Meta-Analyses and Orthodontic Evidence-Based Clinical Practice in the 21^st Century

Search Strategy

For the identification of the studies to be considered for inclusion in this SR, detailed search strategies were developed and undertaken for following major electronic databases: MEDLINE (searched via PubMed), EMBASE, Cochrane Database of Systematic Reviews (Cochrane Reviews), Google Scholar Beta, ISI Web of Knowledge (all databases), Scopus, LILACS (searched via Virtual Health Library), Bibliografia Brasileira de Odontologia (searched via Virtual Health Library), Bandolier, Extenza, African Journals Online, Dissertations and Theses (searched via ProQuest), Theses (University of Glaskow), and German National Library of Medicine (ZB MED). These were based upon the search strategy developed for PubMed but revised appropriately for each database to take into account differences in controlled vocabulary and syntax rules. All electronic searches were conducted on July 22, 2009.

The databases searched and the corresponding search strategies per database are presented in detail in Table 2.

Following the electronic searching as described above, the reference lists of the initially retrieved articles were handsearched to identify possible additional relevant publications that may have been missed in the database searches.

Databases were searched to include all languages and no language restriction was applied during the whole identification process of the published articles.

Inclusion Criteria

The criteria for considering studies to be included in this SR were the following: (a) the studies should have been indexed as MAs by the corresponding databases, (b) they should have analyzed primary data of the individual articles by means of meta-analytical statistical procedures, and (c) they should have investigated issues related to orthodontics.

Exclusion Criteria

Studies investigating subjects not relevant to orthodontics, as well as articles discussing various aspects of cleft lip and palate, were excluded from this SR.

In addition, SRs, all types of original studies, as well as other types of publications, such as narrative reviews, comments on other articles, short surveys, case reports, protocols, book reviews, letters to the editors, etc., were not included in this evaluation, even if they discussed orthodontic related issues.

Nevertheless, since some confusion still exists in the terminology of SR and MA, and several SRs are indexed in the corresponding databases (or even considered by the writing authors) as MAs and vice versa, the full text of each study initially identified as SRs was examined before its exclusion from current SR, in order to confirm that no MAPs have been used for primary data evaluation.

Selection of Articles

The eligibility of the selected articles was verified firstly by reading the titles and abstracts of the articles initially identified in order to exclude articles discussing subjects not related to orthodontics. Then, comprehensive reading of the full-text of the remaining articles was undertaken by the author of current SR in order to include or exclude articles according to the inclusion and exclusion criteria as described above.

Quality Assessment

Quality assessment is usually applied for the evaluation of primary studies, such as RCTs. However, the quality of studies with other study designs cannot be evaluated in the same way as for randomized controlled trials (RCTs) [28-30]. Various criteria have been suggested to critically appraise their validity, which can be applied to other types of studies; however, a great deal of judgment is necessary [31]. Thus, a quality assessment of the MAs included in this SR was undertaken using the following quality criteria: (a) number of the original studies included in the analysis, (b) electronic databases searched, (c) undertaking of hand-searching, (d) cross-checking of references, (e) inclusion of non-English studies, (f) quality analysis of the original studies, (g) heterogeneity analysis of the primary data, (h) publication bias assessment, and (i) statistical procedures used for primary data analysis.

Class II Treatment

Nguyen et al. [32] investigated the risk of traumatic dental injuries of the anterior teeth due to overjet. The authors developed a methodological checklist for observational studies, in order to qualitatively assess the 11 articles that were finally included in the analysis. The relative risk of overjet compared with a reference was expressed as odds ratios (ORs). For each primary study, the ORs and their 95% confidence intervals (CIs) were computed and subsequently these ORs were pooled across the studies. Finally, the influence of the quality of the studies on the pooled OR was addressed. According to this evaluation, the authors concluded that: (a) children with an overjet larger than 3 mm are approximately twice as much at risk of traumatic dental injury on anterior teeth than children with an overjet less than 3 mm, (b) the effect of overjet on the risk of dental injury is less in boys than in girls of the same overjet group, and (c) the risk of anterior teeth injuries tends to increase with increasing overjet size. Regarding the MAPs used in this study, and despite the fact that a possibility of publication-bias cannot be excluded, some additional shortcomings include the following: (a) The studies selected for the metaanalysis showed heterogeneity of the outcomes, with exception of the subgroup between the ages 6 to 18. (b) There was no reference to the number of the control group size for any of the primary studies. (c) No information was given if studies in other language than English were included in the analysis. To conclude, although this MA was performed adequately well, the above mentioned conclusions should be considered with some caution due to the limitations of the source studies and the heterogeneity of the primary data implemented in the analysis.

The effect of functional appliances on the skeletal pattern following orthodontic treatment of patients with Class II, division I malocclusion was the subject of a MA performed by Mills [33]. The author reviewed the cephalometric findings of 26 papers investigating the Andresen-activator and Frankel appliances. In order to increase statistical significance, the findings of the primary studies were combined to produce larger samples, and they were also compared with a control group derived from reports of untreated Class II, division 1 individuals. For this purpose, the author combined the means and the given standard deviations of the primary studies using the student’s t-test for the full figures and the Mann-Whitney U-test for the annual measurements and concluded that: (a) there was no appreciable restraining effect on the forward growth of the maxilla in either of the groups (functional appliances group and control group); (b) a slight mean increase in mandibular growth could be observed, mainly in a vertical direction; (c) no change in the position of the glenoid fossa was evident; and (d) there was a wide individual response and average changes were rarely observed in a patient. Although there was no language bias, no multiple publication bias and no citation bias, the MAPs applied in this study presented the following problems: (a) No search strategy was developed and no searching of the electronic databases was performed. The literature search was incomplete including only papers from American, European and Scandinavian literature, and thus, additional studies may exist that were not included in the MA. (b) No special designed tests for MAs were used; instead, statistical tests designed for primary data analysis (student’s t-test and Mann-Whitney U-test) were used, even though some primary studies did not report standard deviations! (c) There was lack of quality and heterogeneity analysis as well as of assessment of publication bias. (d) The treatment duration of the patients included in the primary studies was different, while the control group was consisted of cases with rather milder symptoms, facts that are associated with selection bias. Consequently, this MA can not be considered as appropriately conducted. Thus, the corresponding conclusions are not supported with strong evidence and they should not be taken onto consideration.

