Understanding providers' perspectives using patient-reported outcomes in a pain management setting: a pilot study

Abstract

Aim: This pilot study evaluated an electronic patient-reported outcomes collection system in pain management to understand providers' experiences using the data, and how it affects their patient interaction and guides their clinical decision-making. Materials & methods: Using stratified convenience sampling, nine semi-structured interviews were conducted with consented pain physicians. The transcribed, de-identified interviews were coded and analyzed. Results: Although most physicians utilized patient reported outcomes (PROs), one-third reported no significant change in their practice since implementation and 56% stated it does not influence their treatment recommendations. Conclusion: Despite the importance of measuring the impact of chronic pain on quality of life, there are significant limitations to the real-world use of PRO that may limit the patient's assessment and care.

Plain language summary

Summary points.

• The importance of measuring the impact of chronic pain on overall functioning and health-related quality of life outcomes in chronic pain management has been established and is recommended by almost every consensus guideline.

• Patient reported outcomes (PROs) can be useful in objectively measuring these domains and guiding clinical decision-making at point-of-care; however, the perspectives of pain management physicians in using these PROs have not previously been reported but are paramount in understanding how it can be best implemented and used in clinical practice.

• The Pain Management division at Weill Cornell Medicine implemented a patient-reported outcomes collection and reporting system using the Patient Reported Outcomes Measurement Information System (PROMIS^®).

• Using stratified convenience sampling, nine semi-structured interviews were conducted on consented chronic pain physicians practicing at Weill Cornell Medicine's Multidisciplinary Spine Center and the transcribed interviews were de-identified, coded, and analyzed.

• Since the integration of PROs, 67% of the physicians report utilizing the data while 33% report not using it or not regularly using it.

• Despite following consensus guidelines in implementing the PROs, a third of the physicians interviewed did not report a significant change in their practice since its implementation and a majority reported that it does not influence their treatment recommendations.

• There were many physician-perceived benefits of the PRO data, including that it provides more detailed information that may have been missed during the clinical encounter, and it helps provide objective value in patients' trends in pain scores and the functional impact of their chronic pain.

• There were several physician, patient and technical-level barriers that were discussed, including that the physicians still need to confirm the questionnaire data with the patients, low patient completion rates and clinician time constraints.

• Despite the value of PROs in measuring the impact of chronic pain on quality of life, further research is still needed to streamline the process for meaningful and efficient PRO data capture in order to improve individualized patient care, treatment outcomes and the impact of chronic pain on a population level.

Chronic pain treatments vary between pain management centers, and even different practitioners within a given practice. Pain management treatments should be selected according to the patient's response to therapy, yet these outcomes are often not objectively measured using validated tools and tracked throughout their course of treatment [1,2]. As endorsed by the pain research community, as well as the Institute of Medicine recommendations and the National Pain Strategy, it is imperative to measure not just pain severity, but also the impact of chronic pain on overall functioning and health-related quality of life [3–5]. These patient-reported outcomes (PROs) can also be useful in guiding clinical decision-making regarding the safe and effective use of medications, including opioids, for chronic pain management. A systematic review by Holmes et al. found that while PROs may help with pain assessment and influence clinical decision-making as well as the evaluation of treatment effectiveness, further research is still needed to guide the best use of the data [6].

Many PROs can be used to provide information about the severity of pain and its impact on a patient's physical and psychological well-being. The Patient Reported Outcome Measurement Information System (PROMIS^®) is a set of tools that was developed through the National Institutes of Health and offers several advantages over legacy instruments. It is a dynamic, psychometrically-validated tool that can be used to assess PROs across a wide range of chronic diseases and demographic characteristics [7,8]. Moreover, it offers significant efficiency when item-response theory and computer adaptive testing (CAT) are utilized to decrease respondent burden and time required to complete all assessments. Specifically, the PROMIS measures PROs across physical, social, mental and global health domains [7,9].

