L. Carballo-Costa1, R. Costas2, A. Moseley3
1University of A Coruna, Department of Physiotherapy, Medicine and Biomedical Sciences, A Coruna, Spain, 2Centre for Science and Technology Studies (CWTS). Leiden University, Leiden, Netherlands, 3Institute for Musculoskeletal Health, The University of Sydney and Sydney Local Health District, Sydney, Australia
Background: The Physiotherapy Evidence Database (PEDro) is the most important and comprehensive bibliographic database indexing clinical practice guidelines, systematic reviews, and randomised controlled trials evaluating physiotherapy interventions. One of the features that make this database unique is that the methodological quality of trials is appraised using the PEDro scale. The PEDro scale is composed of 11 items scored as present or absent, the presence of 10 items is tallied to produce the total PEDro score (0 to 10). It is feasible for all physiotherapy trials to achieve a score of at least 7 out of 10. To the best of our knowledge, it is unknown which topics in physiotherapy trials have better methodological quality.
Purpose: The purposes of this study are:
- to identify the methodological quality of the most frequent topics detected in randomised controlled trials indexed in PEDro.
- to identify the topics reported in trials that have methodological quality of 7 or more.
Methods: Randomised controlled trials indexed in PEDro and published in January 1979 to July 2019 were matched to an in-house version of Web of Science. Bibliographic data and total PEDro score were extracted for all trials. An advanced bibliometric technique of text mining was applied to the title to extract the relevant topics for all trials. Noun phrases that referred to conditions and interventions were selected to name the topics.
Results: A total of 25,083 trials were included in the analysis. The ten most frequent topics detected were exercise, pain, physical activity, stroke, chronic obstructive pulmonary disease, acupuncture, low back pain, balance, diabetes, and knee osteoarthritis. The mean (standard deviation) total PEDro score for these topics ranged from 5.1 (1.4) for physical activity to 6.1 (1.6) for knee osteoarthritis. Trials that included terms related to whiplash and shoulder pain in the title had methodological quality of 7 or more, on average.
Limitations. The results are only applicable for English-language randomised controlled trials evaluating physiotherapy interventions.
Limitations. The results are only applicable for English-language randomised controlled trials evaluating physiotherapy interventions.
Conclusion(s): The most frequent topics detected in the titles of randomised clinical trials indexed in the Physiotherapy Evidence Database (PEDro) had similar methodological quality, all less than 7 out of 10. Highest methodological quality occurred in trials evaluating conditions including whiplash or shoulder pain.
Implications: To know the current situation of the methodological quality of randomised controlled trials by topics allows us to recognise those fields where the design of the studies is good. This information is useful to clinicians because trial quality impacts on treatment decisions. Researchers can explore strategies to improve trial quality in some topics. Finally, these results provide policymakers with information to support research policies oriented towards the promotion of research in those fields where the methodological quality is currently not good enough.
Funding, acknowledgements: This work has not received any funding.
Keywords: Bibliometrics, Randomised clinical trials methodological quality, Physiotherapy Evidence Database (PEDro)
Topic: Research methodology, knowledge translation & implementation science
Did this work require ethics approval? No
Institution: Universidade da Coruna
Committee: Universidade da Coruna
Reason: This is a bibliometric study
All authors, affiliations and abstracts have been published as submitted.
