In this study, we analyzed occurrences of health events (ie, diagnosis codes and types of healthcare provider seen by patients) using the MarketScan research databases prior to the diagnosis of PsA to better comprehend a patient’s journey in obtaining a PsA diagnosis. Overall, nonspecific musculoskeletal manifestations and non-PsA diagnoses were common up to 6 years before the diagnosis of PsA compared with controls without PsA. The diagnoses and codes entered prior to PsA diagnosis varied by provider type. Rheumatologists, general practitioners, and dermatologists accounted for the diagnoses of two-thirds of patients with PsA.
Timely referral of a patient with potential PsA for rheumatology evaluation is critical to improving outcomes for patients with PsA, and this been recognized for some time [9, 12, 13]. The ability to identify PsA symptoms early in the disease course, in addition to recognizing specific trends and patterns in a patient’s pursuit for diagnosis, may permit prompt referral to appropriate specialists, leading to an earlier diagnosis of PsA. These health events may serve as potential predictors of a PsA diagnosis with regard to the many diagnoses received and healthcare providers seen. Identifying the subset of the patient population for screening and referral may be additionally useful to avoid overwhelming rheumatology healthcare providers with high volumes of patients with musculoskeletal symptoms.
Compared with controls, patients with PsA in our study received a variety of additional non-PsA diagnosis codes, suggestive of PsA symptom build-up, in the 6 years preceding diagnosis. An analysis of administrative data in Canada similarly reported a build-up of musculoskeletal symptoms as early as 5 years prior to the diagnosis of PsA [14]. Patients manifesting these symptoms have well-documented difficulties (“unable to do” or “with much difficulty”) with various daily tasks, including the inability to bend down to pick up items from the floor or to walk on flat ground [15]. These, in combination with other negative psychological, social, and economic impacts, contribute significantly to the overall burden of disease and poor quality of life [16–19].
In addition, an abrupt increase in the number of psoriasis diagnoses was observed in our cohort preceding the diagnosis of PsA. Although more than two-thirds of patients with PsA develop dermatologic lesions prior to joint and musculoskeletal manifestations, the symptoms are not often recognized as related to each other [15, 20, 21]. A study comprising 5604 patients with psoriasis or PsA from the National Psoriasis Foundation survey panels reported that almost 70% of patients sought medical help from 1 to 2 healthcare providers and 25% saw ≥ 3 providers within 2 years prior to the survey before receiving a diagnosis of psoriasis [22]. Because the diagnosis journey for psoriasis may portend a PsA diagnosis [23, 24], dermatologists are in a prime position to detect PsA early and rapidly refer patients to rheumatology. Other studies have reported the presence of subclinical inflammation in the joints and entheses of patients with psoriasis without a diagnosis of PsA [25, 26]. Subclinical PsA may be regarded as a stage of the progression of psoriatic disease, characterized by joint inflammation and/or morphological changes detectable by imaging modalities; up to 50% of all patients with psoriasis may have symptoms suggestive of but not yet diagnosable as PsA [15, 20, 21, 27]. Undiagnosed PsA is commonly observed in dermatology offices [28, 29] and has been recognized in almost one-third of patients with psoriasis [24, 29–32]. Patients with psoriasis presenting with PsA manifestations, comprising arthritis, enthesitis, dactylitis, nail disease, and spondylitis, should be evaluated for PsA [33].
Collectively, our study may indicate an opportunity for earlier detection of PsA as characterized by nonspecific musculoskeletal manifestations and skin symptoms and suggests that these musculoskeletal manifestations may be undercoded or underrecognized in clinical practice [27]; improved outcomes have been observed among patients who were promptly treated after receiving a diagnosis of PsA [12, 13, 34]. In creating algorithms using electronic medical records or administrative (ie, insurance) databases to identify patients who may have early PsA, data scientists need to consider that the diagnosis pathways may differ from patient to patient, depending on the sequence of provider specialties consulted by the patient.
Approaching PsA diagnosis, general practice, rheumatology, and dermatology visits were much more frequent, suggesting the need for patient education in recognizing symptoms of arthritic conditions. A delay in diagnosis due to initial medical consults with general practitioners and other nonrheumatology healthcare providers is well recognized [8, 15, 17]. An integrated, multidisciplinary method is critical for comprehensive medical care for patients with PsA. Rheumatologists and dermatologists are in an ideal position to mediate collaborations with other healthcare providers and patient advocacy and support groups for the design of a patient-centric intervention and treatment strategy [35, 36]. An improvement in skin and joint symptoms was observed in more than half of patients with psoriasis and PsA after treatment modifications facilitated by rheumatologists and dermatologists [35, 37]. However, as the number of rheumatologists is predicted to decline by 30% by the year 2030 [38], more innovative approaches to patient care are needed. Integrating the use of technology, such as software that calculates composite targets based on both physician- and patient-reported outcomes to determine disease activity status may be implemented [3]. Validated screening instruments may also be used [39–41], along with patient education and referral guidelines for general practitioners [36, 42]. The future development of applications that tabulate risk or likelihood of disease based on patient data may be a useful undertaking. A more patient-centric disease management method may be implemented with the use of currently available diagnostic tools, such as electronic applications allowing the upload of patient-reported outcomes prior to doctor visits [3]. This may help increase patient awareness of symptoms, enhance self-management, and improve patient-provider communication; the healthcare provider would enter notes from the visit into the same software.
There are limitations with our study that should be considered when interpreting the findings. As with any analysis of claims-based data, miscoding by healthcare providers may have occurred; however, the inclusion of patients with PsA who had received 2 diagnosis codes, instead of only 1 diagnosis code, may decrease the likelihood of miscoding or misdiagnosis. Additionally, some healthcare services may not be included in claims databases, leading to the underestimation of overall disease and economic burden. ICD-9-CM and ICD-10-CM codes are used for billing and may not reflect an accurate diagnosis, and there was no independent confirmation of patient diagnoses. This analysis is informative, regardless, by evaluating the real-world diagnostic experience of a large number of patients.