Perspectives on applying immuno-autonomics to rheumatoid arthritis: results from an online rheumatologist survey

Rheumatologists were recruited from a panel (M3 Global Research) and a rheumatology care management organization (United Rheumatology), and were offered a small gratuity ($20–$40) for completing the survey. All survey contacts were made electronically via email message. Survey panel members were invited to participate directly via M3; rheumatologists recruited via the care management organization received a message stating, “United Rheumatology is helping to field a survey to learn more from rheumatologists about their treatment of rheumatoid arthritis (RA) patients. Please help us by taking this short 10-min survey by clicking the link below.” At the beginning of the survey, respondents were told that the purpose of the survey was to seek “feedback from physicians about their treatment of rheumatoid arthritis patients.” Study qualification requirements included: a primary medical specialty of rheumatology; primarily treat adult patients; been in clinical practice three or more years; and evaluate at least 15 RA patients per month.

Questions were organized into four main sections. The first collected demographic information and current practice management for RA patients. This included demographic and practice setting-related questions. It also included questions about how the interviewee’s RA patients currently respond to treatment and rheumatologists’ interest in new treatments for RA patients. The second section provided information on the potential role of stress in RA, including information on stress biology and findings from a patient survey indicating a large percentage of patients believe stress can trigger their RA. Survey respondents rated their level of agreement with four statements about stress and RA, and answered questions about how they currently assess and address stress in their practice. The third section provided information on the possible role of evaluation of the ANS for the management of RA patients, including the use of next-generation HRV as a tool to assess autonomic state. Pertinent references (about 40) were provided, as well as several paragraphs of information reviewing the available literature. Rheumatologists also were shown a model (see Fig. 1) and a narrative description of several studies showing an association between reduced HRV (indicative of increased sympathetic ANS activity), which in turn may be associated with increased inflammation and suboptimal therapy results. Next-generation HRV was presented to interviewees as an accepted tool to measure autonomic state. Survey respondents rated their level of agreement with statements about ANS and RA, and their interest in being able to quantitatively measure ANS using next-generation HRV. The final section asked respondents to consider their interest in a test to measure autonomic state using next-generation HRV and to indicate on which groups of RA patients they would consider using such a test. All materials shown to respondents are available in Supplement 1.

Survey data were reviewed to check for response errors or concerning response patterns (e.g., users who consistently selected the top answer category to each question, or users who completed in an extremely short amount of time). No invalid responses were noted. Only complete survey responses were used in the analysis.

Descriptive statistics were used to characterize trends in the data. Key survey variables were evaluated in a cross-tabulation analysis against demographic variables including years in practice, gender, race, practice characteristics, and practice setting. Chi-square analyses were used to identify statistically significant differences across respondent groups (e.g., male vs. female respondents). No weighting of survey data was performed.

The survey was sent to 1204 potential respondents (see Fig. 2). A total of 350 respondents began the survey (participation rate = 29%); 55 were excluded due to ineligibility, resulting in an eligibility rate of 84% (295). (Among those who were ineligible: n = 13 declined the conditions for taking the survey; n = 20 had been in practice less than 2 years; n = 13 primarily saw pediatric patients; n = 6 had a specialty other than rheumatology; and n = 3 were screened out for other reasons.) Among eligible respondents, the majority (231/295; 78%) completed the survey. On average, the survey took 11 min to complete.

As shown in Table 1, respondents are primarily male (66%) and white (60%). Most (63%) have been in practice 11 or more years, and more than half (60%) see 81 or more adult RA patients per month. About a fifth (17%) practice in an academic setting. These results are similar to the demographics reported in the 2015 ACR Workforce Study [16], which found that 59.2% of rheumatologists are male, and 20% practice in an academic setting. Rheumatologists are actively engaged in continuing education with 74% saying they participate in multiple CME activities per year and 89% subscribing to rheumatology journals.

Rheumatologists were asked to classify their RA patients according to their current disease activity. A mean of 34% (median = 30%) of patients are categorized as having moderate to high disease activity (range from 0 to 86%). Rheumatologists also were asked to classify for what percentage of their RA patients is it difficult to achieve low disease activity or remission. Their answers range from 2 to 95%, with a mean of 28% (median = 25%) (Table 2).

Rheumatologists express interest in new tools or tests to predict treatment outcomes for RA patients, as well as ways to better understand disease pathways (Table 2). Most respondents (89%) agree there is a need for new tools or tests to assess why some RA patients do not respond to conventional, biologic, or targeted synthetic DMARDs. Likewise, 90% agree they would be interested in a new tool or test to identify RA patients who are less likely to respond to treatment. More than 9 in 10 (92%) also agree they would like to better understand disease pathways that may predict treatment outcomes for RA patients.

