<i>A priori</i> feasibility success criteria.

Background
Physical rehabilitation of critically ill patients may improve physical outcomes; however, the relative benefits and risks with patients requiring vasoactive drugs is currently unknown. A feasibility study is needed to inform the design of a future trial required to address this issue.
Methods
A two-phase exploratory observational feasibility study was carried out:

A retrospective study to clarify the current practice of rehabilitation with patients receiving vasoactive drugs to inform future trial interventions and design.

A prospective study exploring recruitment and outcome measurement. Intensive care patients receiving vasoactive drugs were recruited and asked about the acceptability of a future trial. The feasibility of using an adverse event tool was measured during rehabilitation. Patients were followed up after 60 days to describe the feasibility of measuring outcomes for a future trial.

Results
Retrospective study (n = 78): Twenty-one percent of patients took part in physical rehabilitation whilst receiving vasoactive drugs. Of 321 days with vasoactive drugs administered, physical rehabilitation occurred on 27 days (8%).
Prospective study (n = 40): Eighty-one percent of participants indicated acceptability of being recruited into a future trial (n = 37). Eighty-eight percent of clinicians found it acceptable to randomise patients into either early rehabilitation or standard care. The adverse event tool was implemented by researchers with 2% loss of information. Finally, a 100% follow-up rate at day 60 was achieved for mortality outcomes. Follow-up rates were 70% for the EQ-5D (5 level), 65% for the World Health Organisation’s Disability Assessment Schedule 2.0 and RAND 36-item Health Survey 1.0 and 26% for the 6-minute walk test.
Conclusions
This study found a low frequency of physical rehabilitation occurring with intensive care patients receiving vasoactive drugs. A high proportion of clinicians and patients found a future RCT within this patient group acceptable. Mortality and patient-reported outcomes were the most feasible to measure.