The Pain Management division at Weill Cornell Medicine (WCM) implemented a PRO collection and reporting system using PROMIS in August of 2018 for all patients seen at the multidisciplinary spine and pain clinic, which was seamlessly integrated into the electronic health record (EHR). The specific PROMIS domains measured include ‘Pain interference’ (PROMIS CAT V1.1), ‘Physical function’ (PROMIS CAT V2.0), ‘Depression’ (PROMIS CAT V1.0), ‘Anxiety’ (PROMIS CAT V1.0), ‘Fatigue’ (PROMIS CAT V1.0), ‘Sleep disturbance’ (PROMIS CAT V1.0), ‘Ability to participate in social roles and activities' (PROMIS CAT V2.0), and ‘Global mental and physical health’ (PROMIS Adult Short Form V1.0 Global Health). All providers were trained in using the EHR to locate and graph patient questionnaire responses. Smart phrases were developed and shared with all clinical staff to facilitate importing the data into clinical notes. The data is made available to the physician in real-time to review with the patient during their clinical encounter to help guide point-of-care clinical decision-making. In addition, administrative staff were trained in asking all patients to complete the surveys at the time of appointment scheduling as well as during their reminder phone calls. Questionnaires were automatically sent to patients based on their scheduled office visits, with email and text reminders to complete their e-check-in including the questionnaires at several time points before their appointment. In addition, the PRO data across the multidisciplinary spine center are being aggregated in a chronic pain registry for analysis in quality improvement and research projects. There is no financial incentive for physicians or staff to use these PROs.

Prior literature examined the impact of PROs for chiropractors [10] and physiotherapists [11] with noted preference for functional scales and outcomes over pain scales, and the benefit of increased patient involvement in their care. Similarly, this was reported in surgeons across eight surgical specialties highlighting some of the advantages of PRO data, including that it can modify surgical practice on both an individual and institutional level [12]. Another recent qualitative evaluation looked specifically at orthopedic surgeons and found that participants stated that aggregate PROs are more valuable for departments and individual surgeons than for individual patients [13].

The use of PROs was also evaluated in the primary care setting. One such study used a mixed-method approach of a survey and qualitative analysis and found that the main benefits of PROs included aiding clinical management, utilizing the data as a screening/diagnostic tool and facilitating shared decision-making [14].

Boyce et al. published a systematic review describing healthcare professionals' experience using PROs [15]. They concluded it is most useful for clinical decision-making, but that significant barriers remain including inappropriate infrastructure before implementation and disruption to normal clinical routines. They suggested the importance of utilizing technology and ensuring transparency regarding the rationale for data collection [15].

Few studies looked at the utilization of PROs in routine clinical care from the patient's perspective. One such study that evaluated both the patient and physician perspectives highlighted that patients felt the use of PROs could provide clinicians insight into their experiences living with their respective diseases and could help them to recall issues they want to discuss with their clinician [16]. Also, they noted facilitators of use of the PRO data including clinician enthusiasm and sending patients reminders to complete the questionnaires [16]. Similar trends have been reported in specific disease states, including chronic kidney disease, also adding that electronic PROs would be able to discern signs of deterioration of a patient's condition and emotional well-being, increase patients' self-reflection and could help the clinician focus on concerns prioritized by patients [17].

In addition to evaluating physician and patient perspectives, there have been publications evaluating other healthcare team members' perspectives. In Zhang et al., physicians reported not using PRO results due to a lack of actionable data, citing barriers including lack of financial incentives for use. However, social workers and psychologists more reliably used the PRO data to evaluate mental health and provide additional resources when clinically indicated [18]. In addition, while medical assistants were well trained in the process, there was a perceived lack of benefit as they had little knowledge of the details of the questionnaires [18]. In a survey of staff working in a comprehensive cancer care setting, respondents expressed an overall positive attitude toward PROs with physicians being more knowledgeable than other professionals and with a stronger preference for electronic collection compared with nurses, but low usage in clinical practice was noted overall [19]. It is noted, however, that PROs allow for facilitating difficult, end-of-life communication between clinicians and patients to better tailor end-of-life care in the palliative setting [20].

While the importance of measuring quality of life in chronic pain management has been established and strongly recommended by most consensus guidelines [4,21,22], there are no published studies looking at physician perspectives in the pain management setting in using PROs that we are aware of at the time of this study. These physician perspectives are paramount to understanding how it can be best implemented and used in clinical practice without disrupting of the normal clinical workflows. This is a pilot study to evaluate the implementation of a PRO collection system in the pain management division of an academic medical center. Information is used to understand providers' beliefs and experiences using the data, and how it affects their patient interaction, discussion and guides their clinical decision-making. It is hypothesized that there is a wide spectrum of utilization, from providers who check it each visit and review it with their patients in detail, graphing the results and discussing the implications directly with patients, to providers who ignore all the data and may feel more burdened by it.

