Parkinson disease (PD) is a multisystem neurodegenerative disorder characterized by motor and nonmotor symptoms [1]. Due to its progressive nature, the burden on patients and their caregivers increases over time. Recent studies have demonstrated the effectiveness of online sessions including remotely supervised exercise and telerehabilitation for patients with PD [2-6]. The advancement and prevalence of remote technologies, in part owing to the COVID-19 pandemic, have promoted remote programming and surveillance in patients who have undergone deep brain stimulation (DBS) [7-18]. Several studies have shown that online sessions are highly effective, particularly in reducing the cost and time burden for patients with PD who underwent DBS, as well as their caregivers [8-11,15-18].

In patients with de novo DBS, 2 randomized studies and several observational studies have shown that remote programming is safe, effective, and comparable with in-clinic adjustments, with high satisfaction rates [7,11,12,15,16,18]. These findings suggest that online sessions can support early postoperative management and reduce the need for frequent hospital visits. In contrast, evidence in patients who had chronically implanted DBS devices is more limited. Retrospective and cross-sectional studies indicate that remote programming may reduce travel distance, time, and financial burden, and that patient satisfaction is generally favorable [8,10,13,14,17]. However, these studies are heterogeneous, mostly observational, and lack prospective evaluation.

The clinical needs of patients who had undergone chronic DBS differ from those of de novo cases, as long-term follow-up typically involves clinical review, medication management, and device checks rather than frequent parameter adjustments. Thus, despite accumulating evidence, prospective, population-specific data on the safety and feasibility of online sessions in chronic DBS care remain insufficient. To date, no prospective study has evaluated online sessions in patients with PD who underwent chronic implantation of DBS devices, and detailed data are still lacking. Given these gaps, we aimed to explore whether online sessions could be implemented safely and feasibly in patients with chronic DBS devices, while assessing their preliminary efficacy in maintaining clinical stability and reducing patient burden—particularly travel-related demands—without compromising routine clinical management.

A prospective, single-arm, longitudinal, multicenter, nonblinded, nonrandomized, and self-controlled cohort trial was conducted at Tokushima University Hospital and Kurashiki Heisei Hospital in Japan between 2023 and 2024. All participants had been treated with in-clinic sessions, followed by 2 consecutive online sessions (Figure 1).

This study was approved by the Ethics Committee of Tokushima University (approval number 4409, date of approval August 28, 2023) and the Ethics Committee of Kurashiki Heisei Hospital (approval number R05-008, date of approval July 4, 2023) in accordance with the Declaration of Helsinki. The participants received detailed study information and signed informed consent forms before enrollment. To protect participant privacy, all data were deidentified, stored on secure institutional servers, and accessible only to authorized study personnel. No audio or video recordings were collected during online sessions, and all telehealth communications were conducted through encrypted, institution‑approved platforms. Participants did not receive financial compensation for study participation.

All participants met the following inclusion criteria: (1) adult patients (aged 18 years and older) diagnosed with PD, (2) those who had received bilateral subthalamic nucleus (STN) DBS using the Infinity DBS system with the NeuroSphere Virtual Clinic remote care (Abbott), (3) those who had a follow-up period of at least 3 years postsurgery, (4) those who had never experienced online sessions, and (5) those who are able to operate the online system with or without caregivers and consented to be included in this study. For the sample size calculation, we estimated that 15 participants would provide 80% power (with a 5% probability of type I error) and a large effect size using G*Power 3.1. Therefore, we set the target sample size to at least 17 participants and estimated a dropout rate of 10%.

All measured data are presented as mean (SD). The 2-tailed t test (matched pairs) was used to analyze the 2 groups when the normality of distribution was verified by the Shapiro-Wilk test. If normality was not confirmed, the Wilcoxon rank-sum test was used. Significance and effect size were determined using Hedges g when normality was observed, or Cliff δ when it was not. The Friedman test was applied to compare repeated measures of MDS-UPDRS-III across the 3 assessment points and to evaluate seasonal variation in exposure days across 4 seasons (spring, summer, autumn, and winter).

