Objective: To develop a simple, accurate algorithm for predicting DBS IPG battery life in patients with PD, ET and dystonia.

Background: Battery life for DBS IPGs is difficult to monitor or predict, putting patients at risk and undue burden on caretakers and clinicians. Current DBS battery life calculators are complicated, utilize parameters not frequently recorded by the clinician such as impedance, and not accurate in our hands. I hypothesized that using within-subject comparisons of battery lives would yield more accurate predictions using only DBS program settings typically recorded in the clinic. How well these predictions hold up for each movement disorder over time may also lend insight into differences in long term changes among disorder types.

Method: Analysis included 83 patients (Kaiser Permanente Northern California) implanted with a Medtronic activa DBS system and 2-4 IPG changes. Patients requiring hardware repairs or that had their IPG replaced prematurely were excluded. DBS settings recorded at each clinic visit were averaged across visits weighted by time. For each depleted battery, area-under-curve (AUC) was calculated by the product: (Amp)(PW)(Hz)(#active contacts). Impedance and #poles were ignored, as well as whether the stimulus was constant current (>90% were constant V). AUC was multiplied by (e^battery voltage) to correct for battery voltage at time of change. This product was compared as a ratio among all depleted batteries in a given patient. Predicted battery life was calculated based on these ratios and compared to actual battery life.

Results: Battery predictions had high accuracy for PD and DYT groups. The mean predicted vs actual difference in battery life (in months) was -0.07+/-1.9 for PD patients (n=62) and 0.8+/-1.9 for dystonia patients (n=10). Predictions in PD patients remained accurate across 4 battery changes spanning >8 years (n=9). In contrast, battery predictions for ET patients were inaccurate and highly variable, with a mean difference of 8.7+-18.7 (n=11).

Conclusion: Within-patient comparison of DBS battery lives yields accurate predictions for PD and DYT patients using basic parameters most noted by clinicians, but with limitation of requiring at least one prior depleted battery as a baseline. Predictability remains stable for nearly a decade. Battery life for ET was unpredictable suggesting differences in long-term changes in physiology or patient interaction with the system, compared to PD.

« Back to 2022 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/an-accurate-yet-practical-method-for-dbs-battery-life-prediction/