Chen et al. [34] investigated the efficacy of functional appliances on mandibular growth in patients with Class II malocclusion, aiming to test the hypothesis that functional appliances enhance mandibular growth. The data of 6 articles meeting validity standards were evaluated for 12 cephalometric variables using ANOVA, students t-test for paired data, and 95% CIs. Only the variables Ar-Pg and Ar-Gn differed significantly between the treated and control groups. Weaknesses of the methodology used in this MA include the following: (a) The lack of searching databases additionally to MEDLINE. (b) Only English language papers were evaluated (language bias); (c) The use of statistical tests designed for primary data analysis (student’s t-test, ANOVA). (d) No control and treated group sizes were mentioned in the text (although the results of the analysis were given in box plots). (e) The age of patients at the start of the treatment as well as the treatment duration was different in the primary studies (selection bias). Consequently, no definite conclusions concerning the efficacy of functional appliances on mandibular growth could be reached. This could be attributed to the inconsistencies in the cephalometric variables of the primary studies measuring treatment outcome, as well as to the different treatment durations of the various patient samples.

More recently, Antonarakis and Kiliaridis [35] evaluated the anteroposterior short-term skeletal and dental effects in growing patients with Class II malocclusion following treatment with functional appliances, extraoral traction, or combination appliances. They assessed the primary data of 9 studies which investigated changes referring to maxilla (as expressed by the SNA angle), mandible (as expressed by the SNB angle), intermaxillary relationships (as expressed by the ANB angle), and to the amount of overjet after annualizing the corresponding results due to the different treatment duration observed in the various groups. For the statistical analysis they used the random-effects model to calculate the weighted mean differences (WMDs), while forest plots were drawn to graphically present their results. Heterogeneity analysis was performed by means of chi-square tests and I² statistic. The authors concluded that: (a) Functional and extraoral traction appliances used for the treatment of Class II malocclusion in growing patients are associated with an improvement of the sagittal intermaxillary relationship. (b) Activators, twin block and combination appliances are acting mainly on the mandible, while twin block appliances also have a significant effect on maxillary growth. (c) Extraoral appliances are acting mainly on the maxilla. (c) All appliances under investigation except extraoral traction revealed a large decrease of the overjet. Although the literature search was performed adequately well (however, no information is given if non-English papers were also included) and a quality analysis of the included studies was undertaken, the methodology followed by the authors present some drawbacks. In detail, one of the studies involved patients with Juvenile Rheumatoid Arthritis (JRA), and in two studies there was no blind allocation for patients selected for the treated and the control groups. Further, the authors excluded papers investigating Herbst appliance treatment, although Herbst appliances are in general considered as functional (despite the fact that they are fixed) [59]. The data of these studies might have been used for sub-group analysis and comparison with the other treatment approaches under investigation. It is not rational to include in a MA a group of patients suffering from JRA, presenting specific craniofacial characteristics [60,61] and combine their data with data derived from conventional orthodontic patients and at the same time to exclude a treatment modality (Herbst appliance) that belongs to the subjects of the specific investigation. Finally, no assessment of publication bias was performed. To conclude, the results of this investigation should be treated with some caution.

At the same time period, Harrison and colleagues [36] assessed the effectiveness of orthodontic treatment of 7 to 9 years old children and adolescents presenting prominent upper front teeth. Following implementation of specific inclusion and exclusion criteria, 8 trials involving data from 592 individuals were included in the analysis. Initially, a quality analysis of the included studies was undertaken. Statistical analysis was performed as following: For dichotomous outcomes, the estimates of effect of an intervention were expressed as ORs together with their 95% CIs. For continuous outcomes, the mean differences and standard deviations were used to summarize the data for each group. Further, using the random-effects models, the ORs were combined for dichotomous data, and the mean differences for continuous data. Heterogeneity analysis was also performed by means of Cochran’s test for heterogeneity and the I² statistic. The authors concluded that: (a) early orthodontic treatment, followed by a later phase of treatment when the patient is in early adolescence, does not appear to have any advantages over one-phase treatment in early adolescence, (b) when treatment with functional appliances is provided, there a minor skeletal changes, probably not clinically significant, and (c) the use of functional appliances does not seem more effective than twin blocks. The methodology followed by the authors present no shortcomings and therefore the corresponding conclusions are supported with strong evidence. Thus, they can be seriously taken into consideration when planning orthodontic treatment of children or adolescents presenting prominent upper front teeth.

Class III Treatment

Kim et al. [37] evaluated the effectiveness of maxillary protraction with orthopedic appliances in patients with Class III malocclusion, aiming additionally to determine a consensus for controversial issues such as treatment timing and use of adjunctive intraoral appliances. Fourteen studies met the selection criteria and included in the analysis. In order to combine the primary data, the means and standard deviations of the original articles were summarized and the corresponding results were graphically represented. The statistical analysis of the changes following treatment, as measured by selected cephalometric landmarks, showed no distinct differences between palatal expansion and non-expansion group, except for one variable. Further, examination of the effect of age revealed greater treatment changes in the younger group of patients. These results indicated that (a) protraction face mask therapy is effective in patients who are growing, but to a lesser degree in patients who are older than 10 years of age, and (b) protraction in combination with an initial period of expansion may provide more significant skeletal effects. Although no language bias was evident, this MA was associated with the following drawbacks that weaken the provided evidence: (a) The authors used statistical tests for primary data analysis and not specially designed tests for MAs. (b) The literature search was limited only to a MEDLINE search and no other databases were examined. (c) An ethnic maturation differential may exist in the primary studies. (d) There was lack of a matched control group. (e) There was lack of standardization of the design of the primary studies. (f) No heterogeneity analysis and no quality assessment of the included studies were undertaken. Consequently, the results of this MA cannot be considered that are supported with strong evidence.

