Validation of Turkish psoriatic arthritis screening tool for psoriatic arthritis: A cross-sectional comparative study

Abstract

Psoriasis (PsO) is a chronic immune-mediated disease that affects the skin and nails. Several screening tools such as psoriasis epidemiology screening tool, psoriatic arthritis screening and evaluation, early arthritis for psoriatic patients (EARP), STRIPP, SIPAS, and TOPAS have been developed and validated for identifying patients with suspected psoriatic arthritis (PsA) in dermatology clinics. Herein, we aim to validate and evaluate the performance of TurPAS in patients with PsO compared to the EARP questionnaire. This cross-sectional study was carried out in collaboration between dermatology and rheumatology clinics at Aksaray University Hospital, a tertiary referral center. All patients were examined for dermatologic aspects in the dermatology outpatient clinic and TurPAS and EARP questionnaires were performed by the same dermatologist. The patients who met the CASPAR criteria were classified as PsA in addition to appropriate clinical findings. In our study, 119 patients (n = 72; 60.5% female) with PsO were included. Mean ± standard deviation age of the patients was 44.9 ± 14.3 (range; 18–83) years and the median (IQR) PsO duration was 10 (14) years. Median (IQR) TurPAS and EARP scores were 2 (3) and 3 (5), respectively. Overall, 39 patients (32.8%) met the CASPAR criteria and were classified as PsA. The best diagnostic accuracy was observed in ≥ 3 positive items with 82% and 70% sensitivity and specificity (positive predictive value: 57.1%, negative predictive value: 89%, likelihood ratio: 2.7, diagnostic accuracy: 74%) for TurPAS and ≥ 4 items with 74.4% and 86.3% (positive predictive value: 72.5%, negative predictive value: 87.3%, likelihood ratio: 5.4, diagnostic accuracy: 82.4%) for EARP. In our study, approximately one-third of patients with PsO were diagnosed as PsA. TurPAS had acceptable diagnostic performance compared to the EARP questionnaire.

1. Introduction

Psoriasis (PsO) is a chronic immune-mediated disease that affects the skin and nails. Approximately 1% to 2% of the population suffers from PsO.^[1] Psoriatic arthritis (PsA) is a heterogeneous disease associated with PsO. Approximately 5% to 40% of patients with PsO develop PsA during the disease course.^[2] Entesitis, dactylitis, peripheral arthritis, and axial skeletal involvement are the main features of PsA. Joint involvement of PsA has many subtypes, such as mono-oligoarthritis, polyarthritis, distal interphalangeal joint (DIP) arthritis, arthritis mutilans, and axial disease. Psoriatic patients are also at higher risk for metabolic syndrome and its components such as obesity, diabetes mellitus, hypertension, hyperlipidemia, and coronary heart disease.^[3] Due to its complex and heteregenous heterogeneous nature, “Psoriatic disease” has recently been used as an umbrella term for all disease-related findings reflecting rheumatological, dermatological, and metabolic aspects.^[4] The absence of specific biomarkers, overlapping symptoms with other musculoskeletal disorders, and the fact that joint symptoms can occur independently of cutaneous signs often delay the diagnosis of PsA.^[5,6] Dermatology clinics play a pivotal role in the early detection of PsA, as the condition typically follows or accompanies cutaneous psoriasis manifestations.^[7] The variable clinical presentation and often subtle early symptoms require dermatologists to remain vigilant.^[2] By assessing musculoskeletal symptoms alongside dermatological findings, dermatologists can identify patients at risk and refer them promptly to rheumatology, thereby preventing disease progression and long-term joint damage.^[8]

Early diagnosis and treatment of PsA are essential to avoid irreversible joint deformities, work disability, and other serious complications.^[7] Thus, screening and identifying PsA at an early stage in dermatology settings is critical for improving outcomes in patients with psoriasis.

Referring patients from dermatology to rheumatology clinics is problematic since it is unclear which PsO patients should be consulted. Several screening tools, such as PEST, PASE, EARP, STRIPP, SIPAS, TOPAS, GEPARD, and CONTEST have been developed and validated for identifying patients with suspected PsA in dermatology clinics.^[9–11] However, these tools often assess a limited number of symptoms, excluding key clinical domains such as axial involvement, enthesitis, or dactylitis, which can potentially lead to delayed diagnosis. Additionally, some questionnaires are comprehensive and time-consuming to complete, limiting their practicality in busy clinical environments (e.g., PASE, ToPAS, CONTEST). A widely adopted, standardized, and universally accepted screening tool remains lacking.

