Practice variability in vagus nerve stimulation settings for drug-resistant epilepsy and a pilot quality improvement...

The aim of this study was to access vagus nerve stimulation (VNS) parameters for the management of drug-resistant epilepsy (DRE) and to implement a pilot quality improvement initiative to ensure that uniform and consistent literature-supported parameters are used for optimal seizure control in patients with DRE.

A single-institution retrospective analysis was performed to assess VNS parameters of patients with DRE who underwent implantation from 2013 to 2022, before the intervention was introduced (group 1). Patients who underwent implantation from 2023 to 2024 comprised the postintervention group (group 2), with intervention that included activating VNS intraoperatively and titration to a therapeutic dose of 1.5 mA within 12 weeks. Parameters between groups were compared. Statistical control charts were used to assess process stability. An interrupted time series (ITS) analysis was conducted to assess the impact of the intervention over time, adjusting for preintervention trends.

Eighty-six patients (50% female, mean age 25.4 [SD 15.4] years) were included. Group 1 (n = 62) had delays in initiation of VNS therapy following implantation (median 15 [IQR 10.75] days) and achieving the therapeutic threshold (median 57.5 [IQR 72.93] weeks). Group 2 (n = 24) showed no delay to initiation (median 0 [IQR 0] days) and reached the therapeutic threshold early (median 10 [IQR 6] weeks) compared with group 1 (p < 0.001). Patients for whom the therapeutic threshold was reached during the clinical course increased from 80.65% in group 1 to 91.67% in group 2. Among them, the proportion reaching that threshold within 12 weeks of implantation increased significantly from 14% (7/50) in group 1 to 68.18% (15/22) in group 2 (p < 0.001). Individual and moving range charts, assessed using Levene's test for variance, showed greater process stability and consistency in group 2. ITS analysis showed that the intervention had a sustained and significant impact; the proportion of patients with timely therapeutic parameters increased by 13.26 percentage points every 3 months (p = 0.029). Logistic regression analyses showed the postintervention group and use of scheduled programming had greater odds of achieving timely therapeutic parameters (OR 6.58 [p = 0.001] and 56.77 [p < 0.001], respectively).

Substantial practice variability in VNS parameter settings with suboptimal utilization of therapy existed, which improved following introduction of the pilot intervention. Standardized protocols can ensure timely delivery of therapeutic parameters. These strategies have been underutilized in earlier studies, likely overestimating the time to response from VNS therapy.