Maxillary protraction treatment was also the subject of a MA performed by Jager et al. [38] aiming to quantitatively review the published results concerning the treatment effects of maxillary protraction on craniofacial growth of patients presenting Class III malocclusion. After applying specific inclusion criteria, 12 studies remained for further evaluation. In order to combine the data of the primary studies, the results of various cephalometric measurements were reviewed using the Dstat 1.10 software by calculating a standardized treatment effect variable. The homogeneity of the variances of the different effect variables was evaluated and a composite effect was calculated. According to the results of this investigation, a significant composite effect of the maxillary protraction treatment on some craniofacial skeletal and dental components was evident. However, several of the individual effect variables demonstrated a significant lack of homogeneity. Study characteristics which might possibly account for this variability were the age of the patients at treatment start and the combination of maxillary protraction with rapid maxillary expansion. With regard to the statistical procedures used, the fact that a specially designed test for MA was performed adds to the strengths of this study, while the limitation of literature searching in only one database (MEDLINE) without any information regarding the non-English studies, as well as the lack of assessment of publication bias weaken the power of the analysis.

Treatment of Transverse Problems

The subject of the MA performed by Burke and colleagues [39] was the evaluation of the changes of mandibular intercanine width during and after treatment and post-retention. Twenty six studies that assessed the longitudinal stability of post-retention mandibular intercanine width were evaluated. For the statistical analysis, weighted averages and standard deviations for the means were compared for linear changes in intercanine transverse dimensions during treatment (T1), immediately after treatment (T2), and after removal of all retention (T3). Paired two-tail t-tests were performed between T3 and T1 means on all groups and 95% CIs were computed. The authors concluded that: (a) Regardless of patient diagnostic and treatment modalities or whether treatment was extraction or non-extraction, mandibular intercanine width tends to increase during treatment about 1-2 mm, to decrease post-retention to approximately the original dimension, and to show a net change after post-retention between +0.5 to -0.6 mm. (b) While statistically significant differences could be demonstrated within various groups, the magnitudes of these differences were not considered clinically important. (c) The net change in mandibular intercanine width that was approximately zero, which supports the concept of maintenance of the initial intercanine width in orthodontic treatment. Regarding the methodology used, the following issues should be mentioned: (a) The authors used statistical tests for primary data analysis and not specially designed tests for MAs. (b) No control group size was mentioned. (c) The literature search was limited only to a MEDLINE search and no other databases were investigated. (d) The post-retention period varied from 4 months to 12 years. (e) Quality and heterogeneity analysis as well as assessment of publication bias of the included studies were not performed. Consequently, the results of this MA are not well supported by the methodology used and they should be considered with caution.

Schiffman and Tuncay [40] evaluated the existing literature on maxillary expansion in 2001 in order to understand the appropriateness and stability of this treatment approach. The assessment of the primary data was performed by coding and scoring each study with respect to pre-established characteristics. Following this evaluation, only 6 studies remained for the final analysis. An overall effect size was computed and aspects of study design were analyzed. The results of this investigation were the following: (a) The mean expansion after appropriate adjustment was 6 mm (±1.3 mm). Of the 6 mm average, 4.9 mm was retained while wearing retainers; (b) The 6 mm average expansion with retention yielded a 4.7 mm residual expansion, which subsequently was reduced to 3.9 mm during the short-term post-retention period. (c) In the long-term post-retention period, only 2.4 mm of the residual expansion remained, which was no greater than what is documented as normal growth. The authors concluded that there was inadequate evidence to support the opinion that the expansion achieved beyond of what is expected due to normal development of the maxilla could be retained in the long term. The methodology used in this study, includes the following shortcomings: (a) The literature search was performed in only one electronic database (MEDLINE), no other databases were investigated, and this search was conducted only for the period from 1979 to 1999. (b) There was a language restriction, since only English language publications were included in the MA. (c) The authors did not present any P-values for the measurements given in their analysis. (d) The lack of uniformity of the included primary studies and their improper study design, which did not permit an appropriate statistical analysis. (e) The lack of a detailed analysis of heterogeneity. To conclude, due to these weaknesses the revealed evidence should be regarded with some caution.

Harrison and Ashby [41] performed a MA aiming to evaluate orthodontic treatment procedures used to expand the maxillary dentition and correct posterior crossbites. For the statistical analysis of the 13 primary studies, the ORs, relative risk, relative risk reduction, absolute risk reduction, and the corresponding 95% CIs, were calculated for event data while the WMDs and 95% CIs were calculated for continuous data. The authors also conducted a quality analysis of the included studies. The conclusions of this MA were: (a) Occlusal grinding in the primary dentition, with or without the addition of an upper removable expansion plate, is effective in preventing a posterior crossbite from being perpetuated to the mixed and permanent dentition. This is however only based on data derived from two studies with small sample sizes. (b) No evidence of a difference in treatment effect, as documented by molar or canine expansion, between the test and control intervention was found in the trials which compared banded vs. bonded and two point vs. four point rapid maxillary expansion, banded vs. bonded slow maxillary expansion, transpalatal arch with or without buccal root torque, or upper removable expansion appliance vs. quad-helix. (c) Not adequate data were provided regarding two point vs. four point rapid maxillary expansion to permit an appropriate analysis. Although the procedures used in this study were according to the guidelines for performing MAs, [14] almost all primary trials included in the analysis presented small sample sizes, a fact that weakens in a way the evidence produced by this MA. Therefore, as the authors also concluded, further studies with appropriate sample sizes are needed to assess the effectiveness of the above interventions.

