
|
First author/Year
|
Sample size calculation
|
Detailed sample preparation and handing
|
Allocation sequence
|
Statistical analysis
|
Addressing limitations of study and potential bias
|
Estimated potential risk of bias
|
|
Abdel-Azim, 2014
|
N
|
Y
|
Y
|
Y
|
Y
|
Y
|
|
Alikhasi, 2018
|
N
|
Y
|
Y
|
Y
|
Y
|
Y
|
|
Albaryak, 2021
|
N
|
Y
|
Y
|
Y
|
Y
|
Y
|
|
Amin, 2017
|
N
|
Y
|
Y
|
Y
|
N
|
N
|
|
Menini, 2018
|
N
|
Y
|
Y
|
Y
|
Y
|
N
|
|
Papaspyridakos, 2016
|
N
|
Y
|
Y
|
Y
|
N
|
N
|
|
Kim, 2019
|
N
|
Y
|
Y
|
Y
|
N
|
N
|
|
Rech-Ortega, 2019
|
N
|
Y
|
Y
|
Y
|
Y
|
N
|
|
Tan, 2019
|
N
|
Y
|
Y
|
Y
|
N
|
N
|
Fig. 2 Qualitative analysis with adapted checklist for reporting in vitro
[25]
The date for the linear accuracy of the nine included studies was obtained. The groups
were divided based on the IOS system used in the study. The mean and standard deviation
for each IOS reported in the study were used to perform the meta-analysis model. All
data were imported to statistical analysis software (Stata/MP 17, StataCorp, College
Station, Texas, United States). A random-effect Hunter–Schmidt model was used for
the meta-analysis to compare the mean and standard deviation of each study's conventional
and digital implant impression systems for each IOS at a significance level of α = 0.05.
Results
The initial search resulted in 3,018 articles, which was reduced to 1,255 after removing
duplicate reports. After study selection, 31 articles were selected for full-text
reading, and 9 studies were considered eligible for this review in April 2022 ([Fig. 1]). Because of the heterogeneity of the parametric data in the studies regarding the
trueness, precision, and angular accuracy, the meta-analysis was performed only with
linear accuracy. The reasons for exclusion can be found in [Supplementary Table S2].
The dimensional distortion of trueness and precision was described using deviation/discrepancy
values.[21] This is done through a different methodology of IOS measurement technique, four
studies used superposition of stereolithography (STL) data sets (best-fit alignment
tool), which calculate the horizontal linear distance and the marginal discrepancy.
Other studies analyzed the measurement technique through contact feature mode and
marginal discrepancy evaluated with a microscope. Another intervention characteristic
was the IOS measurement level executed on the infrastructure level, STL level, and
cast level.
The majority of studies included only open tray conventional impressions,[1]
[20]
[23]
[24] except for two studies[4]
[24] which compared both closed tray and open tray to digital scans, one study used only
close tray,[25] unclear in two studies,[2]
[26] and one study[15] evaluated splinted coping impression technique at the implant level, nonsplinted
coping impression technique at the implant level, and splinted coping impression technique
at the abutment level. Vinyl polysiloxane and polyether were used as impression materials
for the conventional methods.
The accuracy in conventional impression from in vitro ranged from 0.46 to 573.63 µm, and IOS impression ranged from 0.56 to 579.92 µm ([Table 3]). Overall, the digital impressions present higher accuracy (137.86 µm) than conventional
impressions (182.51 µm) (p < 0.001; mean difference with 95% confidence interval of 30.06 µm [–36.11 to –24.00]).
The heterogeneity of the study's meta-analysis was 18.01%, which suggests that the
heterogeneity is insignificant, and the summarization of the meta-analysis results
is statistically appropriate([Fig. 3]).[27] Out of the nine studies that were included in the quantitative analysis, five studies[4]
[15]
[20]
[23]
[26] used Trios (3Shape, Copenhagen, Denmark), four studies[1]
[2]
[24]
[26] used True Definition (3M ESPE, St Paul, Minnesota, United States) IOS, two studies[1]
[20] used Cerec Omnicam (Dentsply Sirona, Bensheim, Germany), one study[20] used Carestream 3500 (Carestream Dental LLC, Atlanta, Georgia, United States), and
one study[25] used iTERO (Align Technology, San Jose, California, United States).
