Background and objectives With a growing need for non-opioid chronic pain treatments, pain physicians should understand the proper utilization of neuromodulation therapies to provide the most comprehensive care. We aimed to identify the unmet training needs that deter physicians from using spinal cord stimulation (SCS) devices.
Methods Internet-based surveys were fielded to fellows enrolled in pain fellowships during the 2016–2017 academic year accredited by the Accreditation Council for Graduate Medical Education and past pain fellows identified through pain medicine societies and SCS manufacturers.
Results Current fellows were more likely to have received SCS training during fellowship compared with past fellows (100.0% vs 84.0%), yet there was variability in fellows’ SCS experiences with a wide range of trials and implants performed. Forty-six percent of current fellows felt there was an unmet training need regarding SCS. Deficiency in SCS case volume was the most common barrier that was noted (38.5%), followed by lack of SCS curriculum (30.8%) and lack of faculty with SCS expertise (23.1%). Lack of training was a predominant reason for past fellows choosing not to use SCS devices postfellowship. The majority of current and past fellows (79.5% and 55.4%, respectively) strongly supported direct training of fellows by SCS manufacturers.
Conclusions While SCS training during pain fellowship has become more universal, the experiences that fellows receive are highly variable, and most rely on industry-sponsored programs to supplement training deficiencies. Standardization of SCS procedures may also enable less experienced providers to navigate the SCS treatment algorithm.
- spinal cord stimulation
- pain fellowship
- training curriculum
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Spinal cord stimulation (SCS) is a cost-effective treatment for chronic pain, including pain from failed back surgery syndrome, neuropathic pain syndromes, peripheral vascular disease and angina pectoris.1 2 More than 14 000 neurostimulators are implanted around the world every year, and technological advancements as well as increasing clinical applications have led to widespread use of these devices.3 4 SCS training became a suggested component of pain fellowships accredited by the Accreditation Council for Graduate Medical Education (ACGME) in 2006, with fellows gaining experience through clinical training and cadaver-based courses sponsored by industry manufacturers and pain societies. While the Accreditation Council for Graduate Medical Education (ACGME) Pain Fellowship guidelines (five neuromodulation cases) are not required, some programs do not provide significant or possibly no exposure to neuromodulation to fellows.5
New graduates from pain fellowships will be asked to care for complicated patients with chronic pain, and with the current opioid epidemic, neuromodulation therapies will play a larger role in the management of pain syndromes. While inadequate exposure and lack of training uniformity at academic pain fellowships have previously been identified as reasons for why practitioners decline using neuromodulation devices,6 7 there have been no studies specifically examining current SCS training perceptions and deficiencies.
Therefore, we aimed to assess the unmet training needs that deter physicians from using these procedures through the use of surveys sent to current and former pain fellows asking about their fellowship training experiences and current SCS practices. Correctable deterrent factors may be used by fellowship program directors to revise their training curriculum to ensure that pain physicians are able to meet the rising demand for SCS devices.
After Institutional Review Board (Weill Cornell Medical College, protocol #1507016431) approval was obtained, three web-based surveys hosted by SurveyMonkey Inc (www.surveymonkey.com, San Mateo, California, USA) were constructed. Survey A was a 27-item questionnaire (online supplementary appendix 1) sent via email with a survey hyperlink to program directors for all 100 multidisciplinary ACGME-accredited pain management fellowships to distribute to their fellows (estimated total fellows: 286)8 at the start of the 2016–2017 academic year. This survey was fielded from 10 August 2016 to 31 October 2016. The survey questions addressed current fellows’ SCS experiences prior to fellowship. Fellows were asked to provide their email addresses at the end of the survey for the purposes of recontacting for a follow-up survey (survey B).
Survey B (online supplementary appendix 2) was then sent to all respondents of survey A via email towards the end of the 2016–2017 fellowship year (months 11 and 12 of fellowship). Survey B was a 36-item questionnaire follow-up asking about fellows’ SCS training experiences during fellowship, their future practice goals and thoughts on training by industry manufacturers. This survey was fielded from 7 June 2017 to 6 July 2017.