More recently, Lagravère and colleagues [42] performed a MA of the immediate transverse, anteroposterior, and vertical dental and skeletal changes following orthodontic treatment with rapid maxillary expansion. Fourteen articles satisfying the final inclusion criteria, though with methodological deficiencies, were included in the analysis by calculating of effect size estimates and their corresponding 95% CIs using the fixed-effects model for homogeneous data and the random-effects model for heterogeneous data. The authors concluded that (a) the greatest changes immediately after rapid maxillary expansion were observed in the maxillary transverse plane. (b) Maxillary width increase was due to dental expansion rather than true skeletal expansion. (c) Only few vertical and anteroposterior changes were statistically significant, and none of them was clinically significant. The drawbacks of this MA include the following: (a) The authors made a great effort by searching a large number of electronic databases and checking also the reference lists of the selected articles in order to identify all published studies dealing with rapid maxillary expansion, excluding however non-English published studies. (b) Although a quality analysis of the included papers was undertaken, no heterogeneity analysis and assessment of publication bias was performed. (c) All of the original studies presented methodological deficiencies. Consequently, the results of this investigation should be considered with some caution.

Incisor Intrusion

The quantification of the amount of true incisor intrusion attained during orthodontic treatment was the subject of another MA performed by Ng et al. [43] Two studies were included in the analysis. The fixed-effects model was used to evaluate the amount of incisor intrusion attained by means of the segmented arch technique (SAT). The results indicated that true incisor intrusion was feasible in both arches using the SAT. Using the SAT a 1.5 mm of maxillary incisor intrusion and 1.9 mm of mandibular incisor intrusion could be achieved when applied in non-growing patients. However, the clinical significance of the amount of true intrusion as an exclusive treatment option was questionable for patients presenting severe deep bite. Regarding the methodology used, the very small number (n=2) of the source studies evaluated in this MA which additionally included only adult patients, the small sample sizes, as well as the lack of heterogeneity analysis and assessment of publication bias of the included studies weaken the power of this investigation. Therefore, the extrapolation of the observed values to younger populations appears to be questionable and the overall results of this MA should be viewed with caution.

External Apical Root Resorption

The treatment-related factors of external apical root resorption were also among the subjects recently investigated by Segal et al. [44] Clinical trials in English language with a sample size more than 10 individuals that received orthodontic treatment with fixed appliances, with available pre- and post-operative radiographs, and measurements of external apical root resorption mainly of the maxillary incisors were included in this study. Studies including samples with history of trauma, prior root resorption or endodontic treatment were excluded. In order to access the methodological soundness for each individual study, coding variables were determined along with a grading system and a cumulative "meta-analysis factor" was computed for each study. In addition, appropriateness of these selections and absence of publication bias was tested by a "funnel plot" analysis. Eight articles met the inclusion criteria and further evaluated in this MA. After calculating the weighted means of correlations between the variables, the authors found that apical tooth resorption was highly correlated with total apical displacement (r=0.822) and treatment duration (r=0.852) and concluded that treatment-related causes of external apical root resorption seem to be the total distance the apex had moved and the time it took. Consequently, factors that are associated with the duration of active treatment might result in increased levels of apical root resorption in a pre-disposed individual. Regarding the power of the statistical procedure, the limitation of searching in only one database, the restriction of articles published only in English language (suggesting database or language bias, respectively), as well as the lack of a heterogeneity analysis weaken the overall results of this MA.

Dental Anomalies

The prevalence of dental agenesis of the permanent teeth was the subject of a MA performed by Polder et al. [45]. Multiple regression-analysis (weighted least squares) was performed using date from 28 primary studies in order to evaluate the possible influence of chronological age, sample size, continent and year of publication. The prevalence of agenesis per tooth type, affected patients and number of missing teeth per patient was calculated, as far as it was reported in the corresponding papers. The relative risk was also calculated for the comparison of the prevalence between males and females. Following this evaluation the authors concluded the following: (a) The prevalence of dental agenesis differs in relation to continent and gender: for both sexes was higher in Europe (males 4.6%; females 6.3%) and Australia (males 5.5%; females 7.6%) than in North American Caucasians (males 3.2%; females 4.6%), while in females it was 1.37 times higher than in males for all three continents examined. (b) The mandibular second premolars were affected most frequently, followed by the maxillary lateral incisors and maxillary second premolars. (c) Bilateral agenesis of the maxillary lateral incisors was more frequent than the unilateral agenesis, while the opposite was found for the mandibular and maxillary second premolars, and the maxillary first premolars. Regarding the MAPs used in this study, although the consideration of possible bias (by evaluating the influence of chronological age, sample size, continent and year of publication) add to the value of this investigation, the lack of quality and heterogeneity analysis of the original studies included in the analysis weaken in a way the power of the analysis. Thus, the corresponding results should be regarded with some caution.

Most recently, the author of this paper along with colleagues [46] assessed the characteristic features and dental anomalies that may accompany tooth transposition. The literature search included several electronic databases, while cross-checking of the reference lists of the included articles was also performed. After applying specific exclusion criteria, 5 studies were found eligible to be included in the analysis. Meta-analysis was performed by determining the marginal mean proportions and the 95% CIs estimated by the random-effects model. Quality analysis of the included studies was undertaken, as well as evaluation of heterogeneity, which it was performed using the Q and the I² statistic. After noting the lack of high-validity studies investigating the characteristic features and dental anomalies of tooth transposition the authors found that tooth transposition is significantly unrelated to dental anomalies, such as congenitally missing teeth, peg-shaped or hypoplastic teeth, and impacted teeth, while there might be an association with over-retained deciduous teeth. They therefore concluded that tooth transposition can be regarded as an isolated phenomenon rather than a syndrome, while occasionally it can be accompanied by some characteristic features and dental anomalies, but without distinct associations. The methodology followed by the authors present no shortcomings, since several electronic databases were searched, cross-checking of the reference lists was performed, no language restriction was applied during the selection process, quality analysis of the included studies and assessment of heterogeneity of data was performed, and appropriate statistical procedures were applied. Thus, it can be considered that the results of this investigation are supported with strong evidence.

