Abstract
Rationale: Patients eligible for randomized controlled trials may not be enrolled for various reasons. Nonenrollment may affect study generalizability and lengthen the time required for trial completion.
Objectives: To describe characteristics and outcomes of eligible nonenrolled (ENE) patients in a multicenter trial of mechanical ventilation strategies.
Methods: Within the OSCILLATE trial of high-frequency oscillation (HFO) versus conventional ventilation (CV) in adults with adult respiratory distress syndrome, and with approval from research ethics boards, we collected a minimal dataset on patients who satisfied eligibility criteria but were not enrolled. We categorized ENE patients as ENE-HFO and ENE-CV based on receipt of HFO at any time. We used multivariable logistic regression to assess the association between ENE status and mortality.
Measurements and Main Results: A total of 548 patients were randomized, and 546 were ENE. The most common reasons for ENE were no consent (42%), physician refusal (24%), missed randomization window (15%), and current HFO use (14%). Compared with randomized patients in respective arms of the trial, ENE-HFO patients were younger and had worse lung injury, whereas ENE-CV patients had lower illness severity. ENE status was independently associated with mortality (adjusted odds ratio, 1.39; 95% confidence interval, 1.06-1.84; P = 0.02), with no significant interaction with ventilation treatment group.
Conclusions: Nonenrollment was common, with approximately one ENE patient for every randomized patient. Our study suggests that enrollment in trials of mechanical ventilation may be associated with improved outcomes compared with standard care and highlights the need for prospective tracking and transparent reporting of ENE patients as part of trial management.