To address these limitations, TurPAS was recently developed through collaboration between Turkish rheumatologists and dermatologists to provide a more holistic assessment of PsA’s heterogeneous features. Consisting of 6 items, TurPAS evaluates key clinical domains including synovitis, enthesitis, dactylitis, and axial involvement,.^[12] Its brief administration time, ease of use in clinical practice, and broad coverage of diverse psoriatic phenotypes make TurPAS a promising screening instrument. The current study aims to validate the TurPAS questionnaire by comparing its diagnostic performance to that of the EARP in patients with psoriatic disease.

2. Material and methods

2.1. Study group

This cross-sectional study was carried out in collaboration between dermatology and rheumatology clinics at Aksaray University Hospital, a tertiary referral center. Patients with a diagnosis of PsO who followed up in a dermatology outpatient clinic were included in the study. All patients were examined for dermatological aspects in the dermatology outpatient clinic, and TurPAS and EARP questionnaires were self-administered by the patients following a brief standardized explanation provided by the same experienced dermatologist (ŞBK). Patients were allowed to ask clarification questions without receiving guiding responses. Questionnaire administration was performed under similar conditions for all participants to ensure consistency. Afterward, the patients were referred to the rheumatology clinic for rheumatological evaluation regardless of the questionnaire results. Therefore, the patients who had no musculoskeletal complaints were also included in the study. Prior treatment history of the patients with PsO including conventional synthetic disease-modifying antirheumatic drugs (cs-DMARD) and biological disease-modifying antirheumatic drugs (b-DMARD) was also noted. The TurPAS questionnaire is shown in Table 1. The rheumatological evaluation was performed by a single experienced rheumatologist (MB) who was blinded to relevant the questionnaire results. The patients under 18 years, who had cognitive dysfunction, and/or were unable to read were excluded from the study.

Table 1.

The new screening tool: TurPAS.

Items	Domain	Turkish version	English version
Q1	Joint	El/ayak parmaklarinizda ya da herhangi bir ekleminizde hiç şişlik veya ağri oldu mu?	Have you ever had swelling or pain in your fingers/toes or any of your joints?
Q2	Dactylitis	Resimde gösterildiği gibi el veya ayak parmağinizda sosis şeklinde şişlik oldu mu?	Have you had a sausage-shaped swelling on your fingers or toes as shown in the picture?
Q3	Enthesitis	Topuk ağriniz olur mu?	Do you have heel pain?
Q4	Axial involvement and morning stiffness	Belinizde, sirtinizda veya boynunuzda istirahatle artan, özellikle sabaha karşi kötüleşen veya sabahlari hareketinizi kisitlayan ağriniz olur mu?	Do you have pain in your lower back, back, or neck that increases with rest, worsens especially in the morning, or restricts your movement in the morning?
Q5	Drug use	Eklem şikayetleriniz için zaman zaman ilaç kullanir misiniz?	Do you take medication for your joint complaints from time to time?
Q6	History of rheumatic disease	Size daha önce iltihapli romatizma tanisi konuldu mu?	Have you ever been diagnosed with a rheumatic disease before?

Q = question, TurPAS = Turkish psoriatic arthritis screening questionnaire.

Informed consent was obtained for this study. Ethics Committee approval was obtained from Aksaray University, Faculty of Medicine (Number: 2022/21-11).