Table 3
Outcomes of in vitro included studies
|
Control group (conventional impression)
|
Experimental group (intraoral scanning)
|
|
Author, year
|
Trueness
distortion
(mean ± SD)
|
Precision
distortion
(mean ± SD)
|
Trueness
distortion
(mean ± SD)
|
Precision
distortion
(mean ± SD)
|
|
Abdel-Azim et al, 2014
|
Linear:
250.04 ± 194.85 µm
Angular: NR
|
Linear: NR
Angular: NR
|
Linear:
3Shape Trios: 71.50 ± 30.40 µm
Angular: NR
|
Linear: NR
Angular: NR
|
|
Albaryak et al, 2021
|
Linear:
345.32 ± 75.12 μm
Angular:
0.74° ± 0.65°
|
Linear: 66.97 ± 36.69 μm
Angular: 0.50° ± 0.38°
|
Linear:
Carestream: 123.06 ± 89.83 μm
Cerec Omnicam: 229.72 ± 121.34 μm
3Shape Trios: 209.75 ± 47.07
Angular:
Carestream: 0.26° ± 0.07°
Cerec Omnicam: 0.53° ± 0.42°
3Shape Trios: 0.33° ± 0.30°
|
Linear:
Carestream: 80.43 ± 29.69 μm
Cerec Omnicam: 94.06 ± 69.96 μm
3Shape Trios: 35.55 ± 28.46 μm
Angular:
Carestream: 0.19° ± 0.11°
Cerec Omnicam: 0.30° ± 0.28°
3Shape Trios: 0.22° ± 0.19°
|
|
Alikhasi et al, 2018
|
Linear: 573.62 µm
Angular: 36.92°
|
Linear: NR
Angular: NR
|
Linear:
3Shape Trios 177.50 µm
Angular: 0.475°
|
Linear: NR
Angular: NR
|
|
Amin et al, 2017
|
Linear: 167.93 ± 50.37 µm
Angular: NR
|
Linear: NR
Angular: NR
|
Linear:
Omnicam: 46.41 ± 7.34 µm
True Definition: 19.32 ± 2.77 µm
Angular: NR
|
Linear: NR
Angular: NR
|
|
Papaspyridakos et al, 2016
|
Linear: 110.58 µm
Angular: NR
|
Linear: NR
Angular: NR
|
Linear:
3Shape Trios: 19.38 µm
Angular: NR
|
Linear: NR
Angular: NR
|
|
Menini et al, 2018
|
Linear: 42.5 ± 28.99 µm
Angular: 0.316 ± 0.167°
|
Linear: NR
Angular: NR
|
Linear:
True Definition: 17 ± 2.83 µm
Angular:
True Definition: 0.257° ± 0.242°
|
Linear: NR
Angular: NR
|
|
Kim et al, 2019
|
Linear: 72.2 ± -µm
Angular: 1.19 ± -°§
|
Linear: 26.25 µm
Angular: NR
|
Linear:
3Shape Trios: 177.4µm
Angular:
3Shape Trios: 1.55°
|
Linear:
3Shape Trios: 80.76 µm
|
|
Rech-Ortega et al, 2019
|
Linear: 15.3 µm
Angular: NR
|
Linear: NR
Angular: NR
|
Linear:
True Definition: 28.75 µm
Angular: NR
|
Linear: NR
Angular: NR
|
|
Tan et al, 2019
|
Linear: 0.10 μm
Angular: 20.95
|
Linear: NR
Angular: NR
|
Linear: –0.837 μm
Angular: 288.34°
|
Linear: NR
Angular: NR
|
Abbreviations: NR, not reported; SD, standard deviation.
For the Trios system, the mean accuracy difference was –48.62 µm (–131.39, 34.16)
to the control group. Two studies[4]
[20] demonstrated better accuracy, and one study[15] showed no differences between Trios and conventional impressions. However, two studies[23]
[26] demonstrated lower accuracy for Trios when compared with conventional impressions.
