The neuromodulation registry survey: A web-based survey to identify and describe characteristics of European medical patient registries for neuromodulation therapies in chronic pain treatment

1 Introduction

Neuromodulation therapies have long been used to treat a variety of medical conditions, including chronic pain. These range from transcutaneous electrical nerve stimulation (TENS) to deep brain stimulation (DBS). Neuromodulation therapies, such as spinal cord stimulation (SCS), DBS, and intrathecal drug delivery systems, require implantation of medical devices. The market for these devices has grown significantly [1].

Several implantable neuromodulation modalities are considered standard care in many countries when conservative treatments for certain chronic pain conditions fail. Their efficacy is supported by systematic reviews [2,3]. However, recently, a controversy emerged from two Cochrane reviews [4,5] and a randomized controlled trial (RCT) [6]. They suggest SCS may not be more effective than placebo for pain intensity and physical function. The publications faced criticism for methodological flaws [7,8], prompting calls for more rigorous methodology [9] and medical patient registries [10,11].

While blind RCTs are the gold standard for evaluating treatment efficacy [12,13,14], there are challenges in the context of neuromodulation because best practices require individualized device programing, which inevitably expose study participants to treatment allocation. Additionally, RCTs often focus on short-term efficacy, potentially overlooking patient safety and long-term effectiveness [15]. Often RCTs have strict inclusion criteria that can lead to study populations not reflecting real-world patients, high costs, and logistical challenges. Registries can address some of these limitations by providing long-term real-world data [16,17] and also enabling the supervision of treatment trends and comparison with other clinics.

Patient registries are data systems that prospectively collect and use observational and clinical data to assess specific outcomes for specific populations, serving scientific, clinical, or policy purposes [18]. They are essential for monitoring safety and evaluating risks [15]. The European Union Medical Device Regulation [19] mandates the examination of adverse events and clinical investigations of implanted devices and places surveillance responsibility on manufacturers [20]. Monitoring adherence to guidelines is important in various medical fields [13]. For instance, such monitoring has successfully improved stroke patient treatment [21].

There is published information about the characteristics of neuromodulation registries for chronic pain, which have been described in detail by Meier et al. [22] and Jaeger et al. [23], as well as within the framework of other published research [15,24,25,26,27,28]. However, despite these efforts, there remains a gap in the knowledge of the prevalence and characteristics of such registries. This article aims to fill this gap by presenting the first overview of such registries, thereby contributing to the body of knowledge in this area.

2 Methods

This study is a part of a broader study, named the Eurosonar survey (ES), which investigates neuromodulation practices for treating chronic pain. Specifically, the segment presented here, termed the Neuromodulation Registry Survey (tNRS), focuses on registries monitoring neuromodulation implants used to treat chronic pain. The other findings of the broader ES will be published separately.

In order to verify the novelty of our survey, we conducted a search on PubMed and Embase using the keywords: ((Survey[title]) AND (Neuromodulation)) AND (Registr*[Title/Abstract]). This search did not yield any publications that are similar to ours, confirming that our survey is the first of its kind to provide such an overview.

We conducted the study as a web-based survey using the “Nettskjema” online tool (University of Oslo, Norway) to design the survey and to harvest data. A team of five professionals, including three neuromodulation device implanters (two of whom are active implanters) and three with experience in building registries in the pain or neuromodulation fields, designed the survey. Four team members were medical doctors, and one was a nurse. The survey was piloted by two clinicians to enhance its clarity and relevance.

We recruited respondents from June to September 2023 by distributing invitation emails. After an initial attempt, we found it necessary to change our approach. The first attempt was to identify responders in the Embase database, where emails of corresponding authors are available for extraction. We narrowed an initial list of 7,000 emails to focus on European, human-related pain research involving device implants, excluding device manufacturer employees. We also sourced participants through our network and public online information. This selection process led us to send invitations to 213 email addresses through the Mailchimp newsletter service (Atlanta, Georgia, USA). However, the response rate was minimal, with only four individuals engaging in our initial correspondence and subsequent reminders.

To increase participation, we manually verified invitation recipients’ involvement in neuromodulation for pain management through Google searches. This revealed that many initial recipients were ineligible. After this vetting process, we directly contacted 136 individuals via personalized emails. These mails were sent to recipients in the following 19 countries: Austria, Belgium, Bosnia and Herzegovina, Denmark, Finland, France, Germany, Ireland, Italy, the Netherlands, Norway, Poland, Portugal, Russia, Spain, Sweden, Switzerland, Turkey, and the United Kingdom. We requested only one response per center and manually checked for several answers from each center. Thus, we reached out to 101 distinct centers across Europe. One reminder email was sent to non-responders after 1 month (Figure 1).

Figure 1

Flow diagram of number of centers (n) responding to the survey. The Eurosonar survey, a broader survey, consisted of two parts: (1) a section on clinical practices in neuromodulation for chronic pain treatment (not presented here) and (2) a section detailing Neuromodulation Registries.

To accommodate varied respondent experiences and to avoid forcing answers, we designed the survey with dynamic response options and chose not to include mandatory questions. Consequently, the number of responses varies per question.

The data were analyzed using descriptive methods in IBM SPSS Statistics (version 27) and Microsoft Excel 2016. To maintain confidentiality, we have chosen not to disclose which centers provided which specific responses. Missing data have not been imputed.

3 Results

From the 101 centers across 19 European countries, we received responses from 47 centers on the broader ES, yielding a response rate of 47%. These responses came from centers in 17 out of the 19 European countries invited (Figure 2).

