In Figure 1, the screening procedure is depicted. By electronic means searching, 129 (PubMed: 62; Web of Science: 53; Embase: 14; ClinicalTrials.gov: 0) were selected. After duplicates removal, the title and abstracts of 90 studies were looked at. Of these, 25 articles were removed including meta-analyses, reviews, case series/reports, in vitro studies, and animal experiments, and 58 irrelevant articles were also excluded. Seven articles were subjected to full-text evaluation, and the reasons listed in Table 1 led to the exclusion of three22-24 of them. As a result, this meta-analysis and systematic review included four human studies25-28.
Description of Studies
Tables 2–4 provide a summary of the included studies' primary characteristics. The sockets in one article were in the posterior maxilla areas26, and two studies were in the regions of maxillary anterior teeth25, 27, and one article were in the anterior teeth28,before ridge preservation. The follow-up period lasted 4 to 6 months in all included articles. Among the four articles included, one article took deproteinized bovine bone mineral with 10% collagen and EMD and covered with free gingival graft (DBBMC-EMD+FGG) as the test group and DBBMC+FGG as the control group25; one article took DBBMC and EMD and covered with two layers of a resorbable collagen membrane (DBBMC-EMD+RCM) as the test group and DBBMC+RCM as the control group26; one article took DBBMC and EMD and covered with two layers of a native bilayer non‐cross‐linked resorbable collagen membrane (DBBMC-EMD+NCLM) as the test group and DBBMC+NCLM as the control group27; and another article took alloplastic bone substitute (BoneCeramic [BC]) combined with EMD, and covered with lingual or palatal flap (BC-EMD+LP) as the test group and BC +LP as the control group28
Fig 1 Illustration of the article selection process
Table 1 Excluded Articles with Reasons 22-24
|
Study(year)
|
Articles excluded with justification
|
Alkan, E. A. et al (2016)22
|
Incomplete data
|
Alkan, E. A. et al (2013)23
|
Incomplete data
|
Nevins, M. L. et al (2011)24
|
Incomplete data
|
Table 2 Characteristics of the included articles: Change in Ridge Height25-27
|
Author
(year)
|
Article type
|
Recipient location
|
Groups
(test/control)
|
Number
(test/control)
|
Duration
|
Change in the test group's bone height,
mm, mean ± SD
|
Change in the control group's bone height,
mm, mean ± SD
|
Statistical difference between two groups
|
Faustino Mercado
(2021)25
|
RCCT
|
anterior maxilla
|
DBBMC-EMD+FGG/
DBBMC+FGG
|
21/21
|
4 M
|
Buccal: −0.62 ± 0.9
Palatal: −1.0 ± 1.2
|
Buccal: −1 ± 1.2
Palatal: −0.8 ± 1.0
|
NSSD
|
Jae-Hong Lee
(2020)26
|
RCCT
|
posterior maxilla
|
DBBMC-EMD+RCM/
DBBMC+RCM
|
10/10
|
5 M
|
Buccal: −0.61 ± 0.40
Palatal: −0.69 ±0.62
|
Buccal: −0.90 ± 0.65
Palatal: −0.60 ± 0.37
|
NSSD
|
Jae‐Hong Lee
(2019)27
|
RCCT
|
anterior maxilla
|
DBBMC-EMD+NCLM/
DBBMC+NCLM
|
15/15
|
5 M
|
Buccal: −2.99 ± 1.22 Palatal: −1.21± 0.36
|
Buccal: −3.30 ± 1.48
Palatal: −1.29 ± 0.46
|
NSSD
|
RCCT=randomized controlled clinical trial; DBBMC=deproteinized bovine bone mineral with 10% collagen (Bio- Oss Collagen®, Geistlich); EMD=enamel matrix derivative (Emdogain®, Straumann); FGG=free gingival graft; RCM=resorbable collagen membrane; NCLM=non‐cross‐linked resorbable collagen membrane; M=month; SD=standard deviation; NSSD=no statistically significant difference.