Materials & methods

The principles of grounded theory guided the study design, data collection and analysis. This is a qualitative study, utilizing in-depth open-ended questions via individual interviews. Written informed consent was obtained for all participants and the study was approved by the Institutional Review Board Office at WCM. Stratified convenience sampling was utilized to provide information-rich cases for an in-depth study. The inclusion criteria were full-time or part-time chronic pain physicians at Weill Cornell Medicine's Multidisciplinary Spine Center. These physicians are all board-certified in Pain Management and Anesthesiology and are faculty at WCM.

A standard semi-structured interview guide with probes was used. The interview guide included topics and questions that the interviewer hoped to cover during the interview but did allow for flexibility with individual interviews and throughout the study (See Appendix I for full guide). All questions were piloted in advance, with the entire research team on the interview, and then modifications were made after discussion post-interview, and agreed upon by all members of the study group.

Interviews were conducted via Zoom^© and hosted by a member of the research group who was unbiased and impartial to the PRO implementation process. Only the interviewer could identify the subjects. The process was overseen by a second member of the study group who ensured that the interview remained objective and that nothing from the guide was omitted. Audio taping of the sessions was recorded using Zoom, creating a verbatim transcript. The recordings and transcripts were saved on a secure file. All data were de-identified before analysis.

The data analysis was summative, in which all analysis was conducted at the end of the study, as questions were driven by the purpose of the study. A coding scheme was established and agreed upon by the research team. Dedoose^© software was utilized to assist with coding and analysis. Open coding was utilized to identify concepts from the transcribed quoted data and labels were applied. The properties and dimensions of the data were analyzed. The initial code was used to organize the data from the first individual interviews and then refined during the review and analysis of the subsequent interviews. Categories were developed using axial coding, as we compared and contrasted concepts. Finally, selective coding was used to integrate and refine the data into primary themes and the central or core category was identified as related to the central phenomena.

One independent researcher coded all transcripts and a second researcher checked portions of coding in each transcript to ensure inter-rater reliability and validity. Greater than 90% of the codes were agreed upon in the first round. Any discrepancies were discussed among joint sessions with all co-authors until a consensus was reached to ensure reliability and validity. This was an iterative process with constant comparative analysis which terminated upon data saturation.

Sociodemographic variables were also entered into Dedoose to allow for a mixed methods analysis. The variables evaluated include the academic rank of the physician, gender, years in practice since completion of pain management fellowship, years in practice as an attending physician at WCM, whether the pain physician completed part or all of their medical training outside of WCM, and if they started as an attending at WCM before or after the implementation of the PROs.

Results

After one test interview, nine subsequent interviews with chronic pain physicians at WCM were completed and analyzed. This sample size of nine participants represents all the chronic pain physicians who are employed in the department of anesthesiology at WCM. There are several other practicing pain physicians from other departments, including physiatry and internal medicine, that were not included in this pilot study. Data saturation was achieved at seven interviews. Individual interviews ranged from 10 to 32 min per participant, with an average interview time of 19 min. Of the interviewees, six were males (67%) and three were females (33%). Mostly junior faculty were part of the pilot study with two holding an academic rank of instructor (22%), six assistant professors (67%) and one associate professor (11%). Years in practice since completing fellowship ranged from 3 months to 22 years, with an average of 7.3 years (SD 6.8), with six physicians (67%) completing part of their training outside of WCM. Years in practice at WCM ranged from 3 months to 21 years, with an average of 6.3 years (SD 6.6). Five physicians (56%) started practice at WCM before the implementation of the PROs and four physicians (44%) started after the integration was complete (Table 1).

Table 1.

Physician characteristics (n = 9).

		Number
Rank	Associate Professor	1
	Assistant Professor	6
	Instructor	2
Gender	Female	3
	Male	6
Years in practice	≤5	4
	6–10	3
	11–15	1
	≥16	1
Years at Weill Cornell Medicine	≤5	5
	6–10	2
	11–15	1
	≥16	1
When physicians started practice at WCM in reference to the PROs integration	After	4
	Before	5
Physicians that completed part of their training outside of WCM	No	3
	Yes	6

PRO: Patient-reported outcome; WCM: Weill Cornell Medicine.