A further MAs performed again by the author of this paper and colleagues aiming to synthesize currently existing data and to investigate the prevalence of tooth transposition as well as its relation to gender, dental arch, and quadrant occurrence [47]. Several electronic databases were searched, while cross-checking of the reference lists was also performed in order to identify the potentially relevant studies. After applying specific inclusion and exclusion criteria, 9 studies were included in this evaluation. Quality analysis of the included studies was undertaken, as well as heterogeneity analysis of the primary data by means of the Q and I² statistic. In addition, Publication bias was also assessed by funnel plot analysis. Meta-analysis was performed by determining the event rate and the 95% CIs estimated by the random-effects model and the results were presented by means of forest plots. The analysis of the data of the primary studies revealed that (a) tooth transposition has a mean prevalence of 0.33%, (b) this prevalence seems to be similar between the two genders, and (c) tooth transposition appears more frequently in the maxilla than in the mandible and more unilaterally than bilaterally. As in the previous study mentioned above, the methodology followed by the authors including searching of several electronic databases along with cross-checking of the reference lists of the included articles and the absence of language restriction, quality analysis of the included studies, assessment of heterogeneity, and appropriate meta-analytic statistical procedures, strengthen significantly the power of the analysis and they strongly support the derived evidence.

Frequency of Severe Occlusal Problems

Frazao et al. [48] investigated the prevalence and severity of occlusal problems in populations with deciduous and permanent dentition. The authors evaluated the primary data of 7 articles by calculating the weighted ORs and their corresponding 95% CIs using the fixed-effects model, after checking data heterogeneity by means of the chi-square test with Yates correction. The statistical analysis revealed that (a) the prevalence of occlusal problems was found almost twice as large in patients with permanent than in deciduous dentition (71.3% and 49.0%, respectively), and (b) among the variables under evaluation (including sex, type of school and ethnic group), only the developmental stage of the dentition was significantly associated to the severity of malocclusion. Regarding the statistical methodology used for the analysis of the primary data, although the authors used appropriate meta-analytic approaches, the absence of a quality analysis and of the assessment of publication bias weaken in a degree the evidence produced by this MA.

Nickel Hypersensitivity

Kolokitha and colleagues [49] performed a MA in order to investigate the effect of orthodontic therapy on the prevalence of nickel hypersensitivity and compare it with the prevalence in the general population. Several electronic databases were searched and cross-checking of the reference lists of the articles was also performed. After applying specific inclusion criteria, 8 studies were eligible for the meta-analysis. For the studies reporting prevalence of Ni hypersensitivity before and after orthodontic treatment, the OR of a positive Ni patch test prior to orthodontic treatment to a positive Ni patch test after the placement of orthodontic appliances and the corresponding 95% CIs were calculated, while for the studies reporting the prevalence of Ni hypersensitivity only after placement or removal of orthodontic appliances, and in the general population, the relevant event rates were computed. Further, the random-effects model was used to combine the ORs before and after orthodontic treatment. In addition, evaluation of the validity of the included articles and funnel plot analysis for the evaluation of publication bias were performed, while the Cochran test for homogeneity and the I² test were also calculated to check for heterogeneity and inconsistency, respectively. The results of this analysis indicated that: (a) There is lack of high-validity longitudinal studies of the prevalence of nickel hypersensitivity in patients before and after orthodontic treatment and in appropriate controls. (b) No statistically significant difference between the odds for a positive patch-test result before orthodontic treatment and after the placement of the appliances was observed. (c) Orthodontic patients with no cutaneous piercing or with skin pierced have no statistically significant differences of nickel hypersensitivity after treatment compared with the general population. Consequently the authors concluded that orthodontic treatment is not associated with an increase in the prevalence of nickel hypersensitivity unless subjects have a history of cutaneous piercing. The methodology followed in this MA resulted in strong evidence, since several electronic databases were searched, cross-checking of the reference lists was performed, no language restriction was applied during the selection process, quality analysis of the included studies and assessment of publication bias and of heterogeneity was performed, as well as appropriate statistical procedures were applied.

Orthodontics and Temporomandibular Disorders

The subject of the MA conducted by Kim et al. [50] was the investigation of the relationships between orthodontic treatment and temporomandibular disorders (TMDs) in orthodontic patients following completion of their treatment. Data was extracted from 31 articles and was divided according to the study designs, and TMD symptoms, signs, or indexes. To test whether all primary studies attempted to estimate or observed the same true effect, and whether variability between results of the studies was due to random error only, a statistical test for the hypothesis of parametric homogeneity was conducted. In addition, probabilities of homogeneity and odds on parametric homogeneity were calculated. Due to the extreme heterogeneity of the data of the primary studies, it was not possible to complete the statistical analysis. Thus, a definitive statement concerning the relationships between orthodontic treatment and TMDs could not be concluded. However, although it was not possible to perform a "true" (in other words, a complete) MA, the authors followed, at least during the initial stages an appropriate meta-analytic approach. Despite the extreme heterogeneity of the data of the primary studies, the following issues weaken even more the results of this investigation: (a) Only one electronic database was investigated. (b) Only English language papers were taken into consideration. (c) Not all the primary studies included were of the same quality. (d) Inclusion of studies regardless the age of the patients. (e) Large number of different types of appliances. (f) Different treatment durations of the patient samples. (f) Lack of assessment of the quality and publication bias of the original studies included in the analysis. To summarize, this MA do not provide any evidence concerning the relationships between TMDs and orthodontic treatment.