Figure 2

Overview of neuromodulation registries identified in the survey. Gray countries refer to countries from which no responses were received. See legend for the color designation. The map was created with Mapchart.net. An updated map can be viewed at: https://www.oslo-universitetssykehus.no/fag-og-forskning/nasjonale-og-regionale-tjenester/regional-kompetansetjeneste-for-smerte-reks/neuromodulation-registries---updated-map/.

Of these 47 centers, 27 reported using a registry (although four did not provide details on their registry). This resulted in 23 responses for the tNRS from 11 countries (Table 1).

We identified the existence of three national registries (Belgium, the Netherlands, and the United Kingdom), while 13 centers across seven countries reported using non-national registries. These were single-center, multi-center, and regional registries. Centers in Denmark, France, and Sweden used nationally available software solutions (without having the status of a national registry, to our knowledge). Table 2 shows characteristics of the responding centers.

Centers in Belgium, the Netherlands, Norway, and the UK indicated that registration in the registry is mandatory. Responses from Denmark and the UK showed both mandatory and non-mandatory registrations, thus showing a discrepancy in the respondents. Respondents from centers in Finland, France, Sweden, and Switzerland stated that registration is non-mandatory. Belgium and the Netherlands only provide reimbursement for implant procedures that are registered in the registry. In Denmark, Norway, and the UK, the departments’ leadership mandates the utilization of the identified registries.

Of the 13 different registry solutions identified, 11 are funded from public sources (or no funding). Two respondents, one from Sweden and one from Spain, answer that the registry is funded by device manufacturer(s). Several respondents (five) report that the registry is funded by a combination of public sources, manufacturers, and insurance (Denmark, France, the Netherlands, and the UK).

3.6 Information collected in the registries

The survey included questions on what type of information the registries collect. We divided the information into the following domains: (a) organizational matter, (b) patient characteristics, (c) procedural information, (d) post-implantation treatment trajectory, (e) device programing information, (f) adverse events and complications, and (g) patient-reported information. The responses to these questions are detailed in Table 4 for all registries. We received conflicting responses from different centers within the same registry. To resolve this, we created an index to merge the data. See the legend of Table 4 for a detailed explanation.

Table 4

Information collected by registries surveyed

3.6.1 Organization

We asked if there is dedicated staff to the running of the registry (Table 4). Respondents from the national registries in Belgium and the Netherlands report differently, but most of these respondents informed that there is dedicated staff. Registries in Finland, France, and Switzerland report having dedicated staff for the registry. Meanwhile, registries in Denmark, Norway, Spain, and the UK reported having staff partly dedicated to running the registry.

Steering committees are less common in non-national registries than in the national registries. The registries in Belgium, France, the Netherlands, and the United Kingdom, as well as one registry in Spain, report having a steering committee. The Danish registries report that partly there is a steering committee for the registries, while the registries identified in Finland, Norway, Sweden, and Switzerland report no to the same question.

3.6.2 Data registered in the registry

Most registries record numerous preimplantation variables, including the duration of disease, pain diagnosis, etiology, and daily opioid consumption (Table 4). Eight out of 21 (38%) respond that information on all the patients referred for neuromodulation implants is included in the registry. All registries collect information about the procedures for the trial and permanent implantation, and that is also true for adverse events and complications. On the other hand, information on psychosocial factors that are not collected by Patient-Reported Outcome Measures (PROMs) is collected only by few.

Questions surveying types of patient-registered information are fairly consistent between the registries. Several responders reported using other categories of PROMs. Four out of six responders reported using PROMIS-29, which measures different health domains. For one center, this will be the new alternative to using separate questionnaires for each domain.

Eighteen centers answered the question on time points for when the registries gather PROMs – 100% collect PROMs at the baseline and at 12 months, while 78% collect PROMs at 6 months, and 78% collected them yearly (Table S1, Supplementary materials).

4 Discussion

This study identified 27 centers using registries to monitor neuromodulation implantations for treating chronic pain, primarily from northwestern Europe. Several centers operate regional or local registries, while three registries are national. We also explored perceived challenges of running a registry and the type of data gathered. Due to the low response rate and potential bias toward respondents with an affinity for registries, these findings should be interpreted as a compilation of valuable insights rather than generalizable conclusions for registry work in the neuromodulation field.

5 Limitations

This survey has several limitations. Although the response rate was 47%, consistent with web-based surveys [40], the low absolute number of respondents may lead to potential information gaps. Furthermore, different information on national registries required constructing an index to illustrate the findings, which might not accurately reflect reality. The results from the study may have been skewed from the fact that most respondents were clinicians, leaving out other stakeholders, and that most respondents probably had a special interest in registries. Additionally, the study could have been designed to capture more aspects of data quality. Despite these limitations, we hope this survey offers valuable insights for designing a neuromodulation registry.

6 Conclusions

Our study identified 27 centers utilizing registries, with 23 providing detailed data which we present. Most registries are local, yet many centers in the survey provide data to the three identified national registries. The registries primarily investigate the clinical outcomes and patient-reported outcomes, in line with recommendations. However, more variability exists in collection of other clinical variables essential for benchmarking across centers and for evaluating adherence to clinical guidelines. Data quality remains a significant issue, necessitating further examination beyond this survey to ensure robust registry operations in neuromodulation. We found notable challenges in regard to establishing and maintaining these registries – financial, personnel resources, and practicalities. Many neuromodulation registries appear less developed in comparison to the more established arthroplasty and cancer registries. Addressing these challenges requires cooperative efforts to build structured registries and develop Common COS for registry use. Properly designed registries hold the potential to advance the neuromodulation discipline, influencing policies, improving patient care, and advancing neuromodulation practices.