Background: The Swiss Federal Act on Research involving Human Beings was revised in
2014. To our knowledge, the impact of these new rules regarding research in emergency
situations with persons unable to consent has not been demonstrated yet. The objective of this
work is to assess flow, quality, and effectiveness of informed consent in a clinical trial involving
adults with acute consciousness disorders, the CERTA study (Continuous EEG Randomized
Trial in Adults), aiming at assessing the impact on clinical outcome of continuous
electroencephalography. This analysis could reveal whether the recent laws can be translated
into practice and how they might affect study results.
Methods: We analysed the consent procedure from patients with impaired consciousness
enrolled in the CERTA study. All information was prospectively collected. Patients included in
the CERTA study were adults (18 years) with acute consciousness disorders of any etiology,
treated in an intensive or intermediate care unit of 4 Swiss tertiary hospitals (Centre Hospitalier
Universitaire Vaudois, Hôpital du Valais, Inselspital Bern, Universitätsspital Basel), requiring
an EEG for clinical purposes. The last inclusion criterion was the obtainment of informed
consent for research in emergency situations according to the Human Research Act. Waivers
of informed consent were granted from the Ethics Committee in some specific conditions
(patients dying early, care withdrawal, or relative’s unavailability despite repeated attempts).
Results: 402 patients were randomized. Among those, 7 (1.7%) were excluded (6 double
inclusions, 1 death prior to intervention) and data from 27 (6.7%) could not be used because
of a lack of consent, leaving 368 (91.5%) analysable patients. For each, we found that all steps
of the consent procedure were completed. However, only in 14.1% (52/368 analysable
patients) could a statement of wishes reporting no objection to research before the current
condition be identified in medical charts or with the family, despite at least 3 attempts. In 134
subjects (36.4% of analysable patients), consent was waived according to a per protocol
procedure allowing analysis of patients dying early, while 234 (63.6% of analysable) patients
had an informed consent. Among the latter, 77 (20.9%) were post-hoc, 102 (27.7%) were by
proxy, and 55 (14.9%) patients had both a post-hoc and a proxy consent.
Conclusion: While the consent process in this particular setting seems applicable, certain
points appear difficult to translate into clinical practice. A statement of wishes reporting no
objection to research could be retrieved in only 14.1% of recruited patients, after several
attempts. Informed consent was obtained in only approximately 2/3 of patients, more frequently
by proxy, and was waived in approximately 1/3 according to an exception stipulated in the
protocol, which shows the importance of these waivers for recruitment and the study validity in
such conditions. Also, nearly 7% of enrolments did not comply with the consent process and
had to be completely discarded, not only potentially biasing outcome assessment (only in
surviving patients could consent be refused) but also preventing analysis of reasons for this.
Further trials in this setting should aim at allowing analysis of at least some non-sensitive data
of patients with consent refusals, and specifically foresee inclusion of patients reaching early
an endpoint for the study.