A Protocol for Propofol-Infusion Drug-Induced Sleep Endoscopy

Abstract

Objective

The objective of this study was to outline a protocol utilizing propofol infusion without an initial bolus during drug-induced sleep endoscopy (DISE). We define normative values for final propofol infusion rate (P^final) during DISE and sedation depth values at P^final.

Study Design

Retrospective chart review.

Setting

Tertiary academic hospital.

Methods

A review of patients with obstructive sleep apnea who underwent DISE between 2016 and 2020 was performed. The following patient data were recorded: demographics; DISE procedure details, including P^final, time to P^final, frequency and cadence of infusion rate changes, depth of sedation as measured by Bispectral Index and SedLine values, and hemodynamics; and polysomnography details including apnea-hypopnea index severity and minimum oxygen saturation. A mixed linear model adjusted for age and body mass index was performed for the analysis of effects on P^final. Pearson correlation coefficients determined the strength of association between depth of sedation measured and pattern of collapse on DISE and P^final.

Results

There were 246 patients who met inclusion criteria. P^final resembled a normal distribution (mean ± SD, 156.44 ± 26.69 mcg/kg/min; median, 150 mcg/kg/min). Analysis demonstrated that P^final was influenced by male sex, current smoker status, time to P^final, and number of propofol dose changes (P < .05). Depth of sedation categories measured differently between Bispectral Index and SedLine (55-65 vs 45-55, P < .001). The pattern including severity of collapse on DISE was not associated with P^final (P > .05). No patients required intra- or postoperative respiratory support beyond oxygen via nasal canula.

Conclusion

We describe a propofol slow-infusion DISE protocol that demonstrates safe and reproducible outcomes.