Article Text

other Versions

Procedural education for cancer-related pain in Pain Medicine fellowships: a national program survey
  1. Alex Woodrow1,
  2. Masaru Teramoto2,
  3. Mrinal Thapliyal1 and
  4. Sandy Christiansen1
  1. 1Anesthesiology and Perioperative Medicine, Oregon Health & Science University, Portland, Oregon, USA
  2. 2Physical Medicine and Rehabilitation, University of Utah Hospital, Salt Lake City, Utah, USA
  1. Correspondence to Dr Sandy Christiansen, Anesthesiology and Perioperative Medicine, Oregon Health & Science University, Portland CH15P, Oregon, USA; chrissan{at}


Introduction Cancer-related pain is ubiquitous. Targeted procedural interventions may be an important and underutilized method for improving cancer-related pain and quality of life. The goal of this study was to determine the baseline educational experience and perceived post-graduation comfort of Pain Medicine fellows in performing procedures that can be used for cancer-related pain.

Methods Using Qualtrics, a 16-question survey was distributed to graduating fellows at accredited Pain Medicine Fellowship programs in the USA in June 2022. The fellows’ experience and comfort level performing eight procedures was analyzed using summary descriptive statistics and contingency table analysis with statistical significance determined by Pearson’s χ2 test.

Results Survey respondents represented 30% of fellows among 44% of Pain Medicine programs during the 2021–2022 academic year. The majority of respondents reported no experience performing 7 of the 8 procedures for cancer-related pain (56.1%–90.6%). Graduating fellows reported overall comfort performing sympathetic neurolysis (65.7%), spinal cord stimulator trials (85.7%) and implantations (66.0%), but not intrathecal pump trials (36.9%) and implantations (31.3%), peripheral nerve stimulator implantations (41.7%), vertebral augmentations (31.3%), and vertebral body radiofrequency ablations (16.7%).

Conclusions Experience performing 10 or more of the surveyed procedures, personal interest in treating cancer-related pain, and attendance of cancer-related pain lectures were found to significantly increase comfortability in performing cancer-related pain procedures post graduation, whereas cadaver-based learning did not. This study highlights the need for more robust procedural education for cancer-related pain and identifies procedural experience in non-cancer patients and lectures on cancer-related pain as ways to bridge this educational gap.

  • Cancer Pain
  • Pain Management

Data availability statement

Data are available upon reasonable request. Not applicable.

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, an indication of whether changes were made, and the use is non-commercial. See:

Statistics from

Request Permissions

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.


  • Prior studies describe the educational experiences regarding spinal cord stimulators, ultrasound-guided procedures, and opioid prescribing in Pain Medicine fellowships, yet the educational experience and perceived postgraduation comfort of Pain Medicine fellows in performing procedures for cancer-related pain are limited.


  • Our novel study quantifies the suboptimal experience of Pain Medicine fellows in performing specific procedures that can be used for cancer-related pain and their low level of comfort with independently performing these procedures following graduation.


  • To ensure comfort in independent practice, Pain Medicine program directors should commit to every fellow performing at least ten of each procedure on any combination of cancer and non-cancer patients, or if not feasible, the curriculum should be supplemented with lectures on interventions for cancer-related pain.


Cancer is the second leading cause of worldwide death and as such, cancer-related pain is ubiquitous.1 Between 2005 and 2014, the cancer pain prevalence in oncology patients was 39.3% after curative treatment, 55% during treatment, and 66.4% in advanced disease.2 Direct cancer-related pain caused by the tumor itself can present at every stage of the disease process and can be further complicated by treatments including surgery, chemotherapy, and radiation.3 Not surprisingly, 43% of oncology patients reported undertreated pain.4

While medication management with opioids is the mainstay of pain management for patients with cancer, it has several limitations, namely the opioid tolerance and adverse effects that come with high and sustained doses.5 Using a procedural intervention may be an important underused alternative for improving pain control and quality of life.5–7 There are several options for cancer pain-relieving procedures including targeted neurolysis, intrathecal drug delivery systems, and kyphoplasty among others.7

The Accreditation Council for Graduate Medical Education (ACGME) approved Pain Medicine as a subspecialty of anesthesiology in 1991 with the goal of training physicians in chronic pain management, including cancer-related pain management.8 Today, Pain Medicine fellows include graduates of anesthesiology, physical medicine and rehabilitation, neurology, psychiatry, family medicine, emergency medicine, and other residency programs. Although the ACGME mandates longitudinal experience in cancer pain care, there is no quantitative requirement to graduate. Therefore, the variability in populations-served, faculty interest, and training exposure culminate in a broad range of procedural experiences for fellows.