Rheumatologists accept stress as playing an important role in disease activity and treatment for RA patients (see Table 3). Most (84%) agree that increased patient stress is related to higher disease activity, and 87% agree stress can make RA patients less likely to respond to treatment. Almost 9 in 10 (89%) agree they would like to know more about the role of stress in RA treatment, and 85% would like to be able to measure the effects of stress and the biology underlying stress. Most rheumatologists (90%) report that they currently assess patient stress in their RA patients, primarily via patient conversation (87%). However, only about a third (32%) always or often treat patient stress in their RA patients. Common treatment approaches include lifestyle adjustments such as diet or exercise (67%), in-office counseling (50%), and medications (45%).

A few group differences are noted. Those in an academic setting are more likely to report they use patient-reported outcomes to assess stress (31% vs. 20%, p < 0.05) as are those who attend conferences multiple times per year vs. less often (28% vs. 16–23%, p < 0.05). Rheumatologists in practice from three to five years are less likely to assess patient stress always or often versus those in practice longer (27% vs. 37–39%, p < 0.05). Similarly, stress is assessed more frequently among those who identify themselves as early adopters (39% vs. 28%, p < 0.05) or who attend multiple conferences per year (43% vs. 26–35%, p < 0.05).

Rheumatologists were presented with information on the role of the ANS, including five-minute HRV as an established metric of ANS state, which can be measured quantitatively [11]. With the information provided, respondents were asked to assess whether they agree or disagree with various statements about ANS, HRV, and RA (see Fig. 3). There is a strong level of agreement with all statements given, indicating a belief among rheumatologists that ANS state influences RA presentation and treatment. At least 8 in 10 respondents agree with the majority of the statements, including that: “Autoimmune diseases such as RA may be influenced by ANS state” (89%), “ANS state may interfere with disease control in RA patients” (86%), “Knowing ANS state could be helpful to rheumatologists to better treat their patients” (86%), “I would like to be able to easily and accurately measure ANS state in RA patients” (81%), “I am interested in identifying RA patients with poor autonomic function” (81%), and “Being able to quantitatively measure autonomic state (stress biology) could be useful to me as a rheumatologist in the care of my patients” (80%). In addition, about three-quarters agreed that, “Patients with poor autonomic function may be at risk for not responding adequately to conventional, biologic, or targeted synthetic DMARDs” (76%), and “I would like to be able to quantitatively measure autonomic performance, including sympathetic and parasympathetic activity” (75%).

Several group differences are noted, especially among self-described early adopters. Early adopters (89% vs. 80%, p < 0.05) and those in practice more than 20 years (91% vs. 81–86%, p < 0.05) are more likely to believe ANS state may interfere with disease control. Early adopters also are more likely to think autoimmune diseases are influenced by ANS state (92% vs. 82%, p < 0.05). Those in practice 20 or more years are more likely to think knowing ANS state could be helpful to their practice (91% vs. 82–84%, p < 0.05). Early adopters are more interested in being able to quantitatively measure autonomic performance (78% vs. 67%, p < 0.05) and are more likely to see being able to measure ANS state as useful to their practice (83% vs. 74%, p < 0.05). Early adopters (81% vs. 64%, p < 0.05) and rheumatologists in a multispecialty practice (82% vs. 71–73%, p < 0.05) are more likely to think patients with poor ANS function are at risk of not responding adequately to treatment.

Rheumatologists were asked to consider how interested they would be in an in-office test to measure ANS function. Most rheumatologists are interested in such a test, with 84% saying they are extremely, very, or moderately interested in such a test. Interest is higher among self-identified early adopters and among those who go to multiple conferences per year.

Rheumatologists also were asked to consider whether they would use such a test and for which patient groups they would consider using such a test (Fig. 4). Almost all (94%) indicate they would use such a test. Most commonly, rheumatologists say they would use such a test on patients for whom it is difficult to reach remission or low disease activity (73%), as well as for patients with loss of response to conventional, biologic, or targeted synthetic DMARDs (73%). About two-thirds (65%) also would consider such a test on patients for whom they are considering advancing therapy, and about half (48%) would consider using a test on newly diagnosed patients. In addition, 4 in 10 (43%) would use a test on patients for whom they are considering tapering therapy.