Social media and orthodontics: A commentary on a systematic review

Abstract

Data sources The electronic databases MEDLINE, EMBASE, Google Scholar, the Cochrane Oral Health Group's Trials Register, and CENTRAL from 1946 to 31 March 2018 were searched to identify eligible studies. Information sources in the Grey literature were also searched.

Study selection Randomised and non-randomised studies as well as retrospective studies irrespective of their language were selected by two reviewers independently.

Data extraction and synthesis Data extraction and risk of bias assessments were performed by two reviewers independently. Data were synthesised qualitatively. Quantitative syntheses were not possible because of high heterogeneity.

Results One randomised controlled trial, four cross-sectional studies, and three retrospective studies were selected according to the eligibility criteria. The quality of the evidence in the included studies was predominantly of low to moderate quality. All studies reported on Twitter, three on YouTube, two on Facebook and one study referred also to Google+, Pinterest, Instagram, and Internet Blogs. The included studies reported on the informational value of social media on various aspects of orthodontic treatment and patients' health experiences during orthodontic treatment.

Conclusions The authors of this review concluded that social media represent a basic resource for exchanging knowledge on a wide variety of aspects of orthodontics.

A Commentary on

Papadimitriou A, Kakali L, Pazera P, Doulis I, Kloukos D.

Social media and orthodontic treatment from the patient's perspective: a systematic review. Eur J Orthod 2019: DOI: 10.1093/ejo/cjz029. [Epub ahead of print] PubMed PMID: 31107.943.

GRADE rating

Commentary

Objectives and key findings of the systematic review

The potential role of social media's role in orthodontics is an important research topic not only for patients, but also for clinicians, researchers, and research sponsors. The objectives of this systematic review by Papadimitriou et al.¹ was to assess the interrelationship between orthodontics and social media in relation to (1) the use of social media by patients and potential patients to obtain information on orthodontics (2) the type and quality of the information obtained and (3) how this information is used and how it influences these stakeholders. One randomised controlled trial, four cross-sectional studies, and three retrospective studies were identified to address these objectives. Evidence was predominantly rated as low to moderate quality. The authors concluded that social media represent a basic resource for exchanging knowledge on a wide variety of items on orthodontics.

Methods of our critical appraisal of this systematic review

In this commentary we assessed how this systematic review was reported using the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) checklist.^2,3 We further appraised the methodological validity and the risk of bias of this review with the AMSTAR 2⁴ and the ROBIS^5,6 tools, respectively. All reporting and quality assessments were conducted by two reviewers (RMR and LI) independently. Disagreements between these two operators were resolved through discussions till consensus was reached.

Key findings of our critical appraisal of the systematic review

Tables 1, 2 and 3 report the findings of our critical appraisals and Table 4 summarises the following eight key limitations of this review. First, it was not reported whether research methods were pilot tested. Such tests are important for the fine-tuning of the methods and for the calibration of the two operators who select eligible studies and extract data. Second, adverse effects of interventions were not assessed. Adverse effects should have been assessed to give a balanced perspective on the use of social media in orthodontics. Third, the review protocol was not registered nor was it published a priori. This shortcoming could have introduced bias related to selective reporting of outcomes and the risk of possible post-hoc changes in the review that are not congruent with the original protocol.⁷ Fourth, almost 90% (249/279) of studies were excluded based on a screening of titles only. This method could have excluded potentially eligible studies and therefore have introduced 'selection bias', that is, the eight included studies in this review could be just a selective sample of a larger group of eligible studies. Study selection based on screening of titles only is also not in agreement with Cochrane's typical procedure of selecting studies, that is, examining both the titles and abstracts of research studies to remove the irrelevant reports.⁸ Fifth, the authors gave only the references with rationale of 4, but not of all excluded studies. This lack of reporting jeopardises the reproducibility of the review. Researchers who want to update this review like to see the references of all papers that were excluded in the review with rationale. Sixth, risk of bias for the one included randomised controlled trial was assessed with an outdated 2011 Cochrane risk of bias (RoB) tool for randomised controlled trials.⁹ Since the publication of this 2011 reference, numerous updates of this tool have been published and archived.^10,11 A more recent version of this RoB tool, for example, RoB 2, should have been consulted and implemented for this review. Further, risk of bias in the retrospective and cross-sectional studies was not assessed. Instead quality assessments of these studies were conducted with tools with crucial shortcomings, that is, the quality of the eligible retrospective studies was assessed with a non-validated tool¹² and the quality of the cross-sectional studies was assessed with the Newcastle Ottawa scale which has shown to give low reliability between reviewers and lacks good guidance manuals.^13,14,15 Seventh, the authors reported on the quality of individual studies, but did not give the strength of evidence for each main outcome, which is essential information for end-users of systematic reviews. Eight, the authors did not report on selective reporting within the eligible studies.

Table 1 PRISMA checklist scores for the systematic review by Papadimitriou 2019 et al.¹

Table 2 AMSTAR 2 scores for the systematic review by Papadimitriou 2019 et al.¹

Table 3 Tabular presentation for ROBIS results for the systematic review by Papadimitriou 2019 et al.¹

Table 4 Limitations of the systematic review by Papadimitriou 2019 et al.¹

What are the implications and what should change now?

The eight limitations identified in this critical appraisal and the low to moderate quality evidence identified in the eligible studies jeopardise the trustworthiness of this review. These shortcomings should be carefully considered when translating the findings of this review into practice.

So what should change? Journals, editors, and peer reviewers have an important role as gatekeepers of research quality. These stakeholders can apply a variety of strategies to improve the quality of their journal articles. Strategies before the publication of a review include: 1) not accepting reviews whose protocols were not registered or published a priori; 2) appraising each submitted review with the PRISMA, AMSTAR 2, and ROBIS tools; and 3) assessing the reproducibility of the review. Strategies after the publication of a review include: withdrawals, retractions, and corrections of reviews. Implementing these strategies by journals, editors, and peer reviewers will eventually reduce the overall research waste. Many stakeholders will benefit.

Data sharing

All raw data sheets of our critical appraisals with the PRISMA, AMSTAR 2, and ROBIS tools can be requested from the corresponding author (RMR) at . We will respond rapidly to further clarify any questions on our data and conclusions.

Practice points

• This systematic review concluded that social media represent a basic resource for exchanging knowledge on a wide variety of patient-centred issues relating to orthodontics.

• The quality of evidence in the included studies was predominantly of low to moderate quality.

• Our critical appraisal identified eight key shortcomings of this review, which downgrade the trustworthiness of the findings of this paper. These shortcomings and the low to moderate quality of the evidence should be carefully weighed when implementing the findings of this systematic review.