Interventional cancer-related pain management remains understudied, and the non-specific education requirements for training in cancer pain interventions lead to a national lack of access to interventional pain procedures for patients with cancer-related pain. While there are no other studies to date on interventional cancer pain training during Pain Medicine fellowships, several studies have explored the state of Pain Medicine training.9–11

One study established that while opioid prescribing practices among US fellowships varied, there was a shift away from opioid prescribing towards non-opioid medications and interventions.9 Another study, by Gharibo et al, showed that program directors felt that the most valued source of training in spinal cord stimulation was during the fellowship, followed by manufacturer-sponsored workshops, and lastly conferences.10 Finally, a study by Conway et al established that US and Canadian fellowship programs used ultrasound for peripheral nerve blocks, non-axial musculoskeletal injections, and axial nerve blocks in descending frequency.11

Our study goals were to determine the baseline educational experiences of Pain Medicine fellows in procedures that can be used for cancer-related pain and their comfort in performing these interventions following graduation. By elucidating the baseline experiences, we will propose methods of improving future education processes and offerings to enhance patient access. To accomplish this, we surveyed Pain Medicine fellows at ACGME-accredited programs in the USA.


The survey was engineered using a thorough literature review, statistical planning, as well as discussions with practicing interventional cancer pain physicians and Pain Medicine fellows in the USA. Our study was crafted with the US perspective in mind and thus focused solely on: (1) the educational experiences of Pain Medicine fellows at ACGME-accredited fellowships and (2) procedures described by experts in interventional cancer pain in the literature published in the USA.7 12 13 We established content validity through discussions with stakeholders, including practicing Pain Medicine physicians and fellows. The stakeholders reviewed the questions and stated that the questions will lead to the measurement of what we want to measure.

Eight procedures were analyzed: (1) sympathetic ganglia neurolysis (celiac plexus, superior hypogastric, ganglion impar), (2) intrathecal pump trial, (3) intrathecal pump implantation, (4) spinal cord stimulator (SCS) trial, (5) SCS implantation, (6) peripheral nerve stimulator implantation, (7) vertebral augmentation, and (8) vertebral body radiofrequency ablation. Comfort levels by the number of procedures performed in each of the eight procedures were summarized separately for cancer-related pain, non-cancer-related pain, and cadavers-spine models.

The number of procedures performed was collected as a combination of continuous (0–10 procedures) and categorical scales (>10 procedures) and analyzed on the continuous scale of the number of procedures (ie, 0–10 procedures). Comfort level and the number of procedures performed were also analyzed as categorical variables, with statistical significance determined using Fisher’s exact tests. The categories for a number of procedures were chosen based on the historical ACGME requirement for a quantity of five SCS trial procedures required for graduation. Using this benchmark along with the idea that the other procedures also represent advanced interventional pain procedures, the researchers hypothesized that the 5–10 range would represent adequate experience that would translate to comfortability while the 0 and 1–4 categories would not reach comfortability.

Additionally, Pearson correlation coefficients (r) were computed to examine the associations of the number of procedures performed (continuous scale with 0–10 procedures) among the three categories of experience (cancer, non-cancer and cadaver). Comfort level was collected as a three-level variable: comfortable, not comfortable, and not yet comfortable with plan to seek additional training and converted into dichotomous variables for the data analysis: comfortable and not comfortable.

During the 2021–2022 academic year, a total of 110 ACGME-accredited Pain Medicine Fellowship programs with 362 fellows were identified using the Fellowship and Residency Electronic Interactive Database.14 In May 2022, the study team contacted the Program Directors and Coordinators of all programs to obtain the academic email addresses of graduating Pain Medicine fellows.

In June 2022, using Qualtrics software, a 16-question survey (online supplemental file 1) was emailed to the 239 fellows in 71 programs. For the remaining 39 programs, the study team did not have fellow email addresses, and the survey was emailed to both the program director and the program coordinator with instructions to forward the survey link email to their fellows. The electronic survey landing page included informed consent, an overview of the survey, the estimated time to complete, and potential risks associated with completing the survey. To proceed, all respondents were required to attest that they provided consent.

Supplemental material

To incentivize responses, all fellows received a US$5 Starbucks gift card regardless of survey completion and were asked to provide their academic email address in the last question of the survey to ensure those who received the survey from Program Directors or Program Coordinators also received the gift card.

