Prohibiting deferred consent can lead to research bias in clinical trials within Paediatric Intensive Care Units

Abstract

The conduct of clinical trials under time constraints in the paediatric critical care setting may pose an ethical dilemma for researchers. Many life-saving interventions do not allow adequate time to obtain informed, prospective consent from parents or legal guardians. The rapid decision process in the paediatric critical care environment challenges the ethically accepted gold standard of consent prior to interventional research. In time-critical research, provision is made to obtain consent after the intervention has taken place. This concept is often referred to as deferred consent but may also be referred to as delayed consent, consent to continue, or research without prior consent. Deferred consent aims to progress critical care research by mitigating the limitations of consent faced by researchers. If the intervention pertains to the treatment of critically ill children, the use of deferred consent may be required to perform the intervention pragmatically and to capture a true representation of the study population. Without deferred consent it is possible that patients requiring an immediate life-saving intervention may be missed. This may result in selection bias thus affecting study outcomes and the generalisability of research findings. This thesis examined findings from a nested study to investigate whether differences in consent requirements led to significant differences in the study populations and outcomes of an existing multicentre, interventional, clinical trial within Paediatric Intensive Care Units. The parent trial, the Transnasal Humidified Rapid-Insufflation Ventilatory Exchange in children requiring emergent intubation (Kids THRIVE) trial, examines a system of high flow oxygen delivery during the emergency intubation of children. The Kids THRIVE Study sought approval to use deferred consent in circumstances where prospective consent was not possible. Measures of severity of illness were used to compare participants recruited to the Kids THRIVE trial using deferred consent with those recruited using prospective consent. Severity of illness was measured in ventilator hours, hospital and intensive care length of stay and PIM3 codes. Results indicate that children recruited to the Kids THRIVE trial using deferred consent differed clinically and statistically to those children recruited using prospective consent. Children in the deferred consent group were more unwell and had less favourable clinical outcomes. They were mechanically ventilated for longer (median [IQR] 70.97 [35.58-122.30]vs 48.85 [27.41-120.91] hours, P=0.001); spent longer in intensive care (median [IQR] 8.84 [4.55-18.38] vs 7.17 [4.25-14.29] days, P= 0.020); and had a longer hospital stay (median [IQR] 27.95 [12.69-63.59] vs 19.77 [10.28-39.96] days, P=0.010) than their counterparts in the prospective consent group. PIM3 scores which predict a percentage probability of death did not show any statistically significant difference between the two groups. This research provides evidence that the exclusive use of prospective consent in paediatric critical care research can lead to selection bias by recruiting a study population that is not representative of paediatric intensive care patients. Alternatives to prospective informed consent need to be considered to progress research in this patient cohort.