Research waste in total knee arthroplasty studies: an observational study on registered trials

Timely publication of preregistered study outcomes is not self-evident. Discrepancies can lead to significant research waste.

To assess timely (within 7 years) and consistent publication of preregistered primary outcomes and associated factors of total knee arthroplasty (TKA) studies registered between 2000 and 2017 over time.

An observational study.

ClinicalTrials.Gov, MEDLINE, Embase, Cochrane Library, Web of Science, PubMed and Google Scholar.

Registered TKA trials at ClinicalTrials.Gov between 2000 and 2017.

ClinicalTrials.Gov's required and optional data elements for registering a study and the preregistered and published primary outcome, defined as the outcome stated in the primary outcome field on ClinicalTrials.Gov. We used descriptive statistics, Kaplan-Meier curves and Cox regression analyses.

1352 registered TKA (1072 interventional; 280 observational) studies were included, with 967 (811 interventional; 156 observational) unique references. Regarding the publication of preregistered primary outcomes within 7 years, the results for interventional trials were 0% (2000), which increased to 59.6% (2017). Observational studies were timely published in 0% (2000) and 37.5% (2017). Interventional trials and observational studies not funded by industry were more likely to have timely and consistent publication of their primary outcomes. Drug intervention trials were more likely to be timely and consistently published than procedure-focused trials. Phase 3 interventional trials were more likely, while phase 1 trials were less likely to be consistently published on time.

Despite ongoing efforts to improve publication rates, over a third of interventional trials remain unpublished within 7 years. For observational studies, the rate is even lower, with only two-fifths published on time, contributing to significant research waste.

CRD42021246599.