2.2. Rheumatological evaluation

Baseline laboratory evaluation such as hemogram, liver enzymes, kidney functions as well as acute-phase reactants (erythrocyte sedimentation rate, C-reactive protein [CRP]), autoantibodies (rheumatoid factor, anti-cyclic citrullinated peptides [anti-CCP]) were performed in all participants. Additionally, all patients underwent protocol plain radiography screening for peripheral joints (hand, foot, knee, and sacroiliac joints) at baseline and for other symptomatic joints if needed. Magnetic resonance imaging was also performed if a diagnostic challenge persisted despite the baseline evaluation of peripheral involvement. Axial skeletal magnetic resonance screening including cervical, thoracic, and lumbar spine and sacroiliac joint was performed in patients who had inflammatory neck and/or back pain whose x-ray results were unremarkable. Axial involvement (sacroileitis and/or spine involvement) was defined as ASAS axial spondyloarthritis criteria.^[13] While total axial involvement was defined as axial involvement alone or accompanying peripheral joint involvement, pure axial involvement was defined as the absence of peripheral involvement in patients who had axial disease. The patients who met CASPAR criteria were classified as PsA in addition to appropriate clinical findings.^[14]

2.3. Sample size and selection

We selected our study population through consecutive sampling of patients presenting to the dermatology outpatient clinic with a confirmed diagnosis of psoriasis. This approach was chosen to reflect real-world clinical practice and to capture a representative sample of patients with varying disease severity, duration, and musculoskeletal symptomatology. By including all eligible patients regardless of their musculoskeletal complaints, we aimed to minimize selection bias and enhance the generalizability of the findings. This strategy allowed for the inclusion of both asymptomatic and symptomatic patients, thereby supporting the evaluation of the screening tools across the full spectrum of psoriatic disease presentations.

2.4. Statistical analysis

In our study, the SPSS software version 21.0 (IBM, Armonk) was used for statistical analysis. The normality of continuous variables was assessed using the Shapiro–Wilk test. In descriptive statistics, discrete and continuous numerical variables were expressed as mean ± Standard deviation (SD) or median and interquartile range (IQR), depending on the data distribution. Categorical variables were expressed as numbers and percentages. For comparisons between categorical variables, Chi-square or Fisher exact tests were applied. For continuous variables, normally distributed data were analyzed using Student t-test or paired t-test, while non-normally distributed data were analyzed using the Mann–Whitney U test (for 2-group comparisons) or the Kruskal–Wallis test (for comparisons across more than 2 groups). These nonparametric tests were selected due to their robustness in handling skewed data and violations of normality assumptions. For multiple intergroup comparisons, the Dunn test was applied to identify significant pairwise differences while controlling the type I error rate. Multivariate analysis was conducted using binary logistic regression to compare TurPAS and EARP scores, including the items with a P-value < .2 in the univariate analysis. Receiver operating characteristic (ROC) curve analysis was used to calculate sensitivity, specificity, and the area under the curve (AUC) for diagnostic performance. A P-value of < .05 was considered statistically significant.

3. Results

In our study, 119 patients (n = 72; 60.5% female) with PsO were included in the study. Mean ± SD age of the patients was 44.9 ± 14.3 (range; 18–83) years and the median (IQR) PsO duration was 10 (14) years. Most of the patients had plaque-type PsO (86.6%) and the median (IQR) PASI score was 3 (5). Median (IQR) erythrocyte sedimentation rate : 12.5 (14) mm/hour and CRP: 11.5 (20) mg/L. Median (IQR) TurPAS and EARP scores were 2 (3) and 3 (5), respectively. While 3 patients (2.6%) had rheumatoid factor positivity, anti-CCP was negative in all patients. Overall, 39 patients (32.8%) met the CASPAR criteria and were diagnosed as PsA. Mono-oligoarthritis was in 19 patients (48.8%), polyarthritis in 10 (25.6%), isolated DIP arthritis in 5 (12.8%), overall axial involvement in 9 patients (26.5%), and pure axial involvement in 5 patients (12.8%) in patients with PsA. No patients had arthritis mutilans in our study. Enthesitis was observed in 25 patients (64%) and dactylitis in 10 patients (25.6%) in patients with PsA. Mean ± SD (range) tender and swollen joint counts were 1.05 ± 1.5 (0–6) and 1 ± 1.5 (0–6), respectively (Table 2). After the rheumatological evaluation, 39 patients (32.8%) had a non-PsA diagnosis as follows; osteoarthritis (OA) in 13 patients (33.3%), lumbar disk hernia in 10 (25.6%), nonspecific musculoskeletal pain in 7 (18%), fibromyalgia (FM) in 3 (7.7%), cervical disk hernia in 2 (5.1%), gout and spina bifida in 1 patient (2.6%), respectively (Table S1, Supplemental Digital Content, https://links.lww.com/MD/P768). Phototherapy was used in 7 patients (5.9%), acitretin in 25 (21%), cyclosporin in 6 (5%), methotrexate in 37 (31.1%), b-DMARD treatment in 19 (16%) in patients with PsO (Table S2, Supplemental Digital Content, https://links.lww.com/MD/P768).

