Patient perspectives on health care provider practices leading to an axial spondyloarthritis diagnosis: an exploratory qualitative research study

We performed an exploratory qualitative research study in accordance with best practices [21], in which we conducted six focus groups. Our target population is US adults who have a rheumatologist confirmed diagnosis of axSpA. Also, we utilized purposeful sampling to target a variety of times from when patients first told their physician about symptoms to diagnosis ranging from early (< 2 years) mid (3-7 years) and late (> 7 years). We collaborated with three rheumatologists practicing in Worcester MA, Philadelphia PA, and Aurora CO. Locations were selected to provide patients from different regions in the United States where Rheumatologists were contracted. Rheumatologists followed our recruitment guidelines by alerting potentially qualifying patients. After an opt out period, the rheumatologists provided contact information to UMMS for those patients who were interested in participating. Eligible patients aged ≥18 years had a clinical diagnosis of axSpA verified by the referring rheumatologist and were able to provide informed consent (verified by the referring rheumatologist). People were excluded if they were unable to participate in a two-hour focus group, unwilling to be audio recorded, or did not speak English. One individual was no longer interested and withdrew from the study. Figure 1 depicts how patients were recruited to the study to participate in either an in-depth interview or focus group. A total of 57 patients from Worcester, MA, were contacted. Of those, 22 patients were screened and enrolled. Seventeen patients participated in the study (either focus group or in-depth interview) and 15 patients participated in the Worcester, MA focus group sessions. Twenty-one patients from Philadelphia, PA were contacted and 11 patients were screened and enrolled. Seven patients were included in the study and 3 participated in the focus group session. Twenty-two patients from Aurora, CO were contacted to participate in our study and 16 patients were screened and enrolled Of those 11 patients were included in the study and 8 patients participated in focus groups. The UMMS team confirmed subject eligibility and scheduled participants to participate in one of six focus groups from March 2019 to June 2019: three in in Worcester, MA (March 2019), two in Aurora, CO (April 2019), and one in Philadelphia, PA (June 2019). In Worcester, MA, the first focus group contained 7 participants. The second and third focus group sessions contained 4 participants each. In Aurora, CO, the first focus group interviewed 5 participants and the second focus group contained 3 participants. In Philadelphia, PA, 3 participants participated in the focus group.

A multidisciplinary team developed the focus group guides based on a scoping literature review [22‐30], input from members of the sponsor’s research and two rounds of clinical review by the collaborating rheumatologists [31]. Focus group topics included: early symptoms, diagnostic journey, and advice for primary care doctors on how to reduce axSpA diagnostic delay (Table 5 in Appendix 1). CD, AB and KF had contact with participants during recruitment. Two members of the UMMS research team were present at each focus group, with one author (CD - Associate Professor (female)) who teaches graduate-level qualitative method courses moderating the session. A research assistant set up and tested dual audio recording devices, placed table signs that listed the main topics, handled collection of forms, surveys, and distributed $100 cash cards as (compensation for time and travel). A light meal was provided. At the beginning of the focus group, CD introduced herself and her role. A fact sheet detailing the rights and responsibilities of research participants was reviewed before beginning the discussion. The guideline incorporated semi-structured and open-ended questions to elicit relevant information. At the close of the focus group, participants completed a short background survey to determine if they wanted the opportunity to provide feedback on the draft report and/or would like to receive copies of manuscripts emanating from this research.

Audio recordings were uploaded to a secure UMMS server. CD selected the recording with the best audio quality and uploaded it via secure ShareFile to a HIPAA compliant transcription service. Verbatim transcripts were reviewed for accuracy, errors were corrected, names were replaced with generic references (“my doctor” instead of “Dr. X”), un-identified speakers were labeled, and (if possible) indistinct comments were clarified. Research assistants replaced participant names with ID numbers and then imported the transcripts into NVivo 12 [32]. (QSR International https://​www.​qsrinternational​.​com/​nvivo/​nvivo-products). One author (CD) developed a coding structure based on the focus group topics and her initial impressions of focus group content.

We conducted a thematic analysis of qualitative data [33]. A thematic analysis was conducted using a coding start-list which was derived from the focus group protocol [33]. New codes were also added as they emerged during coding. Overall, an iterative thematic analysis was applied [33]. First, research assistants read all focus group transcripts, listened to all focus group recordings at least once, and wrote a summary. Two members of the research team independently then coded each focus group transcript into one NVivo 12 project. Code reports were generated for each thematic code. Two team members independently prepared summaries of impressions for each code, tagging quotations reflective of the summaries. Impressions were discussed for each code and sub-code and consensus compilations were finally crafted to consolidate the duplicate code summaries and original code reports. Compilations and summaries served as source documents. This paper focuses on those codes relevant to the diagnostic experience. A two-page preliminary report of our findings was emailed on August 7, 2019, to those participants who had requested it.

The six focus groups included 26 patients with axSpA (range: n = 3-7 participants; average length 108 min (range: 80 to 128 min). Most participants were men (n = 16) (Table 1). The average age of participants was 53.5 years (range: 21 to 76 years old), the average age at first symptom was 22.8 years (range: 10 to 55 years old) the average age at which the patient first told their primary care physician about symptoms was 24.4 years (range: 10 to 55 years old) and the average age at diagnosis was 34.8 years (range: 15 to 65 years old) Most participants were non-Hispanic White and 62% had a college degree or higher. Almost all (> 90%) wanted the opportunity to give feedback on preliminary findings or publications emanating from the research.

