Background

The average delay in diagnosis for patients with axial spondylarthritis (axSpA) [1,2,3] ranges from 7 to 10 years [4,5,6,7,8], with estimates as high as 13 years between symptom onset and diagnosis in the United States [9]. The significant diagnostic delay is a well-known feature of spondyloarthritis management [10]. Patients with longer delay in diagnosis (e.g., > 5 years) display more structural damages and limited spinal mobility [11]. In addition, patients may experience distress, depression, and desperation associated with their prolonged search for diagnosis and treatment [12]. Therefore, the improvement of delay in diagnosis for axSpA may slow disease progression and thus avoid or delay serious disability due to the early treatment interventions and access to care [13].

The delay in diagnosis for axSpA is multifactorial including characteristics of the patients, providers, healthcare system, and disease itself. For instance, patients may seek care from different medical professionals without guidance due to varied disease manifestations of axSpA or the lack of proper access to care (e.g., limited rheumatologists). Despite the use of sensitive imaging tool such as magnetic resonance imaging, the delayed diagnosis for patients with non-radiographic axSpA is still significant [14]. On the other hand, physicians may also play an important role in the early diagnosis of axSpA. Despite the low awareness of axial SpA among non-rheumatologist physicians [13], most axSpA patients are diagnosed by non-rheumatologists [15], with rheumatologists having diagnosed only 37% between 2000 and 2012 [16]. Primary care physicians may have difficulty discriminating inflammatory back pain from other types of back pain and may be unaware of other features that are important in making a diagnosis of axSpA.

In this qualitative research study, we sought the perspectives of primary care physicians to gain a better understanding of the reasons for diagnostic delay in primary care settings. We chose a qualitative approach because qualitative data frequently yield surprising insights and provide in-depth detail about participants’ perceptions and experiences that quantitative data often cannot [17]. Such foundational knowledge may be useful in developing interventions to reduce the delay in diagnosis of axSpA.

Methods

The University of Massachusetts Medical School Institutional Review Board approved this study.

Study design

The physician in-depth interviews were part of a larger qualitative study- The SpondyloArthritis Screening and Early Detection (SpA-SED) Study. Our protocol was guided by best practices for the conduct and reporting of qualitative research (COREQ) [18]. The appropriately completed COREQ checklist for the current study is included (Appendix A).

Participant recruitment

We recruited 18 primary care physicians (a sample size likely adequate to achieve saturation) who were willing to participate in a recorded, 60-min in-depth discussion. We used purposive sampling to enroll equal numbers of men and women and family medicine and internal medicine physicians. Thirty-four participants including those who were known to the research team or their colleagues (approached by emails) or who were identified through state and regional primary care professional societies (face-to-face) were invited to participate. Six declined, nine were non-responders, and one physician was not selected because of our need for balance by gender and type of physician. All 18 participants participated in scheduled interviews and completed the study. Participants were offered a cash card of $300 to compensate for their time.

Study setting

The interviews were conducted either in person in a private conference room in Rhode Island or Massachusetts (n = 3) or over the phone (n = 15) between February–May 2019. We conducted phone interviews using Zoom.

Interview guide

A multidisciplinary team developed the interview guide (< 60 min interviews, Appendix B). Questions included experience with back pain, how back pain is evaluated, what laboratory tests are ordered when axSpA is suspected, what referrals participants would make, awareness of axSpA, speculation about what contributes to diagnostic delay in axSpA, etc. The average length of the interviews was 47.1 min (standard deviation: 11.3).

Conduct of interviews

Interviews were conducted by experienced, trained personnel (KLL, DS), with an observer from the research team participating and taking notes. For one interview, the research interviewer (KLL) had previously worked as an investigator on a research project with the physician participant (> 5 years previously). Research staff were trained to assure standardized data collection. The protocol was pilot tested with one physician. The researchers conducting the interviews reported no obvious bias but did report prior assumptions that lack of awareness of axSpA and lack of time would emerge as leading barriers to axSpA screening in primary care settings. No repeat interviews were conducted.