The mean accuracy difference for the True Definition system was –16.62 µm (–22.00,
–11.24) to the control group. Two studies[1]
[24] demonstrated better accuracy, and one study[2] showed no differences between True Definition and conventional impressions. However,
one study[26] demonstrated lower accuracy for True Definition when compared with conventional
impressions. For Omnicam, the mean accuracy difference was –121.17 µm (–142.80, –99.53)
to the control group. The two included studies[1]
[20] demonstrated better accuracy for the Omnicam than conventional impressions. For
Carestream 3500[20] and iTero[25] the included studies showed better accuracy, and the accuracy mean differences were
–222.26 µm (–294.84, –149.68) and –178.54 µm (–336.34, –20.74), respectively. The
accuracy of 200 μm is reported as the maximum clinically acceptable discrepancy threshold.[28] All studies reported a mean accuracy difference from the control group below 200
μm, with the exception of the True Definition in the Tan et al[26] study.
Discussion
This systematic review was based on nine in vitro studies. The data from these studies have an informative value for the clinician
and provide substantial evidence on the different techniques for impression (conventional
vs. digital) of full-arch impression for implant-supported prostheses. The research
hypothesis of this study that there will be no significant differences in the accuracy
of impressions made with IOS or conventional impression materials was rejected. In
general, most IOS and conventional methods had trueness and precision values acceptable
for clinical use. The studies showed that in conventional methods, the trueness and
precision are influenced by impression material, impression technique, cast material,
pouring technique, and measuring technique used by the researcher to evaluate the
discrepancy.[5]
[12]
[13] The deviation values differed in digital methods according to the scanner system,
scanning method, measurement technique, merging technique, scanning strategies, and
learning curves of IOS.[29]
[30]
The scanners are devices that digitize intraoral conditions in a three-dimensional
(3D) file—these exhibit formats with three magnitudes that are important for becoming
the digital impression valuable in dentistry. First, accuracy is defined (International
Organization for Standardization 5725–1) in terms of trueness and precision.[31] The trueness is the deviation of the object scanned with an IOS from its real geometry;
this means that accuracy is described by the mean of the discrepancies between the
object scanned and the IOS scans of the object (target). Lastly, precision represents
the discrepancies between the repeated scans of the same object performed with the
same IOS and parameters. It represents how much a measurement could be systematically
repeated. Precision is described by the mean of the discrepancies among the various
scans of the object performed with the IOS.[32]
In other words, the accuracy of digital (either intraoral or desktop) and conventional
impressions was defined as the closeness agreement between a data set and the accepted
reference value divided into the deviation between measured dimensions and actual
dimensions of the object (trueness) and closeness of repeated measurements (precision).[32]
According to Abdel-Azim et al, conventional impression showed a high linear trueness
deviation value of 250.04 µm. The reason the conventional impression group was less
trueness than the digital impression groups in this study may be associated with the
nonsplinted open tray technique. Still, the Trios IOS accuracy of 71.50 ± 30.40 µm.
Although there is not yet a scientific consensus, a value of 200 µm has been reported
by the literature as acceptable in terms of prosthesis misfit.[33]
In 2018, Alikhasi et al compared the trueness of the nonsplinted open tray and closed
tray techniques using polyvinyl siloxane and the digital impressions made with the
3Shape Trios scanner. The impressions were made from four implants, with two straight
and two 45 degrees distally tilted implants in an edentulous maxillary model. The
results of the comparison showed that the digital technique was found to be the most
trueness with a 177.50-μm linear deviation. The trueness of the open tray technique
was 280 μm, and the closed tray technique showed the highest deviation with 885 μm.[4] A systematic review published by Papaspyridakos et al regarding conventional implant
impressions stated that splinting significantly increases impression accuracy, especially
in partial arch implant cases.[3]
[23]
Finding the reference point for IOS is challenging for digital implant impressions
of full arches. Alikhasi et al present higher discrepancies (177.50 µm) in edentulous
maxilla full-arch with implant abutments for the IOS. The IOS measurement technique
was a superposition of .stl data sets (best-fit alignment tool), and the IOS measurement
level was .stl. Having these different levels (implant analog level or scan body level)
where the files are merged and measured, a fair comparison between study results is
hard to execute. This scanning strategy was different from the other select studies.[2]
[23] The authors started scanning from the palate's reference pin toward all scan bodies'
right tuberosity and lingual surfaces.[4] Next, they scanned the scan bodies' buccal surfaces and occlusal surfaces. Accordingly,
accuracy varies significantly with scan body type, IOS type, scanning strategy, and
modification technique. One possible reason that this study, Alikhasi et al, showed
more deviation was related to the scanning strategy.