Survey C was a 55-item questionnaire (online supplementary appendix 3) sent to pain physicians in the USA identified through pain medicine societies, including the American Society of Regional Anesthesia (ASRA) and the International Neuromodulation Society (INS), as well as SCS industry manufacturers, including Boston Scientific, Nevro and Medtronic. The aforementioned groups and companies reached out to their respective email lists. This survey was fielded from 24 August 2016 to 30 February 2017.
The questions for all three surveys were divided into three sections: (1) training; (2) future/current practice; and (3) industry and standardization. All survey responses were deidentified, and respondents were kept anonymous from the study investigators. Only the study investigators had access to the survey results. Because every question was not completed for each survey, data analysis was based on the non-missing responses to each individual survey question. Distribution of the survey including location, parties involved and number of attempts is listed in online supplementary appendix 3. No financial incentives were provided for completion of the survey as per decision of the investigators to limit bias.
In all surveys, comfort with various procedures was evaluated on a scale ranging from 1 to 10, where 1 indicated ‘very uncomfortable’, 4 indicated ‘somewhat uncomfortable’, 7 indicated ‘somewhat comfortable’ and 10 indicated ‘very comfortable’.
Our study aimed to collect survey data from the entire population of interest (ie, all current and former pain fellows); therefore, we did not conduct an a priori power analysis to identify an appropriate sample size. Descriptive statistics were used to characterize the survey responses. The data were analyzed descriptively and are expressed as percentages of the total number of returned valid responses for each question. Fisher’s exact tests and Mann-Whitney U tests were used, as appropriate. As undercoverage could have been present (but was not quantifiable), in addition to unit non-response (with no data on non-respondents), and in light of the fact that the population surveyed encompassed all individuals within the sample frame (thus becoming a census), we did not weight any analyses nor construct CIs around point estimates. All descriptive statistics, tables and charts were calculated and constructed in Excel (Microsoft, Redmond, Washington, USA), and all tests were two sided and assumed a 0.05 significance level. All analyses were conducted in Stata SE, V.15.
A percentage of 26.6 of all pain fellows in ACGME accredited fellowships responded to survey A and 52.6% of those who responded to survey A also responded to survey B. This resulted in an overall response rate of 14.0%. Given that we were unable to target all practicing pain physicians in the USA, it is difficult to determine the response rate for survey C.
Survey A: current fellows, start of fellowship
Of the respondents, 77.6% previously completed residencies in anesthesiology, 18.4% in physical medicine and rehabilitation (PMR), 2.6% in neurology and 1.4% in neurosurgery (table 1). Only 19.7% had SCS training prior to the start of their fellowship.
Of those with previous SCS experience, 61.5% had directly participated (as first-assist) in two or fewer trial or implant procedures. Thirty-nine percent of fellows with prior SCS experience reported feeling ‘somewhat’ to ‘very’ comfortable performing trials independently at the start of fellowship, but only 31% of the same population felt comfortable performing implants independently.
Seventy-nine percent of current fellows—regardless of SCS experience—indicated that SCS training had been a factor contributing to their choice of fellowship program, with 8.3% reporting that they chose their fellowship solely because it emphasizes SCS training, and another 70.8% reporting that SCS training was one of many factors contributing to their fellowship program choice. Twenty-one percent stated that SCS training had no bearing on their decision.
Future practice/industry and standardization
Sixty-seven percent anticipated applying to postfellowship positions where they would be performing SCS procedures. Sixty percent planned on performing SCS trials and implantations, 27.8% intended to perform trials but not implants, 11.1% were unsure of whether they would use SCS and 1.2% did not intend on taking care of SCS patients. Ninety percent of all current fellows responding to the survey at least ‘somewhat’ supported the direct training of fellows by SCS manufacturers: sixty-seven percent strongly supported this direct training, 23.6% somewhat supported, 8.3% were neutral/had no opinion, 1.4% were somewhat against and none strongly against.
Survey B: current fellows, graduating fellowship
Of the 76 fellows who completed survey A at the start of their fellowship, 40 (52.6%) responded to survey B, which was sent at the end of their training with an additional reminder email. Of those respondents, 72.5% previously completed residencies in anesthesiology, 25.0% in PMR and 2.5% in neurology (table 2). All reported receiving SCS training during fellowship and 77.5% participated in industry-sponsored workshops, with 50% having attended three or more. Fellows participated in SCS courses sponsored by four of the major manufacturers that dominated the SCS marketplace at the time of the survey, with 52.5% of respondents having attended Medtronic courses, 77.5% Boston Scientific, 47.5% St. Jude/Abbott and 30.0% Nevro. Twenty percent of fellows also attended non-industry courses (ie, continuing medical education courses).
