Evaluation of Surgical Learning Curve Effect on Obstructive Sleep Apnea Outcomes in Upper Airway Stimulation

An increasing number of facilities offer Upper Airway Stimulation (UAS) with varying levels of experience. The goal was to quantify whether a surgical learning curve exists in operative or sleep outcomes in UAS.

International multi-center retrospective review of the ADHERE registry, a prospective international multi-center study collecting UAS outcomes. ADHERE registry centers with at least 20 implants and outcomes data through at least 6-month follow-up were reviewed. Cases were divided into two groups based on implant order (the first 10 or second 10 consecutive implants at a given site). Group differences were assessed using Mann-Whitney U-tests, Chi-squared tests, or Fisher's Exact tests, as appropriate. A Mann-Kendall trend test was used to detect if there was a monotonic trend in operative time. Sleep outcome equivalence between experience groups was assessed using the two one-sided tests approach.

Thirteen facilities met inclusion criteria, contributing 260 patients. Complication rates did not significantly differ between groups (P = .808). Operative time exhibited a significant downward trend (P < .001), with the median operative time dropping from 150 minutes for the first 10 implants to 134 minutes for the subsequent 10 implants. The decrease in AHI from baseline to 12-month follow-up was equivalent between the first and second ten (22.8 vs 21.2 events/hour, respectively, P < .001). Similarly, the first and second ten groups had equivalent ESS decreases at 6 months (2.0 vs 2.0, respectively, P < .001). ESS outcomes remained equivalent for those with data through 12-months.

Across the centers' first 20 implants, an approximately 11% reduction operative time was identified, however, no learning curve effect was seen for 6-month or 12-month AHI or ESS over the first twenty implants. Ongoing monitoring through the ADHERE registry will help measure the impact of evolving provider and patient specific characteristics as the number of implant centers increases.