For any MPIC-J items that showed statistically significant pre-post differences, additional exploratory analyses were conducted to evaluate the clinical relevance of the observed changes. Specifically, the minimally important change (MIC) was estimated using the MPIC-J 7-point global change scale as an anchor; receiver operating characteristic (ROC) analysis was performed to identify the optimal cutoff discriminating subjective worsening, and distribution-based estimates (0.5 SD) were also calculated. Correlation analyses were performed to explore associations between changes in clinical outcomes and candidate explanatory variables. Point-biserial correlation coefficients (r) were used for sex, which was treated as a dichotomous variable. For all other continuous or ordinal variables, Spearman rank correlation coefficients (ρ) were calculated. All statistical analyses were performed using R (version 4.2.1; R Core Team) [26], and the significance level was set at P<.05.

Between August 2023 and April 2024, 21 patients were screened for eligibility from 2 hospitals. Three patients were excluded from the study for the following reasons: 2 patients had difficulty operating the equipment and no familial support, and 1 patient found it challenging to attend frequent assessment appointments. Eighteen patients were enrolled, of whom 2 discontinued participation because of hospitalization and infectious disease. The dropout rate was 11% (2/18). All data points were complete, and no missing values were identified throughout the dataset (Figure 2).

The demographic profile was as follows: 50% (8/16) were female, average age was 67.1 (SD 8.4) years (95% CI 62.6‐71.5), average Hoehn-Yahr Scale score was 3.7 (SD 1.0; 95% CI 3.1‐4.2), average duration of illness was 15.8 (SD 4.4) years (95% CI 13.4‐18.1), average duration after bilateral STN-DBS was 6.6 (SD 1.9) years (95% CI 5.5‐7.6), and the average distance from the resident to the clinic was 42.8 (SD 49.7) kilometers (95% CI 16.3‐69.3; Table 1). There was no difference in the duration between each assessment visit of in-clinic and online sessions (mean 138.4, SD 64.5 days, 95% CI 104.1‐172.8 vs mean 152.6, SD 57.2 days, 95% CI 122.2‐209.8; P=.53; Hedges g=0.220). Friedman tests revealed no significant seasonal variation in exposure days across the 4 seasons for in-clinic sessions (mean 56.8, SD 33.2 days, 95% CI 39.0‐74.6; χ²₃=6.73; P=.081; Kendall W=0.140) or online sessions (mean 54.7, SD 36.2 days, 95% CI 35.8‐73.6; χ²₃=6.83; P=.078; Kendall W=0.143). No adjustments were made to antiparkinsonian medications or to any sleep-affecting drugs throughout all study periods.

Our study demonstrated the feasibility of online sessions using the NeuroSphere Virtual Clinic remote care system compared with the conventional in-clinic sessions. Motor symptoms did not change in either group, as assessed using the MDS-UPDRS-III. Motor stability in patients under chronic DBS is expected over short intervals, regardless of visit modality; therefore, the absence of deterioration might not be interpreted as evidence of clinical effectiveness. This study provides a prospective evaluation of the effect of online sessions on patients who have already been implanted with and under chronic DBS. To our knowledge, this is the first study to investigate the effects of online sessions using DBS devices in Japanese patients. While the prospective design improves data quality compared with retrospective reports, the single-arm nature of this study precludes definitive causal inferences.

The only adverse event occurred between the last in-clinic visit and the first online session, making it unlikely that the event was related to the online sessions. Overall, online sessions appeared to be feasible and clinically acceptable from a safety perspective for patients with PD who have chronically implanted DBS devices. Although the fall was not clearly attributable to online sessions, the inability to directly assess rigidity, balance, and gait remains an inherent limitation of remote care. As expected, there was a substantial reduction in both cost and time burden for patients and caregivers [8,9,11,15-18].