The number of SCS trial and implant procedures that fellows directly participated in varied, with 5.0% having performed 1–5 procedures (including trials or implants), 17.5% with 6–10, 22.5% with 11–15 and 55% with 15 or more. Fellows also had more experience with SCS trials compared with implants, with 47.5% of respondents having done 15 or more trials compared with 22.5% for implants. Ninety-five percent felt ‘somewhat’ to ‘very’ comfortable performing SCS trials independently at the end of their fellowship compared with 57.5% for implants. With regards to other surgical implant skills, most fellows felt at least ‘somewhat’ to ‘very’ comfortable with subcutaneous pocket formation for IPG placement, surgical wound closures and infection prevention techniques (all greater than 65.0%). Forty-six percent indicated that there was an unmet need for SCS training (figure 1A), including insufficiencies related to SCS curriculum, faculty with SCS expertise and number of institutional SCS cases (figure 1B).
Most graduating fellows pursued private practice or academic jobs (61.5% and 25.6%, respectively), with 63.2% indicating that they would be taking a full-time pain position (table 3). Ninety-five percent intended on performing SCS procedures, with 66.7% planning to perform both SCS trials and implantations and 28.2% performing only trials. Of the 5.2% who indicated they were unsure or did not plan to manage SCS patients, they stated questionable long-term success rate, maintenance requirements, lack of adequate training and lack of interest as factors. The vast majority (92.3%) felt they received support from their fellowship program for building their SCS practice by being introduced to industry representatives, but fewer fellows had learned how to obtain future support from their own department (69.2%) as well as other departments (ie, surgeons and referring providers).
The number of trials and implants completed during fellowship were not related to the graduates’ future plans to incorporate SCS in their practice (p=0.08 and p=0.06, respectively). Moreover, fellows’ comfort with performing these procedures independently was not a determining factor in their future plans (implants: p=0.14; trials: =0.64).
Industry and standardization
Most fellows (84.6%) believed that SCS procedures (clinic practices and technique) were at least ‘somewhat’ or ‘very’ standardized at their training institutions versus ‘not very’ (10.3%) and ‘not at all’ (5.1%). Eighty-five percent indicated that SCS procedures should be standardized by either pain medicine societies (35.9%), ACGME fellowships directors (25.6%) or through industry and fellowship partnerships (23.1%). The majority (92.3%) also supported the direct training of fellows by SCS manufacturers (figure 2).
Survey C: past fellows (practicing physicians)
We received 163 responses from practicing pain physicians who completed their pain training from 1978 to 2016, with most respondents having completed their fellowships after 2000 (79.8%). Eighty-two percent completed anesthesiology residencies, 10.4% in PMR and 3.7% in neurosurgery. Sixty-three percent of respondents stated that SCS training was a factor for their fellowship choice, while 37.4% said that it had no bearing on their decision. Eighty-four percent received SCS training during fellowship, with 66.3% having attended industry-sponsored courses.
Similar to the graduating fellows, the number of SCS trials and implants completed by past fellows during fellowship varied widely (table 2). More respondents felt ‘somewhat’ to ‘very’ comfortable performing trials versus implants following fellowship (86.6% vs 55.9%). With regards to surgical skills, most past fellows felt at least ‘somewhat’ to ‘very’ comfortable with subcutaneous pocket formation for IPG placement, surgical wound closures and infection prevention after fellowship (all greater than 63.0%). Sixty-two percent felt SCS procedures were standardized at their training institutions, and 57.0% stated that there was an unmet need for SCS training during fellowship (figure 1A).
Most past fellows (62.3%) work in private practice (hospital or office), 31.1% work in academia and 57.6% are engaged in full-time pain practice. Eighty-six percent currently conduct SCS procedures in their practice (table 3).
