Background: Near-infrared spectroscopy (NIRS) has been used to study regional cerebral blood oxygen saturation (rScO2) in patients in the prone position.
Objectives: We aimed to test the hypothesis that head rotation more than 45° would affect the rScO2.
Design: A prospective, controlled, single cohort study.
Setting: University Hospital specialising in spinal surgery.
Patients: Fifty-two patients undergoing spinal surgery in prone position were enrolled and 48 completed the study.
Interventions: NIRS sensors were attached to each side of the forehead. Measurements were conducted during steady-state anaesthesia with the head in the neutral position, rotated left, rotated right and returned to the neutral position. Each series consisted of three measurements: resting on the head support, during head lift (to relieve pressure on the tissue at the sensors) and returned to rest on the head support.
Main outcome measures: The differences in rScO2 between the neutral and the turned head positions.
Results: For both left and right sensors, the median differences in rScO2 between neutral and left or right positions were between 0 and -1 with the head up (P = 0.14 to 0.84). The median differences with the head down were between 3.8 and -0.8, with a significant difference for the left sensor when turned left (P < 0.01) and for the right sensor (P = 0.006) when turned right. Ten patients showed reductions of more than 10 in rScO2 in the rotated (and lifted) positions. When the head was lifted from the head support, the rScO2 was -0.5 to 3.75 units higher, but there was high variability between patients.
Conclusion: We recommend the neutral head position for prone patients.TRIAL REGISTRATION clinicaltrials.gov Identifier: NCT01760369.