After emailing the questionnaire on June 1, 2022, the study team sent weekly follow-up emails asking fellows to complete the survey. The survey closed on June 30, 2022, at midnight Pacific Standard Time.

Statistical analysis

Descriptive statistics were calculated for the survey data using median and IQR as the data for the majority of procedures demonstrated a skewed distribution. Experience in performing procedures and comfort levels were analyzed using summary descriptive statistics and contingency table analysis with statistical significance determined by Pearson’s χ2 test (or Fisher’s exact test in case of low expected frequencies). The relationship between interest level and educational settings on comfort level was examined using Pearson’s χ2 tests/Fisher’s exact tests and point-biserial correlation coefficients (rpb). Lastly, the receiver operating characteristic (ROC) curve analysis15–17 was performed for each procedure to determine the optimal quantity (only including ≤10 procedures performed) for fellows to feel comfortable. An area under the ROC curve (AUC) and 95% CI were calculated for each ROC analysis. All analyses were conducted with Stata/MP V.17.0 (StataCorp LLC, College Station, Texas, USA), and an α level of 0.05 was used for statistical significance.


The survey was sent to 317 individuals, which included 239 fellows, 39 program directors, and 39 program coordinators. A total of 107 survey responses were recorded, representing 30% of Pain Medicine fellows. Survey respondents were from 48 of the 110 Pain Medicine programs in the USA, representing 44% of programs. Overall, 86% (92 out of 107) completed the entire survey. Most survey questions had less than 10% of missing data. Questions regarding educational settings, interest level, and identifiable barriers to physicians’ educational experience had 12.2% of missing data.

The regional breakdown of fellows during the 2021–2022 academic year was 34% in the Northeast, 24% in the Midwest, 20% in the Southeast, 6% in the Southwest, and 16% in the West. The distribution of the 107 survey respondents by region was 29% in the Northeast, 19% in the Midwest, 14% in the Southeast, 5% in the Southwest, and 33% in the West. Comparisons of the breakdown of respondents versus the distribution of fellows highlight a grossly parallel representation.

Most fellows (71.3%) reported that they were very interested (39.4%, n=37) or interested (31.9%, n=30) in working with a cancer patient population. All fellows who reported that they were very interested or interested in working with a cancer patient population completed the entire survey, while all incomplete surveys were from fellows who were neutral or not interested in working with this population.

The overall median number of procedures performed ordered from highest to lowest were SCS trials (median=14) and implantations (median=10), sympathetic ganglia neurolysis (median=4), peripheral nerve stimulator implantations (median=2), vertebral augmentations (median=2), intrathecal pump implantations (median=1.5), intrathecal pump trials (median=1), and vertebral body radiofrequency ablations (median=0) (table 1).

Table 1

Descriptives of the number of procedures performed and its association with comfort level of performing each procedure by type of pain, along with overall comfort level

Fellows reported being most comfortable performing sympathetic ganglia neurolysis (65.7% comfortability), SCS trials (85.7% comfortability), and SCS implantations (66.0%) after graduation and least comfortable performing vertebral body radiofrequency ablations (16.7% comfortability) (table 1).

The data for procedural education for cancer-related pain demonstrated a positive skew for all procedures indicating that most fellows had minimal exposure to this patient population, and many had no exposure at all (table 2). On average, 69.5% of fellows reported no exposure to any procedures for cancer-related pain and 90% reported no exposure to peripheral nerve stimulators and vertebral radiofrequency ablations for cancer-related pain (table 2).

Table 2

Fellow procedural experience (number of procedures performed) in cancer-related pain

When examining fellows’ comfort levels by the number of procedures as a categorical variable (0–10 procedures vs >10 procedures), comfort varied by procedure and by cancer versus non-cancer pain (table 1). For cancer-related pain, fellows who performed >10 procedures for sympathetic ganglia neurolysis, intrathecal pump trials, intrathecal pump implantations, and vertebral augmentations were significantly more likely to report that they were comfortable compared with those who performed 0–10 procedures (p<0.05) (table 1).

Respondents who reported that they were very interested or interested in working with a cancer patient population were also more likely to report that they were comfortable in performing the surveyed procedures.

Most programs had between 4 and 6 fellows (66%, n=62) with 6 or more faculty (86.2%, n=81) and were at a university hospital setting (72.3%, n=68). Correlations between lectures and educational setting with comfort level indicate no quantity of fellows or faculty in the program was significantly associated with leading to increased comfortability for any of the procedures. The number of cancer pain lectures received clustered in two groups: between 2 and 5 lectures (65%, n=61) or >10 lectures (17%, n=16). The quantity of lectures for cancer-related pain (continuous variable with 0–10) that physicians received was significantly associated with feeling comfortable in performing procedures of intrathecal pump trials, intrathecal pump implantations, SCS trials, SCS implantations, and vertebral augmentations (table 3; rpb≥2.51; p<0.05).