Those who reported performing more than 10 SCS cases (trials and implants) during their fellowship were significantly more likely to personally conduct SCS procedures in their practice at present (96.8% vs 78.1%, p<0.01). However, there was great variability in the number of procedures performed each year by past fellows with an active SCS practice. Since graduating from fellowship, 33.3% reported performing 10 or fewer SCS trials annually on average, while 26.8% perform more than 40 trials annually. There was also variability present in the number of implants performed: 42.9% reported performing less than 10 or fewer annually, while 22.7% perform more than 40. Among the respondents performing SCS procedures, 16.0% do not do implantations. Medtronic and Boston Scientific were the most commonly used SCS manufacturers (82.5% and 72.7%, respectively), followed by St. Jude/Abbott (62.8%) and Nevro (41.3%).
Of those who do not perform SCS procedures, 45% pointed to lack of training as their reason for not performing them, 25.0% questioned the long-term success of the devices (reasons not elucidated), 25% indicated lack of institutional support and 15.0% stated (declining) reimbursements. When non-SCS practicing physicians were asked what resources would be necessary to adopt SCS in their practice, 58.8% cited additional training, 41% indicated increased reimbursements (trials and implants), 35.3% indicated learning SCS practice infrastructure and 35.3% noted education identifying appropriate patients. Nonetheless, 94.1% of these respondents were found to make referrals for SCS procedures to other providers. Seventy-four percent of providers indicated that they would like to incorporate more SCS cases into their practice compared with their current volume.
Industry and standardization
Institutional standardization of SCS procedures was not found to be common, with 52.9% noting that there is no standard practice at their current institution, while 41.2% were encouraged to use the same technical approach and 2.9% were required to use the same approach. When asked who should be responsible for developing guidelines to help encourage the standardization of SCS practices, the most common answer was pain medicine societies (47.1%), followed by ACGME/fellowship directors (14.7%), fellowship/industry partnerships (8.8%) and industry workshops (1.0%). Eighty-four percent of respondents supported direct training of pain fellows by SCS manufacturers (figure 2). Fifty-nine percent have taken industry-sponsored courses since finishing fellowship.
With expanding capabilities of SCS devices and a growing need for non-opioid chronic pain treatments, current and future pain physicians will increasingly use neuromodulation therapies to provide the most comprehensive care for their patients. Use of neuromodulation, in particular SCS, can determined by the three ‘C’s’: (1) competence in the techniques, (2) confidence or belief in the therapy helping patients and (3) culture of the practice and whether it supports the use of the therapy. To our knowledge, this is the first study investigating SCS training deficiencies during and after pain fellowship.
Current graduating fellows were more likely to have received SCS training during their fellowship compared with past fellows (100% vs 84.0%), which was expected since neuromodulation training recently became required for accredited pain fellowships in 2006. ACGME strongly suggests that fellows must document hands-on involvement with at least five clinical neuromodulation procedures.5 Despite this, we found substantial variability in fellows’ SCS experiences with a wide range of trials and implants performed during fellowship. For instance, 7.5% of current fellows did not perform any SCS implant procedures, while 22.5% participated in over 15. This lack of training uniformity is consistent with the findings of Gharibo et al,7 who showed large discrepancies in procedure volume at academic institutions. Fanciullo et al 6 also demonstrated a high degree of variability in neuromodulation practice habits at academic institutions, including the length of SCS trials, implantation techniques, opioid weaning and programming algorithms. In this study, most current fellows (84.6%) noted that SCS procedures were at least ‘somewhat’ standardized at their institutions, while past fellows indicated less institutional standardization (62.0%) during their fellowship training. Nonetheless, the majority of current and past fellows believed that SCS procedures should be standardized, with pain medicine societies being the most popular choice as a lead entity. Given the complexity of implantable therapies, we believe that establishing guidelines for the management of these devices would help new providers navigate the SCS treatment algorithm and provide the best patient care. The Neurostimulation Appropriateness Consensus Committee Recommendations (which include many fellowship faculty) found in neuromodulation are a step in the right direction and should be considered for formalized training programs as part of the curriculum.9
In addition, it is not surprising that 46% of current fellows felt that there was an unmet training need during fellowship regarding SCS. Deficiency in SCS case volume was the most common barrier noted (38.5%), followed by lack of SCS curriculum (30.8%) and lack of faculty with SCS expertise (23.1%). Past fellows experienced even more deficiencies in their SCS training during fellowship, with 57% indicating that there were unmet training needs. Lack of training was also a predominant reason for past fellows choosing not to use SCS devices postfellowship.