Macedo et al. [51] performed a MA to evaluate the effectiveness of occlusal splints for the treatment of sleep bruxism with alternative interventions, placebo or no treatment. Among the 32 potentially relevant clinical trials initially identified, only 5 RCTs satisfied the inclusion criteria and were selected for further analysis. In these trials occlusal splint treatment was compared to treatment with palatal splints, mandibular advancement devices, transcutaneous electric nerve stimulation, and no treatment at all. However, the methodological quality of the five trials included was low. In detail, none of the studies gave a clear description of the concealment of allocation and only one study described how the allocation sequence was generated, some inclusion criteria were not specified by the authors, and other outcomes could not be analyzed because they were inappropriately described. In addition, just one outcome (arousal index) was common and it was combined in a meta-analysis which was performed using the random-effects model. For dichotomous outcomes, the estimate of effect of an intervention was expressed as risk ratios together with the corresponding 95% CIs. For continuous outcomes, WMDs to compare groups and standard deviations were used to summarize the data for each group. In addition, quality analysis of the included studies was performed, and heterogeneity analysis was undertaken using the Cochran’s test for heterogeneity and the I² statistic. The results of the MA indicated that no statistically significant differences between the occlusal splint treatment and the control groups were found. Therefore, there is not sufficient evidence to support the effectives of the use of occlusal splints for the treatment of sleep bruxism with regard to sleep outcomes, but their use might be beneficial with regard to tooth wear.

Obstructive Sleep Apnea Syndrome

The subject of the MA conducted by Miles et al. [52] was the investigation of possible significant differences between the cephalometric variables describing the craniofacial skeletal or soft tissue morphology of individuals with and without obstructive sleep apnea syndrome (OSAS). A hierarchical analysis was conducted to examine the quality of evidence within the studies included in the analysis. The MAPs employed in this study included combined means and standard deviations for the OSAS and control groups for effect size plots in order to examine the distribution and consistency of outcomes across studies, Z-scores for statistical significance testing between groups, and receiver operating characteristic (ROC) curves. Following this evaluation the authors concluded that: (a) The literature is characterized by several methodological deficiencies and therefore is equivocal regarding a causal association between craniofacial structure and OSAS. (b) Evidence for a direct causal relationship between craniofacial structure and OSAS is unsupported by the literature, both qualitatively and quantitatively. (c) The rationale for OSAS treatments based on morphologic criteria remain unsubstantiated. (d) The two most consistent strong effect sizes with the highest potential diagnostic accuracies presented the variables related to mandibular structure (Sn/MPA, Go-Gn). (e) Although mandibular body length (Go-Gn) appears to be an associated factor, this does not support causality. (f) More standardization of research methods and data presentation is required. Regarding the methodology used, only one sample study and none of the efficacy treatment studies met all the quality criteria for inclusion in the analysis, as set by the authors. However, MA was performed in the examined variables by including data from studies with control samples and for each variable, data from different number of studies was included in the MA (i.e., for Go-Gn, measurements from 3 studies were combined). Further, although no publication biases were present, the studies included in the analysis were limited to those published in English language and no heterogeneity analysis of the primary data was undertaken. Consequently, it can be stated that the MAPs followed by the authors did not result in strong evidence.

Lim et al. [53] performed a MA in order to determine the clinical effectiveness of oral appliances in the treatment of obstructive sleep apnoea-hypopnoea (OSAH) syndrome in adults. After applying specific inclusion criteria, 17 studies (involving 831 participants) found to be eligible and they were included in the analysis. The quality assessment revealed that all the original studies included in the analysis presented some shortcomings, such as small sample size, underreporting of methods and data, and lack of blinding. For continuous variables, the authors calculated the results of the individual studies using a fixed-effect WMD or a standardized mean difference (SMD) with the corresponding 95% CIs. For dichotomous variables, a fixed-effect OR with the corresponding 95% CIs were calculated. Heterogeneity was assessed using the Breslow-Day's test, the Chi-square test and the I² statistic. When heterogeneity was present, the random-effects model was used to calculate the results. According to the results of this MA the authors concluded that: (a) There is increasing evidence suggesting that oral appliances improve subjective sleepiness and sleep disordered breathing compared with a control. (b) Nasal continuous positive airway pressure (CPAP) therapy appears to be more effective in improving sleep disordered breathing than oral appliances. (c) Although more evidence is needed to further support the results of this investigation, it may be more appropriate to recommend oral appliances treatment to patients with mild symptomatic OSAH, as well as to those patients who are unwilling or unable to tolerate CPAP therapy. Although the evidence produced by this investigation with regard to the appropriate meta-analytic approaches followed by the authors revealed as strong enough to support the use of CPAP therapy in patients with OSAH syndrome, its relative effects compared with oral appliances treatment require further elucidation, since the currently existing trials present a number of shortcomings.

Oral Hygiene

Kaklamanos and Kalfas [54] performed a MA in order to meta-analysis was to compare the effectiveness of powered and manual toothbrushes in reducing gingival inflammation in patients treated with fixed orthodontic appliances. After applying specific inclusion criteria concerning study design, participants’ characteristics, intervention characteristics, and principal outcome measures, 5 trials were considered appropriate to be included in the meta-analysis. The WMD with the corresponding 95% CI was used to express the comparative treatment effect, while the random-effects model was used to combine treatment effects across the studies in each category. The quality of the included trials was evaluated by assessing randomization, allocation concealment, blinding, and handling losses using the Cochran test for homogeneity and the I² statistic. Based on quality assessment and the short experimental period of the 5 trials included in the analysis, and following appropriate meta-analytic procedures the authors concluded that current evidence is insufficient to support the comparative efficacy of powered toothbrushes in reducing gingivitis in patients undergoing fixed orthodontic appliance treatment.