Table 3 Characteristics of the included articles: Change of Ridge Width 25-27
|
Author
(year)
|
Article type
|
Recipient location
|
Groups
(test/control)
|
Number
(test/
control)
|
Duration
|
RW
(mm)
|
Change in the test group's bone width,
mm, mean ± SD
|
Change in the test group's bone width,
mm, mean ± SD
|
Statistical difference between two groups
|
Faustino Mercado
(2021)25
|
RCCT
|
anterior maxilla
|
DBBMC- EMD +FGG /
DBBMC+FGG
|
21/21
|
4 M
|
2
|
−0.40 ± 0.5
|
−0.37 ± 1.1
|
NSSD
|
Jae-Hong Lee
(2020)26
|
RCCT
|
posterior maxilla
|
DBBMC- EMD+RCM/
DBBMC+RCM
|
10/10
|
5 M
|
1,3,5
|
1mm: −1.44 ± 0.54
3mm: −1.21 ± 0.52
5mm: −0.86 ± 0.81
|
1mm: −1.42 ± 0.26
3mm: −1.34 ± 0.72
5mm: −0.50 ± 0.61
|
NSSD
|
Jae‐Hong Lee
(2019)27
|
RCCT
|
anterior maxilla
|
DBBMC- EMD+NCLM/
DBBMC+NCLM
|
15/15
|
5 M
|
1,3,5
|
1mm: −1.57 ± 0.45
3mm: −0.92 ± 0.45
5mm: −0.51 ± 0.21
|
1mm: −1.60 ± 0.30
3mm: −1.05 ± 0.55
5mm: −0.56 ± 0.31
|
NSSD
|
RCCT=randomized controlled clinical trial; DBBMC=deproteinized bovine bone mineral with 10% collagen (Bio- Oss Collagen®, Geistlich); EMD=enamel matrix derivative (Emdogain®, Straumann); FGG=free gingival graft; RCM=resorbable collagen membrane; NCLM=non‐cross‐linked resorbable collagen membrane; M=month; SD=standard deviation; NSSD=no statistically significant difference.
Table 4 Characteristics of the included articles: Histomorphometric Analysis25,28
|
Author
(year)
|
Article type
|
Recipient location
|
Groups
(test/control)
|
Number
(test/
control)
|
Duration
|
%NB
mean ± SD
|
%RG
mean ± SD
|
%STM
mean ± SD
|
Test
|
Control
|
Test
|
Control
|
Test
|
Control
|
Faustino Mercado
(2021)25
|
RCCT
|
anterior maxilla
|
DBBMC-EMD+FGG/
DBBMC+FGG
|
21/21
|
4 M
|
45.1 ± 8.8
|
16.5 ± 6.9
|
20.3 ± 7. 2
|
36.8 ± 8.8
|
34.6 ± 13.8
|
46.5 ± 10.4
|
Bonta, Hernan(2022)28
|
RCCT
|
anterior
tooth
|
BC-EMD+LP /
BC+LP
|
7/7
|
6 M
|
47.30± 9.18
|
32.27±9.11
|
18.12±7.42
|
11.61±13.75
|
34.57±9.43
|
57.21±14.65
|
BC=BoneCeramic®, Straumann; LP=lingual or palatal flap; NB=new bone formation; RG=residual graft; STM=soft tissue and marrow spaces
Risk of Bias
The four RCCTs were evaluated using the Cochrane literature quality evaluation tool for random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting and other bias. The quality evaluation chart was drawn by Review Manager software. Quality evaluation diagram is shown in Fig 2. The overall bias result is that one article is low-risk27 and the other three are medium-risk25, 26, 28. The 3 medium-risk articles were nonblind in the blinding of participants and personnel. However, it is challenging to blind participants and staff members during clinical trials.
1. Change in Alveolar Ridge Height
Table 2 shows an evaluation of the three articles25-27 included's bone height. In all three studies, the changes in buccal and palatal bone height were assessed with cone beam computed tomography (CBCT). According to the meta-analysis, comparing with the control group, the experimental group’s buccal bone height resorption was 0.32 mm lower than the control group, but this difference was not statistically significant; There was also no statistically significant difference in the test and control groups’ absorption of palatal bone height.
1.1 Meta-analysis of the height changes of buccal bone
We can choose the fixed effect for the meta-analysis since I2 = 0. 0% < 50%, and P=0.98 > 0.1, demonstrating that there is no heterogeneity among the selected studies25-27. Comparing with the control group, the experimental group’s buccal bone height resorption was 0.32 mm lower, according to the combined effect of Meta, which was 0.32 (-0.03~0.67). However, neither group was significantly different from the other (z=1.77 p=0.08 > 0.05, figure3). A funnel diagram(figure4) was made to test the three articles' publication bias. Furthermore, the figure 4 bias test indicates that there is no publication bias in the research literature (P= 0.766 > 0.05).
