Introduction: Research in patients with life-threatening illness such as cardiac arrest is challenging since they can not consent. The Food and Drug Administration addressed research under emergency conditions by publishing new criteria for exception from informed consent in 1996. We systematically reviewed randomized trials over a 10-year period to assess the impact of these regulations.
Methods: Case-control study of published trials for cardiac arrest (cases) and atrial fibrillation (controls.) Studies were identified by using structured searches of MEDLINE and EMBASE from 1992 to 2002. Included were studies using random allocation in humans with cardiac arrest or atrial fibrillation prior to enrollment. Excluded were duplicate publications. Number of American trials, foreign trials and proportion of trials of American origin were compared by using regression analysis. Changes in cardiac arrest versus atrial fibrillation trials were calculated as risk differences.
Results: Of 4982 identified cardiac arrest studies, 57 (1.1%) were randomized trials. The number of American cardiac arrest trials decreased by 15% (95% CI: 8, 22%) annually (P = 0.05). The proportion of cardiac arrest trials of American origin decreased by 16% (95% CI: 10, 22%) annually (P = 0.006). Of 5596 identified atrial fibrillation studies, 197 trials (3.5%) were randomized trials. The risk difference between cardiac arrest versus atrial fibrillation trials being of American origin decreased significantly (annual difference -5.8% (95% CI: -10, -0.1%), P = 0.03).
Interpretation: Fewer American cardiac arrest trials were published during the last decade, when federal consent requirements changed. Regulatory requirements for clinical trials may inhibit improvements in care and threaten public health.