ABSTRACT
STUDY OBJECTIVES: Hypoglossal nerve stimulation (HGNS) is a common treatment for obstructive sleep apnea (OSA). Objective assessment of HGNS efficacy measures apnea-hypopnea index (AHI) by multi-amplitude titration polysomnography (tPSG) and/or a single amplitude efficacy sleep study (eHST). Both tests have been used to determine efficacy despite significantly different protocols. This project's aim was to determine differences in objective outcomes in HGNS patients who underwent both tPSG and eHST post-operatively. METHODS: Data from 379 consecutive HGNS patients were retrospectively reviewed. Inclusion requirements were a pre-operative sleep study, a post-operative tPSG, and then an eHST, which at our institution is a type 3 home study. AHI mean and differences were calculated. Wilcoxon rank sum tests were used to analyze differences between tPSG and eHST. Sher15 criteria (post-operative AHI≤15 events/hour and ≥50% reduction from baseline) was calculated and compared by χ2 tests. RESULTS: Ultimately 61 patients met inclusion criteria with an average pre-operative AHI=33.2. When comparing the subject's tPSG versus eHST, tPSG AHI was significantly lower (AHI=8.8 versus AHI=17.6; respectively, p<0.001). There was also a difference in the percentage of patients that met Sher15 criteria when using tPSG (80.3%) versus eHST AHI (45.9%). CONCLUSIONS: HGNS patient's postoperative tPSG AHI was significantly lower than their eHST outcome. This work highlights the importance of reporting the type of post-operative study used in evaluating HGNS efficacy and the need for single amplitude, full-night studies to assess HGNS efficacy more accurately.
ABSTRACT
BACKGROUND AND OBJECTIVES: Mechanical thrombectomy (MT) is crucial for improving functional outcomes for acute ischemic stroke. Length of stay (LOS) is a reimbursement metric implemented to incentivize value-based care. Our study aims to identify predictors of LOS in patients undergoing MT at a high-volume center in the United States. METHODS: This was a retrospective study of patients who underwent MT at a single institution from 2017 to 2023. Patients who experienced mortality during their course of hospital stay were excluded from this study. Extended LOS (eLOS) was defined as the upper quartile (≥75th) of the median duration of hospital stay. Univariate and multivariate analyses were performed, with P values < .05 denoting statistical significance. RESULTS: Seven hundred three patients met criteria for inclusion. The median age of the cohort was 72 years (IQR: 61-82), and 57.2% was female. The median LOS was 6, IQR: 4-10. A total of 28.9% of the cohort (n = 203) patients experienced eLOS. The multivariate regression model identified age (odds ratio [OR]: 0.98, 95% CI: 0.97-0.99), diabetes mellitus (OR: 1.68, 95% CI: 1.15-2.44), and hemorrhagic transformation of stroke (OR: 2.89, 95% CI: 0.39-0.90) as predictors of eLOS, whereas antiplatelet use before admission (OR: 0.55, 95% CI: 0.34-0.89) and higher baseline modified Rankin Scale before stroke were associated with lower odds (OR: 0.59 [0.39-0.90]; P < .05) of eLOS. CONCLUSION: By identifying predictors of eLOS, we provide a foundation for targeted interventions aimed at optimizing post-thrombectomy care pathways and improving patient outcomes. The implications of our study extend beyond clinical practice, offering insights into healthcare resource utilization, reimbursement strategies, and value-based care initiatives.