1.2 Meta-analysis of the height changes of palatal bone
We can choose the fixed effect for the meta-analysis since I2 = 0. 0% < 50%, and P=0.68 > 0.1, demonstrating that there is no heterogeneity among the selected studies25-27. The experimental group’s palatal bone height resorption was the same as that in the control group, according to the combined effect of Meta, which was 0.00 (-0.23~0.23). However, neither group was significantly different from the other (z=0.01, p=0.99>0.05, figure5). A funnel diagram(figure6) was made to test the three articles' publication bias. Furthermore, the figure6 bias test indicates that there is no publication bias in the research literature (P= 0.144 >0.05).
2. Change in Alveolar Ridge Width
The bone width of the three included articles25-27 was evaluated (Table 3). CBCT was used to evaluate the changes in buccal and palatal bone width in all three studies. In one article, 2 mm below the alveolar crest was used as the reference point to measure the width of the ridge (RW). In the other two articles, the authors measured the width of the ridge at reference points of 1mm, 3mm and 5mm under the alveolar crest. Therefore, in this review, changes in RW2 with RW1 and RW2 with RW3 were selected for meta-analysis respectively.
Comparing with the control group, the meta-analysis of RW2 with RW1 showed that the experimental group’s absorption of ridge width was 0.01mm lower, however, neither group was significantly different from the other. Comparing with the control group, the meta-analysis of RW2 with RW3 showed that the experimental group’s absorption of ridge width was 0.09mm lower, however, neither group was significantly different from the other.
2.1 Meta-analysis of RW2 with RW1
We can choose the fixed effect for the meta-analysis since I2 = 0. 0% < 50%, and P=0.97 > 0.1, demonstrating that there is no heterogeneity among the selected studies25-27. Comparing with the control group, the experimental group’s bone width resorption (RW2 with RW1) in the test group was 0.01mm lower, according to the combined effect of Meta, which was 0.01 (-0.20~ 0.21). However, neither group was significantly different from the other (z=0.06, p=0.95>0.05, figure7). A funnel diagram(figure8) was made to test the three articles' publication bias. Furthermore, the figure8 bias test indicates that there is no publication bias in the research literature(P=0.283>0.05).
2.2 Meta-analysis of RW2 with RW3
We can choose the fixed effect for the meta-analysis since I2 = 0. 0% < 50%, and P=0.87 > 0.1, demonstrating that there is no heterogeneity among the selected studies25-27. Comparing with the control group, the experimental group’s bone width resorption (RW2 with RW3) was 0.09mm lower, according to the combined effect of Meta, which was 0.09 (-0.17~0.35). However, neither group was significantly different from the other (z=0.67, p=0.50>0.05, figure9). A funnel diagram(figure10) was made to test the three articles' publication bias. Furthermore, the figure10 bias test indicates that there is no publication bias in the research literature (P= 0.667 >0.05).
3. Histomorphometric Analysis
3.1 Meta-analysis of the fraction of new bone(%NB)
We can choose the random effect for the meta-analysis since I2= 84% > 50%, and P= 0.01<0.1, demonstrating that there is high heterogeneity among the selected studies25, 28. Comparing with the control group, the experimental group’s fraction of new bone was 22.48% higher, according to the combined effect of Meta, which was 22.48 (9.24~35.71). However, neither group was significantly different from the other (z=3.33, p=0.0009<0.05, figure11). A funnel diagram(figure12) was made to test the two articles' publication bias. Furthermore, figure13 showed that there was no data increase or decrease after the cut and fill method, indicating no publication bias.
3.2 Meta-analysis of the fraction of residual graft (%RG)
We can choose the random effect for the meta-analysis since I2 = 92% >50%, and P= 0.0003<0.1, demonstrating that there is high heterogeneity among the selected studies25, 28. Comparing with the control group, the experimental group’s fraction of residual graft was 5.62% lower, according to the combined effect of Meta, which was -5.62 (-28.14~16.89). However, neither group was significantly different from the other (z=0.49, p=0.62>0.05, figure14). A funnel diagram(figure15) was made to test the two articles' publication bias. Furthermore, figure16 showed that there was no data increase or decrease after the cut and fill method, indicating no publication bias.
3.3 Meta-analysis of the fraction of soft tissue matrix (%STM)
We can choose the random effect for the meta-analysis since I2 = 50%, and the P= 0.16>0.1, demonstrating that there is low heterogeneity among the selected studies25, 28. Comparing with the control group, the experimental group’s fraction of soft tissue matrix was 11.90% lower, according to the combined effect of Meta, which was -11.90 (-19.29~-5.73). However, neither group was significantly different from the other (z=3.06, p=0.002<0.05, figure17). A funnel diagram(figure18) was made to test the two articles' publication bias. Furthermore, figure19 showed that there was no data increase or decrease after the cut and fill method, indicating no publication bias.