Given the necessity that patients have general and full access to available treatments, pain fellowship program must ensure that trainees receive comprehensive training. Since certain institutional training limitations such as low clinical volume may be difficult to overcome, additional educational programs have become necessary. The majority of current and past fellows (79.5% and 55.4%, respectively) strongly supported direct training of fellows by SCS manufacturers. Industry-sponsored workshops continue to be a popular resource, with most current fellows having attended a course during fellowship. These courses often offer hands-on cadaver-based instruction in addition to formal didactics. Over 50% of past fellows have also attended industry courses since graduating from fellowship, indicating that many providers continue to receive SCS training even while they are actively in practice. This reflects the need for continued postgraduate education, especially considering the recent advancements made in SCS technology (ie, high frequency and burst stimulation)10 and the expanding clinical applications (dorsal root ganglion and peripheral nerve stimulation).11 12
While it is doubtful that these 1–2 day training courses would suffice as the sole resource for achieving SCS competency, they provide valuable supplemental training. These courses allow for increasing ‘hands-on’ time for procedures, especially with regards to the number of implants from start to finish and further refinement of surgical techniques. Furthermore, advancements in industry-based education among manufacturers have led to innovation in educational techniques including time with plastic surgeons for suturing and wound techniques, presence of lead authors from major studies and even use of virtual simulation in lead placement. However, with a growing number of states establishing laws to reduce industry ties with physicians (ie, Massachusetts and Vermont), industry-sponsored training opportunities may potentially become limited, and fellowships programs will become increasingly pressured to provide more robust training. Pain and neuromodulation societies can also be a valuable resource for providing additional training opportunities.
Although industry-sponsored training programs could bias the treatment of chronic pain states and the selection of these technologies, Gharibo et al 7 demonstrated a paradigm shift in fellowship education with less programs directors viewing manufacturer involvement negatively. Training courses sponsored by four major SCS manufacturers (Boston Scientific, Medtronic, St. Jude/Abbott and Nevro) at the time of this study were widely attended by both current and past fellows, thereby providing exposure to most of the available neuromodulation technologies in the current marketplace. Ultimately, it is the responsibility of the clinician to practice evidence-based medicine and determine the appropriate treatment modality.
In general, current fellows had more experience with SCS trials compared with implants and, unsurprisingly, felt more comfortable performing trials independently by the end of their fellowship. However, we found that the number of SCS procedures performed during fellowship and the level of comfort at the end of their training were not deterring factors for fellows’ future plans, with over 90% of graduating fellows indicating that they intended on performing SCS procedures in their future practice. This is in contrast to the past fellows group who were more likely to conduct SCS procedures in their current practice if they reported performing greater than 10 SCS cases (trials and implants) during fellowship. Although experience and comfort with SCS procedures following fellowship should predictably affect future and current providers’ decisions to use these devices, given the increasing utilization of neuromodulation, current pain fellows may be more likely to embrace it as a mainstay treatment modality. Furthermore, high clinical success rates of trials and implants during fellowship encourages enthusiasm and increased attention devoted to SCS fellowship training.13 14
In addition to the required technical training, using implantable therapies requires close relationships with device representatives, referring providers, surgery centers and surgeons. Knowledge of how to build a neuromodulation practice infrastructure was a resource that was cited by 35.3% of non-SCS practicing providers as being necessary for adopting a SCS practice. Similarly, while the vast majority of current fellows (92.3%) felt they received adequate support from industry representatives, fewer fellows had learned how to obtain the necessary practice or employer support for building a SCS practice infrastructure. Therefore, it is important that SCS training curriculums also provide comprehensive education teaching the logistics of starting an SCS practice.
Current practice beliefs and environment can shape the usage of SCS therapy locally. A small minority of past fellows indicated that they do not perform SCS procedures and do not manage SCS patients with devices. This practice habit could negatively influence new hires to a practice not to use SCS therapy.
