Back

Trends and landscape of cardiovascular disease studies characteristics on ClinicalTrials.gov from 2012 to 2021

Rawlley, B.; Bansal, K.; Dayal, U.; Julka, D.; Salooja, I.; Cordova Sanchez, A.; Gupta, K.; Kumar, S.; Chaudhuri, D.

2023-06-20 cardiovascular medicine
10.1101/2023.06.12.23291308 medRxiv
Show abstract

IntroductionTo analyze cardiovascular disease (CVD) studies from the United States registered on ClinicalTrials.gov focusing on characteristics associated with their external validity; the presence of Data Monitoring Committee/Data Safety Monitoring Board (DMC/DSMB), randomization, blinding, and gender of the principal investigators (PIs). MethodsWe queried the Application Programming Interface (API) of ClinicalTrials.gov to extract data on characteristics of the identified studies, most notably, DMC/DSMB status, Design Allocation, Design masking, and gender of PIs. ResultsWe identified 536 studies pertinent to CVD for this analysis. Out of 536, 471 (88%) were interventional studies (Clinical trials) and 65 (12%) were observational studies with interventions. Amongst these, 261 (49%) reported having a DMC, 312 (66%) reported randomization, 224 (48%) reported masking and 122 (19%) of the PIs were women. No trend was seen in the annual proportion of studies with DMC, randomization, masking, and women as PIs (P-trend > 0.05 for all). Multivariable logistic regression analyses were notable for higher odds of DMC (aOR, 5.28; 95% CI, 2.70 - 10.90; P < 0.05) and blinding (aOR, 2.42; 95% CI, 1.29 - 4.64; P < 0.05) in NIH-funded studies and higher odds of being terminated/suspended or withdrawn in basic science studies (aOR, 2.83; 95% CI, 1.07 - 6.90; P < 0.05). No relation was seen between any characteristics and the study being completed. ConclusionsWe report on the absence of DMCs, randomization, blinding, women as PIs, and lack of cross-gender collaboration in the leadership of CVD studies without any favorable trend over the past decade. This calls for comprehensive efforts to improve these trends and ultimately, the external validity of studies. We also call for an overhaul of the definition of the phase of a clinical trial that centers around a drug being the intervention.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.