Müller et al have investigated the effects of scanning protocols and confirmed that
scanning accuracy differs depending on the protocol used. However, no one considered
the effects of the rotation of the IOSs on their accuracy when performing full-arch
scans.[30] Oh et al, in an in vitro study, assessed the effects of different scanning strategies on the accuracy of the
scanned data. According to the authors, a segmental approach to scanning the region
of interest first or a continuous scan method with the scanner head kept mostly in
a horizontal position throughout the scanning can be used to obtain the full arch
scan data. However, the rotation of IOS in the vertical direction should be minimized
because it affects the accuracy of the stl.[34] Stefanelli et al, in an in vitro study, developed two scanning strategies for a full dental arch. However, the literature
has no consolidated approach for a full arch in digital implant impression.[32]
Similarly, Rech-Ortega et al compared the trueness of two techniques, a direct (or
pick-up) technique with the elastomeric material polyether and a digital scanner (True
Definition, 3M ESPE), by comparing measurements between implant replicas. Neither
technique can be considered accurate in rehabilitations involving more than four implants.
However, both techniques analyzed can be used with relative reliability, as the errors
produced fell within the tolerance range established in the literature as acceptable
(30–150 μm),[2] although it is advisable to make a verification splint before fabricating the definitive
prosthesis.[2] Another reason for this linear and angular distortion presented in the studies is
the methodology to assess the IOS's accuracy; the software works by making a comparison
between the IOS .stl file and the reference scan's STL file using the best-fit algorithm function; this
matching generates linear deviations between the two data sets that can be measured.[32]
Chochlidakis et al, in the first clinical study, compared the trueness of digital
and conventional maxillary implant impressions for fully edentulous patients. They
concluded a positive correlation between implant number and 3D deviation. So, the
angular deviation increases with the number of the implant. According to this study,
a digital impression will be a challenge in full-arch cases. D'haese et al evaluated
the accuracy of full-arch digital impressions compared with conventional impressions
when performed on the abutment or implant level. In this study, the abutment level
impressions were more accurate than implant level impressions because of the flat
connection geometry, which provides a vertical stop.[35]
Digital impressions for full-arch implant-supported prostheses can be as accurate
as conventional impressions, depending on the IOS and software.[9]
[15] Previous in vitro studies investigating the effects of scanning protocols have also confirmed that
scanning accuracy differs depending on the protocol used.[9]
[15] Kim et al[23] used 3D displacement of implant replicas and not those of scan bodies. The discrepancies
are attributed to the inherent errors in converting the scan body position to the
implant replica position using a digital library. Several commercial brands offer
scan bodies for the different implant and IOS systems. Although there is adequate
accuracy when comparing scan bodies' position in a virtual model, the algorithms in
digital planning software still need development to position the implant replicas
accurately.
Lastly, one study[26] showed that the True Definition Scanner might not achieve a clinically acceptable
accuracy threshold as an exception. The clinician should be aware that Midmark purchased
the True Definition trademark from 3M, but the new company did not report significant
improvements in IOS. Digital scanning of full arches for implant-supported complete
prostheses is promising and clinically acceptable; this systematic review ensures
that IOS can be more accurate than conventional impression materials.
Conclusion
Within the limitation of this study, it can be concluded that digital impressions
using IOS present similar or better linear accuracy than conventional impression techniques.
All the IOS included in this study presented acceptable accuracy (below the 200 µm
threshold), except for the True Definition Scanner in one of the studies. The scan
body lack of optimization might be attributed to divergence in the data.
Fig. 3 Forest plot for the accuracy of digital (treatment) and conventional (control) impressive
techniques. The overall heterogeneity and p-values are based on a random-effects Hunter-Schmidt
meta-analysis model with a level of significance of 0.05