In this study, approximately half of the patients thought that online sessions were equivalent to in-clinic sessions, as reflected in TUQ item 15 and the OUQP. Around 70% of patients reported satisfaction with the online sessions (TUQ item 21 and OUQP; Tables 3 and 5). Given that TUQ subdomain “Usefulness” refers to “the positive effects on clinical outcomes” or “reduction in clinical cost” [24], our results indicated that 70.8% (34/48) of the patients felt that online sessions had positive effects on clinical outcomes. Previous retrospective studies assessing patients with chronic DBS implantation indicated that 62% of patients preferred online visits [17]. A cross-sectional survey of teleprogramming after each online session for patients with chronic pain equipped with the NeuroSphere Virtual Clinic remote care showed that 93.8% (15/16) of the patients prefer the online session [9]. Similarly, a cross-sectional survey of teleprogramming following online sessions in patients with PD who had undergone de novo DBS implantation found that 100% of patients were satisfied with the online sessions and expressed a willingness to participate in future [11]. PD is progressive, and the burden of in-clinic visits for both patients and caregivers increases as symptoms worsen. In this context, online systems may help reduce that burden without negatively affecting patients’ core motor symptoms, although this requires confirmation in controlled studies.

One unexpected finding was a potential signal of deterioration in sleep status in patients who received online sessions, which has not been previously reported (Table 2), although the observed change in sleep status consistently fell below the established MIC thresholds, indicating limited clinical relevance. Exploratory correlations suggested that patients with preexisting clinical instability—such as motor worsening or fluctuations in pain-related domains—were more likely to show sleep deterioration.

Interestingly, the duration of online sessions showed a strong association with ΔMPIC-J item 3. Because the online period varied among patients, this association is unlikely to reflect intentional differences in exposure to online sessions. Instead, it may imply that patients with greater instability were more susceptible to sleep deterioration when they spent a longer proportion of the follow-up period in online sessions. These exploratory correlations could imply a potential interaction between patient instability and the extent of online follow-up; however, these findings should be interpreted with caution. Taken together, these findings suggest a need for further investigation into whether sleep deterioration during online-based DBS management may not be solely attributable to the online modality itself but rather an emergence in clinically unstable patients who may instead benefit from shorter online follow-up and earlier return to in-person assessment.

According to a randomized controlled trial, online programming sessions for de novo implanted DBS systems resulted in approximately 38% shorter treatment duration compared with in-clinic optimization and increased patient access to therapy adjustments by nearly 1.9 times [18]. Given the strong therapeutic impact of the initial stimulation, there is concern that the added value of online sessions may be overlooked or inadequately assessed. In addition, Chen et al [15] reported the possibility that physicians may adjust medications less frequently during remote programming than during in-hospital programming for de novo DBS-implanted patients with PD. This raises the possibility of bias due to differences in dopamine levels. In our study, we targeted patients with PD with an average duration of 6.6 (SD 1.9, range 4.5‐10, median 5.5) years, representing the longest follow-up period reported in the literature. In contrast to the evidence in patients who have undergone de novo DBS, studies involving chronically implanted DBS systems have been retrospective or cross-sectional (Table 6). These reports suggest that remote programming can substantially reduce travel time and financial burden and is generally well accepted by patients [10,17] but lacked prospective evaluation of safety, usability, or clinical stability in long-term DBS management. Our study partially addresses this gap by providing preliminary prospective evaluation in patients who had chronically implanted DBS devices, offering a baseline for future comparative trials. It also offers the first population-specific data from Japan, using a device platform not previously examined in this context. By focusing on long-term DBS users, we examined the specific population for whom online sessions are presently feasible in real-world Japanese clinical settings. The inclusion of detailed usability measures (TUQ and OUQP) and the observation of a potential signal in sleep-related symptoms further add domain-specific insights not highlighted in earlier reports. Together, these findings suggest the feasibility and acceptable short-term safety profile of integrating online sessions into DBS care. Although time and cost savings are not direct care quality measures, our descriptive estimates—based on a typical frequency of 6 clinic visits per year—suggested that patients and caregivers might save approximately 28 hours and 47,844 JPY (US $1≈155 JPY).