Models from other disciplines
Interestingly enough, there have been guidelines for pacemakers and defibrillators established since the 1990s and revised in 2003.15 16 These strict criteria have been written by the North American Society of Pacing and Electrophysiology together with the American College of Cardiology to establish three levels of training, from basic management, to more invasive implantation, to level 3 full electrophysiology training. Detailed criteria such as hours of training, number of cases a center should perform greater than 100 yearly implantations to be considered a training program, presence in follow-up visits, programming and surgical revisions including infection management. While this may be more rigid and the absolute numbers need to be modified for neuromodulation, the incorporation of training standards and examination requirements are worth considering and have been attempted in the past.17
As with most survey-based investigations, there are limitations with this study. Our survey response rates were relatively low, with 26.6% of all pain fellows in ACGME accredited fellowships responding to survey A and 52.6% response rate for survey B. The overall response rate for surveys A and B was 14.0%. We recognize the limitations presented by the low response rates, which are lower than some web-based surveys of physicians but on par with others.18–21 However, lacking characteristics about non-responders, we can only point out that our data are not missing at random and that the responses to this survey should be interpreted with this caveat; we cannot make a claim that our results are generalizable.
A further limitation is the inability to calculate a precise response rate for survey C. Though our population was intended to be a census of all pain medicine providers (alumni and fellows) in the USA, the use of the email lists as a tool to saturate the population of these providers introduces the problem of coverage of excluded populations. The email lists cover populations including fellows and resident members, industry and advanced practitioners and inactive members. It is also difficult to know the number of people the survey actually reached due to spam filters and other email settings.22 Next, the survey was not incentivized due to limited industry funding.23 Finally, because the email invitations used for survey C were distributed by SCS manufacturers and pain societies (ASRA and INS), survey participation was likely biased toward pain physicians who use SCS devices in their practice. The majority of the survey C respondents also finished their fellowship training after 2000; therefore, the results for the past fellows group may not be representative of all practicing pain physicians.
In spite of these limitations, we believe that the data received from all three surveys provided valuable information regarding past and current SCS experiences that highlight the improvements that can be made to fellowship and postgraduate training.
In conclusion, while SCS training during pain fellowship has become more universal, the experiences that fellows receive are highly variable, with almost half of the current fellows indicating there was an unmet training need. With the majority of current and past fellows supporting direct training of fellows by SCS manufacturers, it is apparent that industry-sponsored courses can supplement institutional training deficiencies. Moreover, standardization of SCS procedures by pain medicine and neuromodulation societies to establish guidelines for the management of these devices and on the best practice habits may enable less experienced providers to navigate the SCS treatment algorithm.
We would like to thank Tiffany Dunn and Viktorijia Telbis from Boston Scientific, Jey Subbaroyan from Nevro Corp and Andrew Sheliton from Medtronic for their help in funding this study. We would also like to thank the fellowship directors as well as pain societies for their assistance in survey distribution, and lastly, all fellows, those recently graduated and those already practicing, for taking the time to complete our surveys.
Presented at Interim data from this work were presented at the 42nd Annual 2017 Regional Anesthesiology and Acute Pain Medicine Meeting in San Francisco, California, from 6 April to 8 April 2017.
Funding This survey has received a total of $15 000 unrestricted funding from Boston Scientific, Medtronic and Nevro Corp ($5000). All companies in neuromodulation were invited to fund this investigator-initiated study. All final design, protocol implementation and results were under direct control of the investigators. Funding was used for initial costs of survey design and setup, Institutional Review Board Fees, statistical review and travel to conferences for presentation.
Competing interests DP consults for Abbott, Bioness, Boston Scientific, Halyard, Medtronic, Sollis and Nevro Corp. He receives research support from Abbott and Medtronic. TD consults for Abbott, Bioness, Nalu, Vertiflex, Vertos, Jazz, Flowonix and Saluda. TD also has minor equity in Axonics, Saluda, Bioness, Vertos, Vertiflex and Nalu and research support from Abbott, Saluda and Mainstay. AG consults for Medtronic. NM consults for Nevro Corp, Sollis and Biodelivery Sciences. He receives research support from Medtronic and Nevro Corp.
Patient consent for publication Not required.
Ethics approval After Institutional Review Board (Weill Cornell Medical College, protocol #1507016431) approval was obtained.
Provenance and peer review Not commissioned; externally peer reviewed.