The initial part of the interview was conducted to evaluate physicians' baseline beliefs regarding the evaluation and treatment of pain. When asked how the physicians evaluate pain severity, 78% (n = 7) reported relying primarily on pain scores, 33% (n = 3) reported also using physical exam findings, and 22% (n = 2) reported relying on visual cues as well.

In terms of assessment of the quality of life, level of functioning and mood, 22% (n = 2) of the interviewed physicians reported they assess a patient's mood only if they have a psychiatric diagnosis, 33% (n = 3) only assess if the patient specifically mentions their mood or if patients are taking mood stabilizing medications, or if the quality of life, level of functioning and mood information is already available in the EHR. 44% (n = 4) of the physicians assess the impact of pain on activities of daily living and 78% (n = 7) reported assessing the quality of life, level of functioning, and mood in general, while 88% (n = 8) reported assessing activity level, sleep or work to measure one of those domains. In terms of what data is used to guide treatment, 55% (n = 5) of the interviewed physicians reported using diagnostic testing data, imaging results or the medication list, while 22% (n = 2) reported using pain scores and 44% (n = 4) used previous treatments to tailor current treatment recommendations.

In terms of documentation and tracking assessments of pain, 33% (n = 3) of interviewed physicians reported documenting in the same note and 67% (n = 6) followed up on how patients are doing at the current visit and compared with prior visits or follow-up on prior discussions. 22% (n = 2) reported explicitly documenting the PROMIS questionnaire data in their clinical note, and 33% (n = 3) reported documenting and tracking mood while only 22% (n = 2) reported documenting sleep, activity level or level of functioning. Only 22% (n = 2) reported documenting and tracking pain scores, the impact of pain, or the overall impact on quality of life.

In terms of how the physicians elicit patient expectations, values, and preferences, 56% (n = 5) ask about patients' viewpoints on treatment, 33% (n = 3) ask about patient preferences regarding treatment options, and 22% (n = 2) discuss managing patient expectations.

When asked how they ensure that the pain assessments are repeatable and reliable, 56% (n = 5) reported a lack of confidence in the duplication of the assessments. Furthermore, specific standardized assessment methods were not used by the physicians. As a workaround, 33% (n = 3) repeatedly asked the same questions and 22% (n = 2) reported that it must be reliable as it is based on their own assessment.

The majority of interviewed physicians reported minimal to no experience using PROs before their implementation at WCM in August of 2018. Since the integration of PROs at WCM, 33% (n = 3) of physicians interviewed reported not using or not regularly using the PRO data, while 67% (n = 6) do use it (Figure 1). 44% (n = 4) of physicians reported it is most important or they review it in greater detail for new patients, while 33% (n = 3) do not review it at all for follow-up patient visits. 55% (n = 5) of physicians review the PRO questionnaire data before the patient visit, while 33% (n = 3) review it simultaneously while talking with the patient during the clinical encounter. 44% (n = 4) of the physicians review it directly with the patients while 33% (n = 3) do not review it with the patients or only if there is a major outlier that is unusual from prior visits (Figure 2).

Figure 1.

Use of patient-reported outcome data following integration at Weill Cornell Medicine.

PRO: Patient-reported outcome.

Figure 2.

Review of patient-reported outcome data.

Some physician-perceived benefits of the PRO data are that 22% (n = 2) believed it provides more detailed information that may have been missed otherwise during the clinical encounter, and another 33% (n = 3) of the interviewed physicians reported it helps provide objective value in patients' trends in pain scores and functional impact, allowing tracking of their progress over time (Figure 3). Despite some of these noted advantages, 33% (n = 3) of the interviewed physicians reported no significant change in their practice overall since implementation, specifically with 56% (n = 5) reporting it does not influence their treatment recommendations and 22% (n = 2) reporting it is helpful specifically when prescribing opioids or to tailor their treatment, especially again emphasizing the importance of utilizing these measures for new patients.

Figure 3.

Benefits of patient-reported outcomes.

When evaluating the interview responses and sociodemographic variables, physicians with a higher academic rank tend to be more likely to report that the implementation of these PROs did not change their practice relative to those with a lower academic rank. Similarly, this trend is apparent when looking at increased years in clinical practice is associated with an increased likelihood that the physician reports no significant change in their practice with this PRO implementation.

