Abstract
We appreciate the concerns raised in the letter by Mr. Oudmaijer et al and note that we considered many of those same issues when planning our review. The key difference between our approach and the one they propose is a fundamentally different assessment of the heterogeneity of the literature. Studies on injury associated with artificial turf vary according to a vast number of factors. These include
Sport
Level of play, ranging from youth sports to professional leagues
Sex of the participant
Practice included or only gameplay
Definition of injury
Reporting standards for injury
Ad hoc versus predetermined cohorts
Industry sponsorship
Type of artificial turf
Mathematically combining the results of studies with this much variation simply should not be done. There is no heterogeneity statistic that would be reassuring enough to overcome that obvious concern. We do agree that the quality of the included studies as assessed by risk of bias and effect size is a significant issue that would matter if a global meta-analysis were itself a valid approach.
Our approach in dealing with the variability in the literature was to accept the variability and report everything. Formal quality assessment does not overcome the problem of heterogeneity. For instance, the studies of professional American football are of much higher quality than those of youth soccer, but it is difficult to argue that studies of either sport are relevant to each other. Additionally, quality assessment tools may reflect the bias of the assessor. Most recently, the meta-analysis on soccer injuries cited by the authors rated industry-sponsored studies with selected cohorts as high quality and included them in their pooled analysis.
How accurately one can reduce a set of disparate studies to a mathematical outcome is a matter of debate. We believe the literature on injuries involving artificial turf is suitable for qualitative analysis only. Nevertheless, we did identify broad patterns and likely further areas of research while avoiding the pitfalls and obfuscation involved in assigning mathematical weights to studies that simply should not be combined.
Footnotes
One or more of the authors has declared the following potential conflict of interest or source of funding: G.G. has received consulting fees and travel expenses from Pargon 28 Inc. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.