Two patients were excluded from this study because of insufficient caregiver support and lack of digital literacy. Caregiver support is essential for patients with PD at these stages. Patients with PD may experience difficulty using small screens on tablet computers or smartphones due to motor impairments and cognitive dysfunction. These limitations may partly explain the relatively low agreement rates in the “Reliability” and “Interface quality” subdomains of the TUQ. These subdomains assess, “how easily the user can recover from an error and how the system provides guidance to the user in the event of an error” and “how pleasant the telemedicine technology or computer system was to use for the consumer,” respectively (Tables 3 and 4) [24]. It remains difficult for clinicians to establish online consultation systems in hospitals because of limited medical resources. Administrative challenges include difficulties obtaining official prescriptions and a lack of medical insurance coverage [17]. The OUQP showed that it was sometimes difficult to provide rapid treatment resolution (31.3%; Table 5). This was, in part, due to the inability to touch the patients directly (eg, the inability to check the rigidity of the extremities). These challenges should be addressed to improve feasibility. Despite these concerns, it is surprising to see many patients satisfied with online sessions and inclined to use in the future as shown in the TUQ subdomain “Satisfaction and future use” (Table 4).

This study has some limitations. First, this was a single-arm longitudinal study without a comparator group, which limits causal inference and the ability to contextualize the observed changes. Long-term patients with PD equipped with DBS devices with no experience of online sessions are rare, making recruitment challenging; therefore, a single-arm design was the most feasible approach for this population. Second, the small sample size may have reduced the statistical power. Although it was determined a priori using conservative assumptions, the study may still have been underpowered to detect smaller effects. Third, the observation period for the online sessions was short and longer online sessions may yield different effects. Fourth, the study population was necessarily selective, as patients without caregiver support or with low digital literacy were unable to participate. This introduces a potential selection bias and limits the generalizability of our findings to more vulnerable groups. At the same time, patients who can realistically engage in remote DBS care in routine clinical practice are typically those with caregiver support and sufficient digital literacy; therefore, the present cohort also reflects the subset of patients who had undergone chronic DBS for whom online sessions are currently feasible. Finally, the findings apply specifically to patients with bilateral STN-DBS using the Infinity/NeuroSphere remote care system within a Japanese health care setting with adequate infrastructure, which further limits the applicability of our findings to other devices, stimulation targets, or health care systems. The absence of randomization, blinding, or parallel controls also raises the possibility of observer, performance, detection, and expectation bias. Thus, we must evaluate our results carefully because of these limitations.

Sharma et al [27] described ideal candidates for online sessions as those who have a supportive caregiver, do not have cognitive or psychiatric impairment, or have impediments to in-person visits. Given that establishing a relationship with the DBS care team is essential for patients who had undergone de novo DBS [27] and that the postoperative levodopa equivalent daily dose reduction rate may be lower in patients managed online [15], online sessions might be most suitable for patients with chronically implanted DBS devices. The absence of statistically or clinically meaningful worsening in all aspects—with only a modest and clinically limited change in sleep status—in patients with PD and chronic DBS implantations supports the feasibility of incorporating online sessions, or a hybrid of in-clinic and online sessions, into DBS care. Such approaches could be considered as a potential means to alleviate the burden on patients and caregivers. Future studies incorporating validated sleep-specific instruments, such as the Pittsburgh Sleep Quality Index, objective measures including actigraphy, and longer follow-up periods, will be essential to determine whether the sleep changes identified here represent transient fluctuations or clinically relevant deterioration. Despite some limitations, the online sessions showed potential feasibility and an acceptable safety profile in this preliminary evaluation, suggesting their potential role in reducing the time and cost burden for patients.