Table 3

Correlations between the number of lectures for cancer-related pain and comfort level of performing each procedure

There was a significant association between performing a procedure and comfort performing it after graduation for all procedures (p≤0.002; table 4). The comfort level related to the quantity of each of the surveyed procedures in non-cancer-related pain performed during training, analyzed using the ROC curve analysis, is summarized in online supplemental file 2. For certain procedures, such as sympathetic ganglia neurolysis and intrathecal pump trials, some fellows reported that they were comfortable performing the procedure post graduation in the cancer population even though they had never done the procedure in training for cancer-related pain, only in the non-cancer population (table 4). A similar trend was observed for SCS trials and SCS implantations.

Supplemental material

Table 4

Associations of experience (ever performed procedure) and comfort level of performing each procedure among fellows


Understanding baseline Pain Medicine training at ACGME-accredited fellowships is critical to planning future curricula that will aim to close national knowledge gaps related to interventional cancer-related pain management. Optimal future curriculum needs to involve both changes within fellowships and through supplemental offerings, including conference workshops and lecture series. The primary goal of this survey study was to quantify cancer pain procedure experience and assess the threshold of comfortability in using those procedural skills in future practice.

The response rate of this study was similar to the response rate to a study by Pak et al that targeted Pain Medicine fellows.18 Other studies evaluating Pain Medicine fellow education that relied solely on program directors to provide data, demonstrated higher response rates.9 10 19–21 While surveys targeting program directors may lead to higher response rates, using an intermediary to evaluate fellow experience introduces potential limitations. These include possible pressure for program directors to portray their program positively and may inaccurately estimate procedural experience since most programs have multiple attendings and clinical sites.

This study surveyed perceived comfortability of the fellows to infer the likelihood that they will perform these procedures on cancer-patients on graduation. Our study team was specifically interested in establishing quantitative thresholds for each procedure at which point most fellows stated they were comfortable performing the procedure in their future practice. In the past, procedural volume has been used as a benchmark ACGME requirement for graduation. For example, when neuromodulation was introduced, a minimum of five SCS trials were required for graduation.22 However, recently the ACGME Common Program Requirements policies removed discrete procedural volume requirements for graduation.23

Regardless of the presence or absence of discrete procedural volume in ACGME graduation requirements, we found a direct relationship between the number of procedures performed and comfortability for every procedure with the most performed procedures being SCS trials (n=1293, 85.7% comfortability), SCS implantation (n=925, 66.0% comfortability), and sympathetic ganglia neurolysis (n=649, 65.7% comfortability) (table 5).

Table 5

Total number of procedures performed by type of pain and overall comfort level

Empirical thresholds of comfortability were only able to be obtained for SCS trials and implantations with five and four procedures, respectively, as these procedures were associated with higher quantities of experience (online supplemental file 2). Though limited to only two types of procedures, these thresholds are consistent with the prior graduation requirement by the ACGME Program Requirements for spinal cord stimulation.

In lieu of an empiric threshold, a categorical analysis of procedure volumes with reported comfortability identified a threshold of 10 as the most likely volume needed for fellows to feel comfortable performing the procedure for cancer-related pain in independent practice for sympathetic ganglia neurolysis, intrathecal pump trials, intrathecal pump implantations, and vertebral body augmentations (table 1).

As an example, 100% of the fellows who performed >10 intrathecal pump trials (n=5) reported that they would feel comfortable performing them after graduation compared with only 34.7% of fellows (n=33) who performed <10. The difference in comfortability between those who performed more than 10 sympathetic ganglia neurolysis (n=16, 100% comfortable) and those who did not (n=52, 59.8% comfortable) was less disparate, suggesting that there were other mechanisms that increased comfortability among those fellows (table 1).

Notably, with the exception of sympathetic ganglia neurolysis and intrathecal pump implantations, ≤6.3% of respondents reported performing 5–10 of the remaining survey procedures and ≤7.1% reported performing 10 or more (table 2), highlighting a significant gap in medical education in the USA.

A positive correlation was also observed between performing a large volume of certain procedures such as sympathetic ganglia neurolysis, intrathecal pumps, and vertebral augmentations in non-cancer patients with comfortability also performing these procedures in the cancer patient population. This suggests that institutions with a high volume of interventional exposure to both cancer and non-cancer patients may yield higher overall comfortability among their fellows.

