Pooling Results by Means of Meta-Analysis

Abstract

In this issue, we present 2 different meta-analyses that aimed to answer the same question, namely, “Is tibiofibular syndesmosis transfixation best performed with metallic or bioabsorbable screw fixation?” Keep in mind that a meta-analysis is a quantitative review that pools data from publications and attempts to define the heterogeneity of the data to be pooled into a larger sample. Data that are too heterogenous may not be able to be pooled. With regard to the meta-analyses that we present in this issue, one author group came to the conclusion that absorbable fixation was preferable to metallic fixation, whereas the other group concluded that metallic fixation, if not removed, is preferable. The Table briefly compares key elements of each of these reports. Moreover, a critical review of each report shows that, in the van der Eng et al (1van der Eng D.M. Schep N.W.L. Schepers T. Bioabsorbable versus metallic screw fixation for tibiofibular syndesmotic ruptures: a meta-analysis.J Foot Ankle Surg. 2015; 54: 657-662Google Scholar) report, a return to the operating room to remove metallic fixation devices was considered a routine component of the surgical intervention and not a complication. In the Xie et al (2Xie Y. Cai L. Deng Z. Ran B. Hu C. Absorbable screws versus metallic screws for distal tibiofibular syndesmosis injuries: a meta-analysis.J Foot Ankle Surg. 2015; 54: 663-670Scopus (13) Google Scholar) report, to the contrary, a return to the operating room for metallic hardware removal was considered a complication. This distinction is made clear in each of the reports, although this would not be evident upon just a cursory inspection of the title and abstract for each article. Readers are advised and encouraged to keep this type of comparison in mind when they read any set of articles that focus on the same subject, in particular, systematic reviews and meta-analyses of the same topic.TableKey elements of 2 meta-analyses aiming to answer the question, “Is tibiofibular syndesmosis transfixation best performed with metallic or bioabsorbable screw fixation?”Xie et al 2Xie Y. Cai L. Deng Z. Ran B. Hu C. Absorbable screws versus metallic screws for distal tibiofibular syndesmosis injuries: a meta-analysis.J Foot Ankle Surg. 2015; 54: 663-670Scopus (13) Google Scholarvan der Eng et al 1van der Eng D.M. Schep N.W.L. Schepers T. Bioabsorbable versus metallic screw fixation for tibiofibular syndesmotic ruptures: a meta-analysis.J Foot Ankle Surg. 2015; 54: 657-662Google ScholarDatabases▪Medline via PubMed▪Embase▪Cochrane Library▪Chinese Biomedical Database▪MEDLINE via Ovid▪Google ScholarMethod of reportPreferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (PRISMA statement)Preferred Reporting Items for Systematic reviews and Meta-Analyses guidelines (PRISMA statement)Search termsabsorbable, bioabsorbable, biodegradable, degradable, polyglycolide, AND distal tibiofibular syndesmotic injury, tibiofibular syndesmosissyndesmo* (all fields) and absorbable (all fields)Inclusion▪RCT or quasi-randomized trials▪Any language▪Any time▪RCT and comparative studies▪English language▪January 1994 to June 2014Outcomes▪Complications▪Functional scores▪Complications▪Functional scoresAbstractionTwo independent reviewers, a third tie breaker, using Cochrane guidelinesTwo coauthors, using Cochrane guidelinesVariables of interest▪Publication details▪Study design▪Number of patients, Patient age and gender▪Duration of follow-up▪Duration of surgery▪Complications▪Functional outcomes▪Publication details▪Study type▪Number of patients▪Patient age▪Fracture type▪Complications▪Functional outcomesComplications of interest▪Infection▪Foreign body reaction▪Heterotopic osteoarthritis▪Broken screws (Not all metallic screws were routinely removed in the studies included in the review.)▪Infection▪Foreign body reaction▪Heterotopic ossification▪Arthritis followed by arthrodesis▪Bioabsorbable screw remnant requiring removal▪(All metallic screws were routinely removed in the studies included in the review.)Quality assessmentCochrane tool for assessing risk of biasCochrane tool for assessing risk of biasStatistical plan▪Cochrane Review Manager software, version 5.2▪Relative risk (dichotomous variables)▪Δ mean (continuous variables)▪95% CI, α ≤ 0.05▪I2 >50% considered heterogenous, heterogeneity explored using X2 (p < .1)▪Cochrane Review Manager software, version 5.3▪Relative risk (dichotomous variables)▪Δ mean (continuous variables)▪95% CI, α ≤ 0.05▪I2 <30% considered homogenous, heterogeneity explored using X2 (p < .1)Potentially eligible studies362 published reports18 published reportsIncluded studies▪16 RCTs▪N = 1208▪615 (50.9%) absorbable▪593 (49.1%) metallic▪4 RCTs▪N = 260▪137 (52.7%) absorbable▪123 (47.3%) metallicResults Functional outcomesNo difference between fixation groupsNo difference between fixation groups Overall complicationsNot described in “overall” termsNo difference between fixation groups (subgroup analysis showed statistically significantly more when PLLA used) Fixation breakageNo difference between fixation groupsNone observed ReoperationStatistically significantly more with metallic fixation (considered a complication)Significantly more prevalent in metal (not considered a complication) Foreign body reactionStatistically significantly more with absorbable fixationObserved InfectionNo difference between fixation groupsObserved PainNo difference between fixation groupsNot mentioned Heterotopic ossificationNo difference between fixation groupsObserved Absolute number of complicationsMore in the metallic group (hardware removal considered a complication)More in the absorbable fixation group (hardware removal not included as a complication) Crude costNo difference between fixation groupsNot mentioned Duration of surgeryNo difference between fixation groupsNot mentioned Time to fracture healingNo difference between fixation groupsQualitatively no differenceSuture-button fixationNot consideredNot consideredPublication biasFunnel plots suggest notFunnel plots suggest notConclusionsAbsorbable fixation preferable because complications related to metallic fixation are more serious than foreign body reaction and require removalIf do not take metallic screws out, then absorbable offers no advantage Open table in a new tab