Dentition Analysis

The Moyer’s method of mixed dentition analysis was the subject of a MA performed by Buwembo and Luboga in order to assess its applicability in different ethnic groups [55]. The authors performed pooling of the correlation coefficients of the 7 primary studies included in the MA and calculated the weighted mean correlations, its variances, the variances in the population correlation, and the population correlation coefficients (by means of a chi-square test). Overall, correlation coefficients were found to be in the borderline of variation with a P-value of 0.05. Splitting the articles into Caucasian and Asian groups also gave borderline P-values of 0.05. Consequently, the authors concluded that Moyer’s method of mixed dentition analysis may present population variations and they therefore proposed prediction tables to be developed for specific populations. Regarding the methodology of this study, the search strategy including only one electronic database, the small amount of studies for the various ethnic groups and their different characteristics, the lack of a quality and heterogeneity analysis as well as of publication bias assessment of the original studies included in the analysis weaken the overall results of this investigation. In addition, no statistical outcomes were presented concerning the results of the MAPs conducted for the comparison of the Caucasian and Asian ethnic groups, although such results are discussed in the text. For all these reasons, the results of this MA should be regarded with caution.

The aim of the MA performed by Liversidge et al. [56] was to investigate the timing of individual tooth formation stages according to the Demirjian’s method in children from eight countries (i.e., Australia, Belgium, Canada, England, Finland, France, South Korea and Sweden). The authors included data from studies known to them, without trying to identify potential relevant articles by searching the major electronic databases or through hand searching of dental and orthodontic journals. Nevertheless, data of permanent mandibular developing teeth from panoramic radiographs were combined and the age-of-attainment was calculated using logistic regression for tooth stages for each group, boys and girls separately. The overall mean age-of-attainment was calculated using a weighting factor inversely proportional to the variance of the mean. The results of this MA indicated that: (a) The mean ages for each group and total sample were significantly different in 65 out of 509 comparisons. Some of these were of small sample size but there was no consistent pattern. (b) Apex closure of the first molar took place significantly later in children from Canada and this might explain differences found in the dental maturity score. Consequently the authors concluded that their results suggest no major differences in the timing of tooth formation stages using the Demirjian’s dental maturity method between the children of the above mentioned 8 countries. However, these results and conclusions should not be taken into consideration, since the MAPs followed by the authors presented significant drawbacks including: (a) lack of electronic and hand searching, (b) lack of inclusion criteria, (c) lack of quality analysis of the included papers, (d) lack of assessment of publication bias, and (e) lack of heterogeneity analysis.

Cephalometric Analysis

In this MA, Trpkova et al. [57] tried to assess the magnitude of lateral cephalometric landmarks identification error for 15 landmarks. The statistical analysis of the data of the 6 primary studies included in this investigation included weighted average of the estimates in order to combine studies reporting means and standard error, as well as one-way analysis of covariance in order to combine studies reporting standard deviations. The results of this study included a measure of systematic and random errors involved when locating landmarks on lateral cephalometric radiographs and they were presented as standard mean errors and 95% CIs for the repeatability and reproducibility of the 15 cephalometric landmarks investigated. The authors concluded that a 0.59 mm of total error for the x-coordinate and 0.56 mm for the y-coordinate are acceptable levels of accuracy and only the cephalometric landmarks B, A, Ptm, S, and Go on the x-coordinate, and Ptm, A and S on the y-coordinate presented with insignificant mean error and small value for total error. Therefore, these landmarks may be considered to be reliable for cephalometric analysis of lateral radiographs. Regarding the methodology used in this MA, the following limitations were present: (a) The literature search was included only one electronic database (MEDLINE) and no other databases were investigated. (b) No information was given if a language restriction was applied during the identification process of the papers to be included in the analysis. (c) No quality and heterogeneity analysis as well as assessment of publication bias of the primary data was undertaken. (d) No information was given regarding the age of the patients of the primary studies. Thus, it can be concluded than the methodology used by the authors rather weakens the evidence produced by this MA and consequently the above mentioned results should be treated with caution.

Orthodontic Education

The purpose of the study of Al-Jewair et al. [58] was to compare the efficacy of computer-assisted learning (CAL) with traditional methods of learning in orthodontic education. Participants considered were undergraduate or postgraduate orthodontic students or orthodontic educators. Comprehensive electronic and manual searches were conducted, and after applying specific criteria 9 studies (7 randomized controlled trials and 2 prospective studies) assessing CAL in teaching orthodontic diagnosis and treatment planning were included in the analysis. Initially, the sample size, mean percentage, and standard deviation percentage were obtained from the original studies. The statistical effect size was calculated using WMDs and the corresponding 95% CIs. Heterogeneity analysis was performed using the Cochran test for homogeneity and the I² statistic. Due to the presence of heterogeneity, the random-effects model was used for the MA. The authors couldn't draw any definitive conclusions about the time efficiency of CAL. However, the evidence derived from current well conducted MA, indicates that when CAL is applied for teaching orthodontic topics related to diagnosis and treatment planning, it is at least as effective as conventional modes of orthodontic learning, and adds an additional small but significant gain in student knowledge acquisition. They therefore suggested that further studies are needed to examine other important outcomes, such as the efficacy of CAL on performance in clinical procedures or in different orthodontic subject areas, cost-effectiveness, knowledge retention, the burden it imposes on students, and its effect on educators in terms of attitudes and effort involved.