Common themes included the lengthy trial and error approaches that led to misdiagnosis and initiation of failed therapies; dismissal of intermittent symptoms by healthcare providers; and the need for patients to research their own diagnosis because providers stopped trying to make an accurate diagnosis (Table 2). Although some participants experienced a timely diagnosis, many others described long circuitous journeys to search for a diagnosis, which involved seeing primary care and specialist physicians, chiropractors, and physical therapists. Patients estimated time from first symptoms to being diagnosed averaging 12 years (range 0-37 years) and time from their first mentioning symptoms to a physician to being diagnosed averaging 10 years (range 0-30 years). Some participants had been prescribed therapeutic trials of many different medications for symptom relief, in the absence of a clear diagnosis.

Participants appreciated that their early symptoms could have been attributed to a wide variety of other conditions, making it difficult for physicians to diagnose them. Yet, participants wanted physicians to reach a diagnosis as rapidly as possible so that they could begin appropriate treatment. The slow, progressive nature of axSpA, with symptoms that could not readily be attributed to an antecedent event, compelled patients to impress upon their physicians that something was genuinely wrong. To emphasize the seriousness of their illness experience, patients often needed to use dramatic language to convey the urgency and legitimacy of their symptoms (e.g., pain level of 10 as “suicidal” (46 year old male); assigning a 1-10 pain scale rating for a muscle spasm “25… it’s horrible” (54 year old male)).

The waxing and waning patterns of pain resulted in some patients delaying seeking medical care and confused some physicians. Participants felt strongly that primary care physicians should not stop trying to establish a diagnosis, regardless of how challenging it might be. In fact, when primary care physicians abandoned their diagnostic quest, patients experienced this as being profoundly negative. During the typically lengthy time that participants waited to be diagnosed, participants experienced frustration and mental suffering. Many participants left with several of their questions not being answered and/or not being believed or taken seriously. Participants were eventually diagnosed with axSpA, but for many of them, this was the result of their tenacity and/or that of their family members to determine the source of their suffering. Some participants conducted their own research and were confident enough to challenge their physicians. However, some resented having had to resort to researching their symptoms on their own.

Participants discussed their early experiences from when their symptoms emerged, and they started seeking explanations and relief from their primary care physicians (Table 3). A recurring theme was that participants felt that they were neither “heard” nor believed. Participants described a variety of negative experiences ranging from primary care physicians minimizing or dismissing their complaints to primary care physicians concluding that their symptoms were imaginary. Participants’ advice to primary care physicians was straight-forward: allow patients to explain their symptoms, listen to them, and believe them. Further, participants highly valued clinicians who were there for them and persistent, who partnered with them in the diagnostic quest, and who followed through to make a correct diagnosis. Beyond listening to and believing the patient, some participants stressed the importance of primary care physicians gaining a complete picture of the patient’s experience with their illness.

Participants articulated the desire for a definitive diagnostic test. Most reported that they had been tested for the HLA-B27 antigen and were positive. Participants would have preferred to have received HLA-B27 testing earlier in the diagnostic process. Although many understood that this test did not establish a diagnosis of axSpA definitively, it often confirmed their final diagnosis.

Participants suggested that primary care physicians would benefit from additional education about axSpA (Table 4). Many participants reported that their primary care physicians admitted to them that they did not know what might be causing their symptoms. Conversely, some participants described primary care physicians who kept researching their symptoms and educating themselves to figure out what was happening to their patient. Some participants stressed the importance of a partnership between the provider and the patient but acknowledged that this was difficult to achieve during appointments of short duration. Some participants spoke of physicians who assumed that they were exhibiting drug seeking behavior, hypochondriasis, or psychosomatic illness. Since axSpA patients are often young and otherwise healthy, physical findings might not have been present at the time of their clinical encounter.

Many participants lamented the long wait time to see a specialist. Patients thought that referral to a specialist could accelerate the time to diagnosis but perceived that the shortage of rheumatologists contributed to these delays. Patients often described their visits with a rheumatologist, once they finally took place, as having been short and hurried. However, in at least one case, coordination of care among treating physicians, including a rheumatologist, resulted in effective diagnosis and treatment.

Overall, participants conveyed the feeling that they had been widely misunderstood by many physicians, which for many resulted in a lengthy delay in being diagnosed and receiving effective treatment. Nearly all participants described experiences in which they had to advocate for themselves and/or rely on the advocacy of family members to accelerate the diagnostic process. This often took the form of “chasing a diagnosis” over the course of years. Some felt that they had to diagnose themselves, receiving confirmation only when a doctor agreed to proceed with HLA-B27 testing or referral to a rheumatologist.

“Doctors, I mean, they care, but they don't. Like, if you don't just, you know, get to the point, be there for yourself, like, advocate for yourself, they're just going to then walk out the door and say nothing's wrong.” 28 year old female

Qualitative research provides insights that cannot be obtained by typical survey research or analyses of claims or electronic health records. We designed our guidelines to be aligned with best practices and we recruited participants from three geographic locations. All participants had received a diagnosis of axSpA, but the time since diagnosis was varied and sometimes was quite long. Information regarding the diagnostic journey was based on patient recall and the extent to which these experiences are reflective of patients without a confirmed diagnosis of axSpA is unknown. Most of the participants were highly educated and all spoke English. The experiences herein may not reflect those of patients with limited English proficiency or with lower levels of educational attainment. Finally, patients agreeing to participate in a focus group may have more confidence and ability to speak up and express their feelings and opinions. Patients who are more insecure or reserved might decline participation in a focus group and thus their views would not be represented.