In addition, several other trends were apparent when looking at interview responses relative to years in practice and academic rank. Those physicians who were interviewed with fewer years in practice were more likely to ask about psychiatric comorbidities and medications while more experienced attendings tended to only ask if the information was already available in the EHR. Similarly, as the years in practice increased there was a trend that physicians were less likely to ask about the impact of pain on activities of daily living but more likely to ask about the impact of the patient's pain on their ability to work. Other notable trends include that more experienced attendings preferred to elicit patient preferences via conversation during the clinical encounter compared with using the questionnaire data, and they also seemed to be more concerned regarding time constraints during the visit relative to less experienced and younger physicians.

There were several physician, patient and technical-level barriers that were discussed during these interviews (Table 2). 11% (n = 1) reported being very concerned about increasing the length of the clinical note and how this data would be interpreted by insurance reviewers as well as any potential medical liability. 78% (n = 7) reported that they asked similar questions to the patient questionnaires on their own, and 33% (n = 3) reported that they still needed to confirm all the questionnaire data with the patient. Some reported patient-level barriers include that 78% (n = 7) of interviewed physicians reported concerns about low completion rates, especially for follow-up patient visits, thought to be due to creating frustration or a burden on their behalf. 33% (n = 3) reported that it takes a long time for the patients to complete the questionnaires, and 22% (n = 2) reported that patients rather discuss these concerns with their provider than complete the questionnaires. Other noted patient barriers include limited access to computers, misinterpretation of the questions, language barriers and that older patients may experience technical difficulties. Some noted organizational and technical barriers include that 67% (n = 6) of interviewed physicians reported they are limited because of time constraints and 33% (n = 3) reported difficulty finding the responses in the EHR. When asked about the impact of these PROs on the clinical visit flow, 44% (n = 4) reported no significant impact, 22% (n = 2) said that it saves time or allows for efficiency while 11% (n = 1) reported that it is burdensome on the provider and slows them down, and another 11% (n = 1) reported that it only impacts their clinical flow if it is a requirement for all patients to complete. In evaluating these barriers, it appeared that attending physicians with less experience tended to be the ones reporting that it saves them time, is more efficient, and doesn't affect their clinical visit flow compared with some of the physicians who have been out of training for a more prolonged time.

Table 2.

Summary of patient-reported outcome barriers.

Aspect	Barrier	Number of physicians that reported code
Physician	Increasing length of clinical note	1
	Concern regarding PRO data interpretation by insurance reviewers	1
	Concern regarding medical liability	1
	Ask similar questions to patient questionnaires	7
	Need to still confirm questionnaire data with the patient	3
Patient	Low completion rate	7
	Burdensome to complete	4
	Long time to complete the questionnaires	3
	Rather discuss concerns with provider than complete the questionnaires	2
	Limited access to computers	1
	Misinterpretation of questions	1
	Language barriers	1
	Technical difficulties	2
Organizational	Time constraint	6
Technical	Difficulty finding PRO data in EHR	3

EHR: Electronic health record; PRO: Patient-reported outcome.

Several suggestions for improvements were offered. In terms of patient-level factors, 33% (n = 3) suggested reminding patients to complete the questionnaires, incorporating the use of iPads to allow for completion at the time of visit, or visually depicting to patients their progress over the course of treatments. Suggestions for improvement of physician-level factors include additional training and allowing providers to view the patient interface. Suggestions for structural improvements include 33% (n = 3) of physicians recommending an improvement in the integration of the PROs with the EHR interface, 22% (n = 2) proposing tailoring the questions based on patients' previous responses, and 11% (n = 1) asking to include specific patient goals in the questionnaires. If patients have not completed the questionnaires, 44% (n = 4) of the interviewed physicians expressed to the patients the importance of completing it in treating and understanding their pain and encouraged them to do so, while 33% (n = 3) of the physicians reported providing no specific intervention, statement or guidance to their patients. Specific quotes from the interviews for the key questions are included in Appendix II. A word cloud from the interview analysis is shown in Appendix III.