To summarize, the evaluation of the MAPs followed in the various studies discussed in this SR, revealed partially some deficiencies that weaken the evidence produced by these studies, and consequently a more critical appraisal of their results is recommended. The main deficiencies observed include: (a) Biases during the procedure of identifying and selecting the primary studies. (b) Language or publication biases are either present or the possible measures taken (or not) to avoid these bias are not reported by the authors. (c) Lack of evaluation of the homogeneity of the primary data, and if performed, the existence of heterogeneity. (d) Lack of information regarding sample and control group sizes, which eliminates the possibility for other researchers to repeat the MA using the same information and data provided or to re-evaluate the scientific progress by applying a cumulative MA. (e) Limitations in the treatment groups, such as the existing differences concerning the age of the patients, the analysis of treatment groups with developmental differences and the comparison of different types of interventions. (f) The small number of the original articles that was possible to be included in almost all the MAs, due to the lack of high quality research articles in orthodontic literature.

Taking into consideration only the well conducted MAs that used appropriate search strategies as well as meta-analytical statistical procedures to synthesize primary data without methodological deficiencies, and at the same time included in the corresponding analysis only original articles with a verified quality of an adequate level and presented no biases or significant heterogeneity of the primary data, it can be concluded that at the beginning of the 21^st century the orthodontic related issues that are supported with adequate or strong evidence include the following:

Class II treatment- (a) The risk of anterior teeth injuries tends to increase with increasing overjet size and children with an overjet larger than 3 mm are approximately twice as much at risk of traumatic dental injuries on the anterior teeth than children with an overjet less than 3 mm. In addition, the effect of the amount of overjet on the risk of dental injuries is less in boys than in girls of the same overjet group. (b) Early orthodontic treatment, followed by a later phase of treatment when the patient is in early adolescence, does not appear to have any advantages over one-phase treatment in early adolescence. (c) When treatment with functional appliances is provided, although minor and probably not clinically significant skeletal changes are expected, there is however no definite evidence concerning their efficacy on mandibular growth and in addition their use does not seem more effective than twin blocks. Thus, even if some positive effects of the functional appliances on the skeletal pattern and especially on mandibular growth can be assumed, there is still no strong evidence to support these findings.

Class III treatment - A significant composite effect of the maxillary protraction treatment on some craniofacial skeletal and dental components was evident. However, no definite conclusions can be drawn regarding the effect of the different parameters, such as the age of the patients at treatment start and the combination of maxillary protraction with rapid maxillary expansion, on treatment outcome.

Treatment of transverse problems - (a) Occlusal grinding in the primary dentition, with or without the addition of an upper removable expansion plate, is effective in preventing a posterior crossbite from being perpetuated to the mixed and permanent dentition. (b) There is no evidence of a difference in maxillary expansion, between banded vs. bonded or two point vs. four point rapid maxillary expansion, banded vs. bonded slow maxillary expansion, transpalatal arch with or without buccal root torque or upper removable expansion appliance vs. quad-helix.

External apical root resorption - The total distance and the time period of orthodontic tooth movement could be considered as treatment-related causes of external apical root resorption, and consequently, all factors that contribute to a prolonged duration of active orthodontic treatment might enhance the phenomenon of apical root resorption, especially in pre-disposed individuals.

Dental anomalies - (a) The prevalence of congenitally missing teeth differs in relation to continent and gender: for both sexes was higher in Europe and Australia than in North American Caucasians, while in females it was 1.37 times higher than in males for all three continents examined; the mandibular second premolars are affected most frequently, followed by the maxillary lateral incisors and maxillary second premolars; and bilateral occurrence of congenitally missing maxillary lateral incisors is more frequent than unilateral, while the opposite was found for the first maxillary premolars and the second mandibular and maxillary premolars. (b) Tooth transposition can be regarded as an isolated phenomenon rather than a syndrome, while occasionally it can be accompanied by some characteristic features and dental anomalies, but without distinct associations; tooth transposition has a mean prevalence of 0.33%; this prevalence seems to be similar between the two genders; tooth transposition appears more frequently in the maxilla than in the mandible and more unilaterally than bilaterally.

Frequency of severe occlusal problems - The developmental stage of the dentition is significantly associated to the severity of malocclusion, while the prevalence of occlusal problems is almost twice as large in patients with permanent than in patients with deciduous dentition.

Nickel hypersensitivity - (a) No statistically significant difference exists between the odds for a positive patch-test result before orthodontic treatment and after the placement of the appliances. (b) Nickel hypersensitivity of orthodontic patients with no history of cutaneous piercing or with skin pierced prior to orthodontic treatment does not differ statistically significant after treatment with the corresponding hypersensitivity of the general population. (c) Orthodontic treatment is not associated with an increase in the prevalence of nickel hypersensitivity unless subjects have a history of cutaneous piercing.

Obstructive sleep apnea syndrome - (a) There is increasing evidence suggesting that oral appliances improve subjective sleepiness and sleep disordered breathing. (b) Nasal continuous positive airway pressure (CPAP) therapy appears to be more effective in improving sleep disordered breathing than oral appliances. (c) Although more evidence is needed, it may be more appropriate to recommend oral appliances treatment to patients with mild symptomatic OSAH, as well as to those patients who are unwilling or unable to tolerate CPAP therapy.

Orthodontic education - When CAL is applied for teaching orthodontic topics related to diagnosis and treatment planning, it is at least as effective as conventional modes of orthodontic learning, and adds an additional small but significant gain in student knowledge acquisition.

Oral hygiene - Current evidence is insufficient to support the comparative efficacy of powered toothbrushes in reducing gingivitis in patients undergoing fixed orthodontic appliance treatment.

Orthodontics and temporomandibular disorders - There is not sufficient evidence to support the effectives of the use of occlusal splints for the treatment of sleep bruxism with regard to sleep outcomes, but their use might be beneficial with regard to tooth wear.

Incisor intrusion, Dentition analysis, Cephalometric analysis - No evidence based conclusions could be drawn.