Article Text
Abstract
Aims The number and size of randomised controlled trials (RCT) in UK paediatric intensive care units (PICU) is expanding due to improved infrastructure support and regulatory incentives for trials involving children. PICUs need to continue to expand trial capacity whilst minimising disruption to patients, parents and staff. Enrolment into multiple trials could lead to increased recruitment and quicker delivery of results for optimal care. The acceptability of co-enrolment to patients and practical issues for staff have not yet been assessed. We describe how CATCH (Catheters in children) - a RCT comparing impregnated with standard central venous catheters (CVC) for reducing bloodstream infection in PICU - has addressed challenges to co-enrolment.
Methods Experiences of success and failure of co-enrolment at PICUs participating in CATCH were recorded. We obtained reasons for PICUs declining to co-enrol children into multiple studies.
Results Eleven months after CATCH began recruiting, 4/7 units were also recruiting to another trial (CHiP-Control of Hyperglycaemia in Paediatric intensive care). Of these, 3/4 declined co-enrolment due to concerns about jeopardising CHiP recruitment, asking too much of parents, and overwhelming amounts of information to explain two trials to parents.
Within the one unit adopting co-enrolment, emergency admissions were firstly randomised to CATCH if a CVC was required, then approached for CHiP if eligible, and finally approached for CATCH as deferred consent was allowed in the study protocol. Of 93 emergency patients randomised to CATCH, 37 were eligible for CHIP. Of these, 17 consented to both; 1 consented to CHiP but refused CATCH; 12 refused CHiP but consented to CATCH; 5 declined both; 2 died awaiting CATCH consent.
Research nurses found it easiest to consent sequentially although simultaneous consent was successful on both occasions when this was attempted.
Conclusions Co-enrolment is feasible and acceptable for this low risk, pragmatic trial. Lack of recruitment due to co-enrolment was minimal, and consent appeared to be more influenced by the type of treatment involved. Strategies need to be developed to combine presentation of information for multiple trials.