Data collection and management

The study data were collected and managed using REDCap electronic data capture tools hosted at the University of Massachusetts Medical School [19]. REDCap (Research Electronic Data Capture) is a secure, web-based application designed to support data capture for research studies, providing: 1) an intuitive interface for validated data entry; 2) audit trails for tracking data manipulation and export procedures; 3) automated export procedures for seamless data downloads to common statistical packages; and 4) procedures for importing data from external sources.

Analytic strategy

We conducted a group method of data analysis known as immersion / crystallization [20]. Four team members independently listened to selected interview recordings, read in-depth transcripts, and wrote analytic notes for each. The transcript texts were subjected to line-by-line coding with NVivo qualitative software [21]. The codebook was modified by team consensus. Searches for alternative interpretations were conducted again and discussed before final decisions were made about how to report the findings of the study. This manuscript focused on one parent node in our coding scheme: Reasons for Diagnostic delay.

Feedback from participants

We prepared a preliminary report of our findings and emailed participants who indicated they were willing to review and provide feedback (Appendix C).

Results

Of the participants, 44% were women, 66.7% were non-Hispanic white and 83.3% attended an allopathic medical school in the United States (Table 1). All physicians considered that the length of diagnostic delay was problematic, and all agreed that this delay was unacceptable. Physicians believed that multiple factors were contributing to the diagnostic delay in axSpA. Figure 1 depicts the multifactorial components perceived by primary care providers participating in the study. Barriers to early diagnosis included disease characteristics (e.g., slow disease progression; Table 2), patient factors (e.g., multiple other conditions, back pain not chief complaint, Table 3), physician characteristics (e.g., lack of rapport/trust, Table 4), and structural/system issues (e.g., lack of time, Table 5). Primary care physicians noted that, because back pain is multifactorial and very common and axSpA is not, axSpA may be missed. Physicians explained that the slow progression of the disease may contribute to the diagnostic delay. The intermittent nature of symptoms and range of symptoms severity related to axSpA also was noted as a factor contributing to diagnostic delay. Additionally, physicians pointed out that it takes time for radiographic evidence to appear and that there are no other definitive diagnostic tests.

Table 1 Characteristics of physicians
Full size table
Fig. 1
figure 1

Factors influencing timely diagnosis of axSpA

Full size image
Table 2 Disease characteristics contributing to diagnostic delay for patients with axSpA
Full size table
Table 3 Patient characteristics contributing to diagnostic delay for patients with axSpA
Full size table
Table 4 Physician characteristics that may reduce diagnostic delay for patients with axSpA
Full size table
Table 5 System/structural characteristics contributing to diagnostic delay for patients with axSpA
Full size table

A common theme was that patients do not interact frequently enough with the health care system and the lack of continuity of care (Table 3). Primary care physicians uniformly noted that patients often have multiple issues to discuss and back pain is mentioned only at the very end of the visit. Physicians did not discuss why patients waited until the last moment to share their complaints of back pain. Some physicians explained that they are likely to attribute back pain to other conditions (Table 4); most emphasized the need to establish trust and a good rapport with patients. The importance of setting appropriate expectations with patients was noted, since back pain may not be addressed adequately in one visit. Primary care physicians lamented that the limited amount of time available for appointments impedes their ability to diagnose axSpA in a timely fashion (Table 5). Further, physicians noted that the lack of continuity of care may contribute to diagnostic delay. Patient concerns about the expense of copayments expected when undergoing testing needed to evaluate axSpA may also contribute to the diagnostic delay. The lack of dedicated spine clinics and the shortage of rheumatologists were reported as barriers to timely diagnosis.

