The comparability of clinical characteristics between single-arm trials and objective performance criteria in clinical research
2School of Acupuncture-Moxibustion and Tuina, Beijing University of Chinese Medicine, China
3First Clinical Medical College of Beijing University of Chinese Medicine, China
4Dongzhimen hospital, Beijing University of Chinese Medicine, China
Background: Single-arm trials (SATs) with objective performance criteria (OPC) or performance goals (PG) have turned out to be a rational design for clinical effectiveness evaluation and facilitation of medical device approval.
Objectives: To compare the baseline characteristics and outcomes between single-arm trials (SATs) and objective performance criteria (OPC) in clinical research.
Methods: We searched PubMed, Embase, the Cochrane Library, and four databases in China from inception to April 2022. SATs with the OPC as an external comparator were included without any restriction on interventions. Furthermore, the eligibility criteria, demographic characteristics data, as well as the outcomes of the SATs, and their matched OPC sources were compared.
Results: We identified 869 citations from seven databases and 76 SATs with OPC were included. Most interventions were medical devices (80.3%, 61/76), of which 80% of studies focused on vascular diseases. Sixty-two (81.6%) SATs set one OPC, and 14 SATs (18.4%) had multiple OPC for different outcomes. Fifty-eight SATs (76.3%) provided the obtained sources of OPC through references. All SATs and their matched OPC belonged to the same disease categories, with some of them being kind of different in the types or stages. Age differences existed with a range from 3 to 20 years in 44.8% (22/49) of included studies between the SATs and their matched OPC. Moreover, whether there was an age difference was statistically associated with the effect size (χ2=4.191, p-value=0.004). Fifteen SATs (25.9%) selected OPC with a more serious condition. For the outcome selection, 76.5% were completely consistent between SATs and their matched OPC in measures; however, 35.9% of them were inconsistent in observation time points.
Conclusions: The inconsistency in some demographic characteristics data and outcome indicators selection has been identified between the SATs and their matched OPC. Comparability of the most important prognostic factors between SATs and their matched OPC should be critically evaluated not only before the determination of OPC, but also when designing statistical analysis plans and drawing conclusions to enhance the reliability of OPCSATs.