Relative and bedside nurse assessment of comfort and communication during propofol, dexmedetomidine, or clonidine-based sedation: pre-planned analysis within the A2B RCT

Background Optimising comfort and ability to communicate for mechanically ventilated (MV) intensive care (ICU) patients is a priority for clinicians, ICU patients and their relatives. Current usual care is propofolbased sedation plus an opioid analgesic. The alpha2-agonists dexmedetomidine and clonidine are potential alternative sedatives. Objective(s) To explore whether nurses and relatives perceive patients sedated with dexmedetomidine and/or clonidine appear more awake, comfortable, and cooperative than patients receiving only propofolbased sedation. Design and methods Sub-study within an open-label three-arm trial. Setting and participants 41 ICUs in the United Kingdom. 1437 adults receiving propofol ±opioid for sedation-analgesia within 48 hours of starting MV, expected to require ≥48 total hours of MV. Interventions Light sedation was targeted in all patients unless clinicians requested deeper sedation. In intervention groups algorithms promoted alpha2-agonist up-titration and propofol down-titration followed by sedation primarily with allocated alpha2-agonist. Usual care was propofol-based sedation. Intervention continued until patients were successfully extubated (primary outcome), or other pre-defined end-points. Outcomes For each 12-hours care period nurses responded to two ‘yes/no’ questions: Is the patient able to communicate pain? Is the patient able to cooperate with care? When the patients’ personal legal representative visited, they were asked for ‘yes/no’ responses to three questions: does the patient appear awake? Does the patient appear comfortable? Does the visitor feel they can communicate with the patient? Intervention versus propofol group responses were compared fitting a generalised linear mixed model, with results expressed as Odds Ratios (OR; 95% confidence intervals); ORs >1 indicated greater probability of a ‘yes’ response. Results Nurse responses were available for >90% of trial patients (mean (SD) 12 (12) care periods per patient). Comparing dexmedetomidine versus propofol groups, the OR for a ‘yes’ response to ‘communicate pain’ was 1.38 (1.08 to 1.75), and for clonidine versus propofol was 1.13 (0.89 to 1.43). For ‘cooperate with care’ comparing dexmedetomidine versus propofol groups, OR was 1.14 (0.98 to 1.32), and for clonidine versus propofol 0.96 (0.83 to 1.12). Relative responses were available for 32-34% of trial patients across groups (mean (SD) 3 (3) days per patient). For the ‘appear awake’ question, the dexmedetomidine versus propofol group OR was 1.48 (1.04 to 2.10), and clonidine versus propofol 1.35 (0.95 to 1.91). For ‘appear comfortable’ the dexmedetomidine versus propofol group OR was 0.64 (0.38 to 1.09), and for clonidine versus propofol 0.78 (0.45 to 1.34). For the ‘feel they can communicate’ comparison the dexmedetomidine versus propofol group OR was 1.00 (0.68 to 1.47), and for clonidine versus propofol 1.05 (0.71 to 1.54). Limitations Interventions were unblinded, with risk of bias; missing data may not have been at random. Conclusions Nurses perceived patients receiving dexmedetomidine-based sedation could better communicate pain than with propofol-based sedation, and relatives perceived patients appeared more awake. No differences for the other questions were found, or for the clonidine versus propofol comparisons, although some uncertainty remains due to the wide confidence intervals. Future work Additional mixed methods research of sedation quality with different agents from staff and relative perspectives. Study registration: NCT03653832