Analysis of Reference Practices among Practicing Orthopaedicians in India

Analysis of Spin in RCTs of Spine Surgery Using ORG-LOC Grading Tool

BACKGROUND

Even in highly credible research models, such as randomised control trials (RCTs), many pitfalls do exist that a practitioner must be aware of, to get the actual sense of the research. The one such pitfall that is much more common but ironically less explored is the Writers' bias or the spin. Particularly in the abstracts, it is a potential source of deception to the readers.

METHODS

We selected 250 recent RCTs from the top 5 spine journals. Baseline data and CONSORT Adherence Score (CAS) were collected. Abstracts of the RCTs were graded using the level of confidence (LOC) grading tool developed by the Orthopaedic Research Group. All possible associations of spin were studied to assess the significance.

RESULTS

The median CAS was 11 (IQR 10-12). Only 47.6% (n = 119) articles had High LOC with no or one non-critical spin in abstract. 12.4% (n = 31) had Moderate LOC and 28% (n = 70) had Low LOC. The rest had Critically Low LOC with more than one critical spin. Of the variables analyzed in multivariate regression analysis, only CAS had a (negative) correlation with the LOC of the abstracts.

CONCLUSIONS

Spin-based grading of RCTs is the need of the hour to aid readers to interpret the true essence of research papers. 40% of the RCTs in top spine journals had low to critically low LOC. Objective structuring of abstracts with adherence to CONSORT guidelines is the way forward to prevent spin.

"Only 50% of randomized trials have high level of confidence in arthroscopy and sports medicine"-a spin-based assessment

PURPOSE

Pioneering works on the quality appraisal of RCTs have recognized and addressed most of the issues that affect the RCT quality but some issues such as "Writers' bias" or "Spin" are yet to be sorted out. Spin, particularly in the abstracts, is a potential source of deception to the readers. The purpose of this study is to grade the RCTs of arthroscopy and sports medicine based on a spin in their abstracts, analyze the prevalence of spin, and explore methods to remove spin.

METHODS

250 recent RCTs from the top 5 arthroscopy and sports medicine journals were selected. Baseline data of the articles were collected. Consort Adherence Score (CAS) was calculated. The abstracts of the RCTs were graded using the Level Of Confidence (LOC) grading tool developed by the Orthopaedic Research Group. The association of the spin grade with other characteristics of the articles was analyzed.

RESULTS

The median CAS for the included studies was 9 (IQR 8-10). It was found that only 49.6% (n = 124) articles had high LOC with no or one non-critical spin in the abstract. 20.8% (n = 52) had Moderate LOC with more than one non-critical spin. 19.6% (n = 49) had at least one critical spin and 10% (n = 25) had more than one critical flaw making their results have Low and Critically Low LOC, respectively. Of the ten variables analyzed in multivariate regression analysis, it was found that CAS was the only significant factor that determines the level of confidence in the abstract of RCTs CONCLUSION: Spin is prevalent in abstracts of sports medicine and arthroscopy journals with 50.4% having some form of spin. Grading the LOC of the RCTs based on spin is the necessity of the day for the readers. Only 49.6% of the RCTs had high LOC. Objective structuring of the abstracts would help eliminate spin in the future.

LEVEL OF EVIDENCE

Level 1.