Discrepancies between registered protocol and final publication in exercise interventions for chronic low back pain: a meta-research study

Date & Time
Wednesday, September 6, 2023, 12:30 PM - 2:00 PM
Location Name
Session Type
Understanding and using evidence
Bargeri S1, Basso G2, Geraci I2, Castellini G1, Chiarotto A3, Gianola S1, Ostelo R3, Testa M2, Innocenti T3
1Unit of Clinical Epidemiology, IRCCS Istituto Ortopedico Galeazzi, Milan, Italy, Italy
2Department of Neurosciences, Rehabilitation, Ophthalmology, Genetic and Maternal Infantile Sciences (DINOGMI), University of Genova-Campus of Savona, 17100 Savona, Italy., Italy
3Department of Health Sciences, Faculty of Science, VU University Amsterdam, Amsterdam Movement Sciences, The Netherlands., The Netherlands

Background: Discrepancies between registered protocols and final publications are common in randomized controlled trials (RCTs) of various medical fields, undermining their integrity and credibility. However, this has not yet been assessed in exercise RCTs for chronic low back pain (cLBP).
Objectives: We aimed to assess the prevalence of discrepancies between the registered protocol and final publication in cLBLP exercise RCTs.
Methods: This is a meta-research study, prospectively registered (doi: 10.1101/2023.02.27.23286399). We started from the RCTs included in the 2021 Cochrane review “Exercise therapy for chronic low back pain” and identified in its update (n=421) to select all RCTs reporting a protocol registration on a primary register of the World Health Organization International Clinical Trials Registry Platform or in ClinicalTrials.gov. Standardized data collection forms were developed to record information from both registration and publication. We detected discrepancies for primary and secondary outcomes, outcomes measurement, and timepoints between the registered protocol and final publication adapting a previous published method. We used descriptive statistics to assess the proportion of RCTs with and without a discrepancy.
Results: Overall, we included 121 RCTs reporting an available protocol registration. Preliminary results from the first 20 studies showed that most protocols were registered on ClinicalTrials.gov (n=11; 55%) with less than half (n=9; 45%) prospectively registered. Overall, we found 31 different discrepancies in reporting outcomes between protocol registrations and publications: 6 for primary outcomes (n=2 in outcome definitions, n=2 in measurements, n=2 in timepoints) in 4 RCTs (20%) and 25 for secondary outcomes (n=17 in outcome definitions, n=5 in measurements, n=3 in timepoints) in 17 RCTs (85%). The most frequent discrepancy was addition of primary (n=2) or secondary outcomes (n=10) (Table 1).
Conclusions: From our preliminary findings, we found widespread discrepancies in one fifth of RCTs for primary outcomes and in four fifths of RCTs for secondary outcomes. The presence of such discrepancies can threaten the validity of RCTs as well as systematic reviews or guidelines for clinical practice.
Patient, public and/or healthcare consumer involvement: The assessment and reporting of efficacy and safety outcomes should be mandatory to fulfil ethical obligations toward patients for fully informed decision-making in the healthcare system.