Introduction

Autoimmune extra-intestinal manifestations are frequent in both patients with Crohn’s disease (CD) and ulcerative colitis (UC) [1]. The strong genetical susceptibility is accountable of the presence of co-disorders, and patients and family member with inflammatory bowel diseases (IBD) frequently suffer from rheumatological and dermatological diseases [2].

The presence of an inflammatory arthropathy is the commonest extra-intestinal manifestation in patient suffering from IBD, ranging from the 4% to the 23% of the subjects, classified in the context of spondyloarthritis (SpA) [3]. The spectrum of the clinical features of the inflammatory articular involvement is broad, showing patterns of pure spinal involvement (axialSpA [axSpA], including both ankylosing spondylitis [AS], and non-radiographic axial SpA [nr-axSpA]), and patterns of exclusive peripheral arthritis and/or enthesitis.

However, despite the well-known relationship between gut and joints, despite availability of the Assessment of SpondyloArthritis international Society (ASAS) classification criteria [4] and of sensitive diagnostic techniques, such as magnetic resonance imaging (MRI) and ultrasonography (US), the reported delay to diagnosis is still from 8 to 11 years [5]. Moreover, gastroenterologists not always ask about joint complaints.

During the last years different referral models for axSpA have been proposed [6,7,8,9,10,11,12], only one dedicated to SpA-IBD related [12]. The lowest common denominator among these models is the axial disease.

However, axial disease may be absent in patients with a SpA related to IBD. In a Canadian cohort, in the 30.3% of the patients suffering from IBD was detected a peripheral SpA, versus the 27.3% of the subjects that showed an axSpA [13].

Keeping in mind the abovementioned considerations, the main objectives of this study were to develop and to test in a preliminary way a new self-administered screening tool, called DETection of Arthritis in Inflammatory boweL diseases (DETAIL) questionnaire, to recognize signs and/or symptoms of peripheral and/or axial inflammatory involvement in patients suffering from IBD not previously diagnosed as having a SpA.

Materials and methods

DETAIL questionnaire development

The goal of this study was to develop a 5–8 items questionnaire, with dichotomous answer, to be filled in quickly in the waiting room of the gastroenterologists, easy to understand, and without the need of laboratory or imaging tests.

The screening tool realization followed two major steps: the item pool development and the item pool reduction. The first step was closed in within a consensus among the authors in May 2016. In this context, a list of items was identified, coming from the already existing referral models for axSpA [6,7,8,9,10,11,12]. Moreover, the items coming from the screening questionnaires available for psoriatic arthritis (PsA) were included [14,15,16,17,18,19,20,21], due to the similarities of the peripheral articular inflammatory involvement. The generation of new items was also encouraged, in order to cover all the possible clinical expressions of SpA associated to IBD. After the elimination of duplicates, of composite questions, of the items requiring figures or special equipments, a list of 30 item was drawn up (Table 1).

Table 1 The list of the 30 items coming from the referral models of SpA generated by the consensus of the authors
Full size table

Then, from June 2016 to October 2016, 112 experts in the field of SpA or IBD were requested by email to rate the importance of each of the 30 items/questions for the detection of signs or symptoms articular or spinal inflammation. We got response from 95 (84.8%) physicians, respectively 55 rheumatologists, 21 gastroenterologists, and 19 internists.

The importance of each item was rated according to a Likert scale: 0 = irrelevant, 1 = slightly relevant, 2 = quite relevant, 3 = very relevant.

After the collection of all the opinions, for each item, the mean relevance scores were computed.

Items were retained in the DETAIL questionnaire if they satisfied at least a mean score of 2, and if were rated as quite relevant or very relevant at least by the 70% of the experts. The questions that satisfied the criteria for the inclusion in the final questionnaire (frequency ≥ 70% and mean relevance score ≥ 2.0) were six (Table 2).

Table 2 The six top-rated items (frequency ≥ 70% and mean relevance score ≥ 2.0) included in the final questionnaire
Full size table

In October 2016, the final version of the DETAIL questionnaire, made by the six top-rated items, was available. The English version was translated in Italian, then translated back in English by a mother tongue English speaker.

Preliminary testing

From October 2016 to April 2017, the pilot testing involved IBD patients coming from three gastroenterology units. During the gastroenterology routine visit, each patient with CD or UC, was asked to fill in the DETAIL questionnaire. IBD patients aged > 18 years, without a previous diagnosis of SpA, able to read and understand Italian language were included in the study. Subjects suffering from cancer or lymphoproliferative disease, uncontrolled diabetes, unstable ischemic heart disease or congestive heart failure, and renal failure were excluded.