Discussion

This study sheds light on the challenges faced by primary care physicians that may contribute to the delay in diagnosing axSpA, which typically presents with the insidious onset of back pain that often is attributed to other, more prevalent conditions [22, 23]. The subjective quality of pain may be difficult for patients to describe and is not easily measured objectively. In our study, primary care physicians noted that axSpA patients may not interact frequently enough with the health care system to facilitate a correct diagnosis. Consistent with previous research [16], lack of diagnostic criteria, lack of definitive biomarkers, and lack of radiographic confirmation of axSpA until later in the disease course [13, 24] were identified as factors contributing to axSpA diagnostic delay.

Several physician-related factors that may contribute to diagnostic delay were identified, including a lack of awareness about axSpA among non-rheumatology health care professionals [9]. Lack of trust, poor rapport, ineffective communication, and misaligned expectations between providers and patients were factors that likely contribute to diagnostic delay. Among patients with immune-mediated inflammatory diseases, distrust of health care providers has been associated with poor medication adherence [25] while, conversely, good patient-provider communication has been shown to improve health outcomes [26]. Our participants also reported that patients may not realize that axSpA often is not diagnosed during a single visit, but rather involves a journey that occurs over a period of time. Physicians mentioned that infrequent contact with the health care system and lack of continuity of care contributed to the diagnostic delay. To our knowledge, research has not been conducted on whether discordant expectations between patients being evaluated for axSpA and their primary care physicians contributes to diagnostic delay. Nevertheless, improved alignment of patient expectations to care processes may improve patient continuity of care.

Primary care physicians mentioned system-level factors that impede timely diagnosis of axSpA, including the short appointment length and difficulty arranging rheumatology referrals. The median length of a primary care visit was 15.7 min, during which a median of six topics were covered with ~ 5 min spent discussing the primary complaint and 1.1 min on each remaining topic [27]. A musculoskeletal or rheumatologic condition accounted for the chief complaint in only about 8.3% of primary care visits [28], which is consistent with reports by providers in our study that mention of back pain may be withheld by patients until the end of the clinical encounter. Patients with axSpA often have multiple conditions, such as hypertension and depression, which might supersede back pain as the main focus of a primary care visit [29]. Given the challenges of reduce time with patients and increased complexity of patient issues, automated tools may be useful to primary care physicians to identify uncommon diseases such as axSpA.

Delays in diagnosis of axSpA also have been attributed to late referral of patients with inflammatory back pain by general practitioners to rheumatologists [15, 16, 30, 31]. However, the average wait for a rheumatology appointment is 4 months [32]. Difficulty accessing rheumatology care is exacerbated by the shortage of and decline in the number of practicing rheumatologists [33]. By 2025, the demand for rheumatologists is projected to exceed supply and a shortage of 3845 rheumatologists is predicted [34, 35]. This shortage results in burdens for patients, including excessive travel time – which may exceed 90 min for some patients [36]. The American College of Rheumatology has proposed multiple strategies to address this workforce shortage, but the extent to which these will be successful remains uncertain.

Limitations must be considered. Most of the participants practiced in Massachusetts and Rhode Island and had academic affiliations. Systems factors may be specific to local context. Physicians with academic affiliations may be more aware of axSpA. Patient factors were reported by the physician participants and are derived from experiences with their own patients. It is possible that physicians only had experience with patients with ankylosing spondylitis. Misinterpretations of direct quotes may have occurred, although the analytic approach used reduced this likelihood. However, despite these potential limitations, our findings appear to be aligned with previous research [16].

Conclusions

Aiming to provide a better understanding of the reasons for diagnostic delay in primary care settings, this study identified primary care physicians’ perceptions of several root causes of diagnostic delay in axSpA. These contributing factors of delay in diagnosis for axSpA were believed to be multifactorial and were attributed to characteristics of patients, providers, the healthcare system, and the disease itself. While in our study primary care physicians felt that the diagnostic delay reported in the literature is lengthy, it must be noted that Primary care physicians believed that strategies to address these barriers and reduce delay in diagnosis of axSpA are needed. Early diagnosis of axSpA is difficult and as such solutions to shorten the diagnostic journey must consider the multitude of contributing factors (i.e., patient, disease, physician, system).