4.1.4. Side Effects

Thirty-seven of the 87 included studies reported side effects (N = 785). Namely, 7.89% of patients complained of paresthesia (N = 62), 5.47% of gait and balance problems (N = 43), 2.03% of ataxia (N = 16), 15.28% of dysarthria (N = 120) and 0.12% of diplopia (N = 1). Only voltage exhibited a significant correlation, with dysarthria r(118) = 0.43, r2 = 0.43 (18.49%), *p* = 0.01.

### *4.2. Experimental DBS Programming*

### 4.2.1. Experimental Tremor Titration

The experiment was conducted in eight ET patients (Table 1, Supplementary Figure S2). None of the patients was on any tremor medication following DBS surgery. Baseline tremor control could not be documented in one patient (ET6) and was excluded from the corresponding statistical analysis. Given that subjective improvement was nevertheless achieved and sustained in the thr medium term, ET6 is still documented in Table 1. The experimental paradigm showed significant tremor reduction compared to baseline stimulation, according to both subjective (*t(6)* = −2.95, *p* = 0.02) and objective (*t(6)* = −3.07, *p* = 0.02) measurements (Figure 2). Notably, the tremor improvement perceived by the patients corresponded to the accelerometer measurements. On average, the results were achieved after 23.8 ± 8.03 min per patient. Upon follow-up, 45% of the patients reported medium-term tremor control superior to baseline (*p* = 0.01). None of the patients reported the new setting as being worse than the previous. As far as side effects are concerned, four patients (ET2, ET4, ET7, ET8) reported stimulation-induced ataxia and/or dysarthria at baseline. Following experimental re-titration, side-effects were resolved in all but one patient (ET4), who reported re-emergence of ataxia upon follow-up.


Abbreviations: ET = essential tremor, F = frequency (Hz), TDAmp = arbitrary unit of measurement representing tremor severity, P = pulse width (μs), V = voltage (V), Vim = ventral intermediate thalamic nucleus, ZI = zona incerta.

### *J. Clin. Med.* **2020**, *9*, 1855

**Table 1.** Patient demographics and deep brain stimulation (DBS) parameters. Baseline tremor control was measured prior re-titration with the patients retaining their current DBS settings. Next, ten random combinations of stimulation parameters were tested in eight essential tremor (ET) patients. The random combinations

**Figure 2.** Experimental ET-DBS titration. Ten random combinations of stimulation parameters were tested and characterized subjectively (patient reported outcomes) and objectively (accelerometer measurements). The tremor outcomes during the experimental parameter exploration are categorized as follows: worst (least tremor reduction), off (stimulation turned off), baseline (initial settings), best subjective (best random settings according to the patient), best objective (best random settings according to accelerometer measurements). The novel programming strategy afforded significant tremor reduction (*t(6)* = −2.95, *p* = 0.02) in the absence of side-effects. (\*) represents a statistically significant (*p* < 0.05) difference.

### 4.2.2. Random DBS Parameters

An example of the randomly generated parameters is given in Supplementary Figure S2. Parameter optimization was achieved with significantly broader pulse widths according to both subjective *Z* = −2.37, *p* = 0.01 and objective *Z* = −2.52, *p* = 0.01 measurements. Conversely, significantly lower frequencies proved superior to baseline settings in both subjective *Z* = −2.53, *p* = 0.01 and objective ratings *Z* = −2.52, *p* = 0.01. No significant difference was detected for either subjective *Z* = −1.12, *p* = 0.26 or objective *Z* = −1.54, *p* = 0.12 ratings of voltage. More optimal tremor reduction was achieved with higher levels of TEED, in both subjective *t(7)* = −2.12, *p* = 0.07 and objective *t(7)* = −3.08, *p* = 0.01 measurements.