Furthermore, when examining procedures that were done on patients with cancer, the only procedure for cancer-related pain that approached a third of respondents completing five or more was sympathetic ganglia neurolysis. Surprisingly, while the majority of survey respondents did not have experience performing seven out of the eight procedures specifically for cancer-related pain (table 2), most of them endorsed comfort in performing all but one of the procedures in independent practice on patients with cancer (table 4). Thus, analysis of comfort level showed that experience with doing the procedure on non-cancer patients translates to comfort in performing the procedure in the cancer patient population. When the lectures in cancer-related pain interventions increased, so did the comfort level, indicating that when increasing procedural volume is not feasible, didactics may fill the gap in procedural education (table 3).

In terms of other learning mechanisms, an average of one-third of fellow procedural experience came from cadaver labs or spine models (table 5). Despite this, there was no statistically significant relationship between cadaver procedural experience and comfortability (table 1), which may suggest that lectures are a more effective mechanism for increasing comfort compared with cadaver training (table 3). One explanation for this phenomenon is that cadaver workshops typically assign multiple learners to a single cadaver station, which limits the time each person can practice the technique, thereby decreasing the utility of the exercise. Alternatively, since the procedures used in both cancer and non-cancer patients are largely the same, the cancer-specific knowledge obtained through lectures may have given fellows the confidence needed to apply their already developed procedural skills in a much different patient population.

The findings of this study are limited by three main factors. First, while our response rate was higher than a similar study18 and the geographical distribution of respondents was representative, less than one-third of all graduating fellows nationwide responded. However, as 44% of the programs with parallel geographical distribution were represented, the study team is confident that the results of these studies are meaningful, generalizable, and should be used for educational planning. Second, because the study team did not conceal the question being studied, it is likely that there was selection bias regarding who completed the survey, as most respondents (71.3%) expressed interest in working with a cancer patient population following graduation. Third, while the procedures identified in this study do not encompass all procedures used to treat cancer-related pain, we believe that they capture a reflection of the majority of current practice patterns and emerging techniques in the USA. Future studies may examine quantities and comfortability of peripheral nerve ablation and intrathecal neurolysis in the treatment of cancer-related pain, as they remain standard of care for much of the world and their inclusion in ACGME curricula would bolster procedural options for cancer-related pain in the USA.

As a final note, while it is important for fellows to feel comfortable performing a given procedure after graduation, comfortability does not equal competency, underlining the importance of fellowship oversight and evaluation prior to graduation. Thus, while producing fellows who are comfortable working with a cancer pain population will increase access to quality of life improving procedures, care must also be taken to ensure these fellows have adequate training particularly given the unique considerations when treating patients with cancer.24 Establishing minimum requirements for cancer pain patients in ACGME graduation requirements may help accomplish this.


Our study highlights the need for more robust procedural education for cancer-related pain. This suggests that alternate experiences such as performing these procedures on non-cancer patients as well as attending interventional cancer pain lectures may yield comfortability in procedures even if fellows are not exposed to patients with cancer during the fellowship. Second, experiences with performing the procedures on cadavers were not associated with comfort after graduation, which may indicate that the return on investment of these workshops is minimal. Last, our study findings suggest that fellows who perform at least 10 of a given procedure on cancer and non-cancer patients combined may feel comfortable performing them in practice. Future research should focus on capturing a larger portion of the fellows training in the USA to better identify barriers to procedural education.

Data availability statement

Data are available upon reasonable request. Not applicable.

Ethics statements

Patient consent for publication

Ethics approval

The Oregon Health and Science University Institutional Review Board reviewed the study and determined that the proposed activity is not research involving human subjects and approval is not required (STUDY00024250). Participants gave informed consent to participate in the study before taking part.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Contributors SC designed the study, distributed the survey, collected the data, wrote the manuscript, prepared the manuscript for submission and is the guarantor for the study. AW designed the study, wrote the manuscript, and prepared the manuscript for submission. MTeramoto designed the study, analyzed the data, wrote the manuscript, and prepared the manuscript for submission. MThapliyal collected the data, wrote the manuscript, and prepared the manuscript for submission.

  • Funding This research received grant funding for survey incentive gift cards from the Pacific Spine and Pain Society.

  • Competing interests SC serves on the Executive Committee for the Pacific Spine and Pain Society. This relationship has been reviewed and managed by Oregon Health & Science University (OHSU). AW, MTeramoto, and MThapliyal do not have any competing interest to disclose.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.