After that, within 2 weeks from the DETAIL compiling, patients were referred to one of two rheumatology units. In this setting, patients were evaluated by a physician (MDC or MML) blinded to the DETAIL results. The rheumatologic assessment was leaded in accordance to a standard protocol, comprehensive of a complete history and physical examination (tender and swollen joint counts on the 68 and 66 joints, cervical rotation, tragus-to-wall distance, lumbar lateral flexion, modified Schober’s test, intermalleolar distance), and laboratoristic assessment (acute phase reactants, HLA B-27 if needed). X-ray, MRI, and US were performed if needed. The rheumatologists were required to confirm or to exclude the presence of a SpA according to the ASAS criteria [4], and if appropriate, to diagnose the presence of other muskuloskeletal disorders. The study was approved by the Hospital Clinic ethics committee (Comitato Unico Regionale – ASUR Marche), and all patients gave their written informed consent for anonymous analysis of data.

Statistical analysis and DETAIL questionnaire interpretation

The performance of the DETAIL questionnaire was evaluated through the calculation of the post-test probability. The post-test probability evaluation of a screening questionnaire is allowed by the knowledge of sensitivity, specificity, and the pre-test probability (the prevalence of the disorder). In this study, the post-test probability was calculated using the Bayesian Analysis Model method, defining for each item of the questionnaire the positive likelihood ratio (LR+), and with the graphics representation on the Fagan’s nomogram. The nomogram explanation starts drawing a straight line connecting the pre-test probability to the LR for the test results (given by the LR+ product of the items answered as “yes” in the questionnaire). The intercepted point of the straight line on the right axis of the nomogram represents the post-test probability of disease.

Results

Patient characteristics

In this cross-sectional cohort, 128 consecutive IBD patients (72 men—56.3%, and 56 women—43.7%) were evaluated. In particular, 77 were suffering from CD and 51 were suffering from UC. Demographic and treatment features are described in Table 3.

Table 3 Demographic, clinical and treatment characteristics of the study cohort
Full size table

After the rheumatologic assessment, 21 (16.4%) patients have been diagnosed with a SpA (subjects fulfilling the ASAS criteria). This percentage represents our pre-test probability.

In the majority of the subjects diagnosed with SpA (11 patients), the exclusive features of the articular inflammatory disease were represented by the peripheral involvement (eight patients synovitis, three enthesitis). In 10 patients, an axial disease was detectable: in five of them, the axial involvement represented the only one manifestation of the disease, while in the other five, a combination of axial and peripheral disease was detectable. Among these 10 patients with axial involvement, four exhibited a radiologic sacroiliitis, while in six, sacroiliitis at MRI was detectable. Seven of the 10 subjects with axial disease were HLA-B27 positive, versus three of the 11 with exclusive peripheral disease (Fig. 1).

Fig. 1
figure 1

Flow-chart showing the characteristics of the patients diagnosed with spondyloarthritis. DETAIL, DETection of Arthritis in IBD, Inflammatory BoweL Diseases; SpA, spondyloarthritis; ASAS, Assessment of SpondyloArthritis international Society; MRI, magnetic resonance imaging

Full size image

DETAIL replies in patients suffering from a SpA

Analyzing the DETAIL questionnaire among the 21 patients diagnosed having SpA, three answered 6/6 yes, two filled in 5/6 yes, eight put 4/6 yes, one answered 3/6 yes, five filled in 2/6 yes, and two replied 1/6 yes.

DETAIL replies in patients without a SpA

In the majority (107 patients) of the subjects with IBD, a SpA was not diagnosed. Exploring the answers to the DETAIL questionnaire in these patients, 40 replied 0/6 yes, 28 answered 1/6 yes, 24 put 2/6 yes, 8 filled in 3/6 yes, four replied 4/6 yes, two answered 5/6 yes, and one put 6/6 yes.

Alternative diagnoses

Overall, 67 partipants filled in at least 1/6 yes without suffering from a SpA. In this group, focusing on the alternative diagnoses of the musckuloskeletal complaints, the most frequent disease was represented by osteoarthritis (OA) (47 subjects—36.7% of the whole cohort), in particular, lumbar spine OA in 24 patients, with a coexisting Modic type 1 lesion in four, and peripheral OA in 23 subjects. In nine participants, an isolated plantar fascia or Achilles tendon enthesopathy was detectable; in two, the complaints were attributable to a primary Raynaud’s phenomenon. One patient was diagnosed with a gouty arthritis, and in one subject was detected a hip complex regional pain syndrome type 1 (CRPS-1). Interestingly, in all the seven subjects with a DETAIL score ≥ 4 without suffering from a SpA was diagnosed with a fibromyalgia (FM).