Discussion

The primary themes of the analysis include physician utilization of PRO data and the evaluation of barriers and facilitators to effective data usage. Despite the theoretical benefits and availability of real-time PRO data, there have been significant barriers to consistent usage and variability in terms of acceptance. Only a minority of the chronic pain management physicians at WCM seem to use it routinely, even less regularly documenting it in their clinical notes and reviewing with patients at point-of-care, and there are a number of challenges and barriers that were highlighted.

In terms of baseline practice patterns, several interesting findings were apparent. While seven physicians reported relying on pain scores for evaluation of pain severity, only two physicians reported documenting and tracking these pain scores consistently. This points to the idea that pain management physicians often rely on subjective pain scores to evaluate chronic pain patients; however, documentation around these pain variables and tracking over time is quite inconsistent and may benefit from a more standardized and objective approach.

Interestingly, there seems to be a significant difference in views and uptake across the group relative to academic rank and years out of training. It is not surprising that physicians with a higher academic rank and increased years in clinical practice tended to report that the implementation of these PROs did not significantly change their practice relative to those with a lower academic rank and fewer years of experience. It has been well established that it is historically challenging to change physician behavior, which may be even stronger in those with many years of experience [23–25]. In addition, those physicians who were interviewed with fewer years in practice were also more likely to ask about psychiatric comorbidities and medications which may represent a cultural change in education around pain management and the more recent recognition of the importance of psychiatric comorbidities on the pain experience. Older and more experienced physicians also seemed to prefer to elicit patient preferences via conversation during the clinical encounter compared with using the questionnaire data, and they also seem to be more concerned regarding time constraints during the visit relative to less experienced and younger physicians, who were more likely to report it improved efficiency, which may reflect the younger generation's increased comfort with technology.

Some of the noted facilitators and barriers of use are similar to what has previously been described in the literature. Specifically barriers of burdening patients and physicians, time constraints for the clinical encounter, technical barriers including integration with EHR, and patient barriers including age, computer literacy, access to the internet and language barriers have been previously described [10,12–14,16,17]. Interestingly, the issue of medicolegal liability, with concerns about charges of clinical negligence and litigation if they did not address electronic PROs was also discussed in prior studies as well [17].

This is a pilot study used to potentially inform a national study. The limitations should be considered within the context of this study. We describe the experience at a single academic institution from the viewpoint of the physicians. The results seem to be plausible and valid as data saturation was reached at seven interviews and two additional interviews were conducted. While the small sample size in this study limits definitive conclusions based on sociodemographic characteristics, it is interesting to note a few trends that should be explored in further research. Additionally, the patient or ancillary staff perspectives are not captured in the study as that was outside the scope of this targeted study. These factors may limit the generalizability of these findings, but this pilot study suggests the need for further exploration and research in this area. It is a particularly relevant and timely issue as the burden of chronic pain in the USA and the opioid epidemic continue to grow at alarming rates and using PROs can help guide clinical decision-making regarding when and how to use opioids most effectively and safely.

Conclusion

It is widely accepted that measuring pain severity in and of itself is insufficient and that physicians also need to assess the impact of chronic pain on a patient's overall level of functioning and health-related quality of life. PROs can be especially useful in objectively measuring these domains and guiding clinical decision-making; however, multiple patient, technical and physician-level barriers still exist to the optimal use of PRO data in the routine care of chronic pain patients. To optimize its impact, further research is needed to streamline the process for meaningful and efficient PRO data capture, as well as define best practices for clinician use in terms of review with patients and documentation of the PRO data in clinical notes to improve individualized patient care, treatment outcomes, the patient experience and the impact of chronic pain on a population level.

Funding Statement

A grant from The Applebaum Foundation (Grant Number: 61500265) awarded to LR Witkin supported the study.

Author contributions

LR Witkin, SY Jiang and R Gotian conceptualized the study and developed the study methodology. All authors conducted participant data collection. LR Witkin and A Kasubhai contributed to the analysis and interpretation of the data. LR Witkin prepared the manuscript. All authors critically edited, reviewed, and approved the final draft of the submitted manuscript.

Financial disclosure

A grant from The Applebaum Foundation (Grant Number: 61500265) awarded to LR Witkin supported the study. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Competing interests disclosure

The authors have no competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Writing disclosure

No writing assistance was utilized in the production of this manuscript.

Ethical conduct of research

The authors state that they have obtained appropriate institutional review board approval from Weill Cornell Medicine for the research described. In addition, they have obtained written informed consent from the participants involved.