DETAIL performance

Applying the Bayesian analysis, of the six items of the DETAIL questionnaire, the best LR+ was found in item 2 (LR+ 3.82), exploring dactylitis, and in item 6 (LR+ 3.82) and item 5 (LR+ 3.40), two questions exploring inflammatory low back pain. Enthesitis (item 3—LR+ 2.87) and peripheral synovitis (item 1—LR+ 2.81) gave similar results, while item 4, exploring the duration of low back pain, resulted in the worst performance (LR+ 1.99) (Table 4).

Table 4 Sensitivity, specificity, positive and negative predictive values, likelihood ratios, and post-test probabilities of the six items of the DETAIL questionnaire
Full size table

In the Fagan’s nomogram representation, the disease probability in a given patient is the product of the LR+ of each item answered as “yes”. A diagnostic test can be considered meaningful if the post-test probability is over 75%. For the DETAIL questionnaire, the presence of at least three affirmative questions results in a post-test probability ≥ 75% for each possible scenario.

It can be affirmed that the presence of three on six items answered as “yes” is the cutoff point for the rheumatological referral (Fig. 2), given that it reflects a high probability for the occurrence of a SpA.

Fig. 2
figure 2

The final graphical representation of the DETection of Arthritis in Inflammatory boweL diseases (DETAIL) questionnaire

Full size image

DETAIL feasibility and acceptability

The mean time to fulfill the DETAIL was 0.8 ± 0.3 min (range 0.6–2.2 min), and it was correctly completed by the vast majority of the subjects (less than 3% of questionnaires with missing values).

The majority of the subjects (87.5%) rated that the tool was easy understandable and to be filled in.

Discussion

One of the main objectives of the clinical rheumatologist is the early detection of an inflammatory joint disease. During the last years, the availability of international classification criteria next to the powerful imaging techniques is simplifying the SpA diagnostic process.

However, certain works still describe a substantial diagnostic delay [5].

SpA adversely and deeply affects the health-related quality of life. Even in the early phases of the disease, inflammation and the subsequent articular damage translate into pain, impaired motility and altered physical function [22]. This inevitably results in a serious impact on work ability and in social partecipation [23], and the severity of disease has to be measured in many domains of daily life [24].

Nowadays, the diagnostic delay is considered unacceptable, particularly in those patients in whom SpA coexists with other inflammatory diseases, as in the case of IBD. On the other hand, the limitations in terms of time and resources make nearly impossible that all IBD patients can be assessed by a rheumatologist.

It is mandatory to maximize the cooperation between gastroenterologists and rheumatologists, optimizing the number of patients to undergo on musculosckeletal assessment.

From this perspective, the aims of this work were to develop and to test in a preliminary way a new screening tool for the early detection of signs and symptoms of an inflammatory articular disease in patients with IBD. In order to be widely applicable, a referral tool has to be easy to understand by a patient and to be filled in quickly. The DETAIL questionnaire meets these specific requirements, and also showed good screening properties. From the data obtained, it can be gathered that the affirmative answer for three of the six dichotomous items results in a post-test probability of suffering of SpA of at least 75%. The local prevalence of SpA in IBD patients (16.4%) is in line with previous experiences [3].

Many referral models have been built during the last years, recentely compared in the SPACE cohort [25]. All the 13 referral strategies performed well, but it is concluded that the ideal referral strategy could be different from country to country. Moreover, simple strategies work well as more complex referral models [26].

The main focuses of all these models are the inflammatory back pain (IBP) and the HLA-B27 presence. On the other hand, in different cohorts, it was documented that peripheral joint involvement is more prevalent than the axial [3, 27]. The DETAIL questionnaire explores both the peripheral synovitis and the IBP.

The major limitations of the questionnaire were found in subjects with coexisting musculoskeletal diseases. A primary confounder for DETAIL questionnaire interpretation is represented by OA, due to the high prevalence of this condition in the general population. However, OA symptoms seem to not deeply alter the DETAIL screening properties, influencing only few questions. Thus, the DETAIL items look positively to distinguish between the inflammatory and mechanic features of pain.

Physicians, on the contrary, should be aware of FM, revealed up to the 30% of IBD patients [28]. In this subgroup of patients, the questionnaire is practically worthless. In five DETAIL items, the word “pain” is mentioned. As the items stands, for subjects suffering from FM, it is almost impossible to differentiate between inflammatory and chronic widespread pain. It is the rheumatologist’s duty to attest the presence of synovitis, enthesitis and, potentially, central sensitization.

After all, the presence of a coexisting FM is a critical aspect in the interpretation of the instruments that are patient-centered, expecially in the field of SpA [29].

The DETAIL questionnaire has to confirm the preliminary findings in a larger and multicentric scale in which the characteristics of the IBD (especially in terms of localization, severity, disease duration), and the role of the treatment have to be better typified. In a broader validation study, it is also necessary to evaluate reliability performing a test-retest. In future studies, the screening properties should to be compared to those of the already available instruments.