Table 2.

Clinical and laboratory features of study participants.

Variables	Results
Age (year), mean ± SD (range)	44.9 ± 14.3 (18–83)
Gender, female, n (%)	72 (60.5)
Duration of PsO (year), median (IQR)	10 (14)
Family history of PsO, n (%)	39 (32.8)
Family history of PsA, n (%)	5 (4.2)
Psoriasis subtypes
Plaque	103 (86.6)
Palmoplantar	6 (5)
Guttat	6 (5)
Invers	2 (1.7)
Nail (only)	2 (1.7)
Special areas of PsO, n (%)
Scalp	76 (64)
Intergluteal	13 (11)
Axillary	12 (10)
Nail (overall)	49 (41.2)
PASI score, median (IQR)	3 (5)
ESR, median (IQR)	12.5 (14)
CRP, median (IQR)	11.5 (20)
RF positivity, n (%)	3/114 (2.6)
Anti-CCP positivity (n = 108)	0
Comorbidities, n (%)
Diabetes mellitus	20 (16.8)
Hypertension	18 (15)
Coronary heart disease	7 (5.9)
Hyperlipidemia	8 (6.7)
Tender joint count, mean ± SD (range)	1.05 ± 1.5 (0–6)
Swollen joint count, mean ± SD (range)	1 ± 1.5 (0–6)
TurPAS score, median (IQR)	2 (3)
EARP score, median (IQR)	3 (5)
Diagnosis of PsA, n (%)	39 (32.8)
PsA subtypes, n (%)
Mono-oligoarthritis	19 (48.8)
Polyarthritis	10 (25.6)
DIP arthritis	5 (12.8)
Pure axial involvement	5 (12.8)
Total axial involvement, n (%)	9 (26.5)
Enthesitis, n (%)	25 (64)
Dactilytis, n (%)	10 (25.6)

anti-CCP = anti-cyclic citrullinated peptides, CRP = C-reactive protein, DIP = distal interphalangeal phalanx, EARP = early psoriatic arthritis screening questionnaire, ESR = erythrocyte sedimentation rate, IQR = interquartile range, PASI = psoriasis area severity index, PsA = psoriatic arthritis, PsO = psoriasis, RF = rheumatoid arthritis, SD = standard deviation, TurPAS = Turkish psoriatic arthritis screening questionnaire.

In univariable analysis, family history of PsA (10.5% vs 1.3%; P = .037; Odds ratio [OR]:5.5), median CRP levels (16.5 vs 3.5 mg/L; P = .003), TurPAS and EARP scores (4 vs 2; P < .001) and (6 vs 2; P < .001) were higher in patients with PsA than those without (Table S3, Supplemental Digital Content, https://links.lww.com/MD/P768). Median TurPAS (4 vs 2) and EARP (6 vs 2) scores were also higher in patients with PsA compared to patients who had a non-PsA diagnosis (P < .001 and P < .001, respectively). In multivariable analysis, TurPAS (P = .003, OR: 3.8; 95% Confidence interval [CI]: 1.56–9.4) (Table S4, Supplemental Digital Content, https://links.lww.com/MD/P768) and EARP (P = .019, OR: 2.06; 95% CI: 1.13–3.76) scores were associated with the diagnosis of PsA in separate models. In an additional multivariable model including both screening tools simultaneously, only the EARP score remained a statistically significant independent predictor of PsA (OR: 1.82, 95% CI: 1.27–2.68, P = .002), whereas the TurPAS score was not significant (OR: 1.21, 95% CI: 0.66–2.20, P = .540) (Table S5, Supplemental Digital Content, https://links.lww.com/MD/P768).

ROC analysis revealed AUC: 0.834, P < .001, 95% CI: 0.751–0.917 and AUC: 0.862, P < .001, 95% CI: 0.784–0.940 for TurPAS and EARP, respectively (Fig. 1). Although EARP had a slightly higher AUC than TurPAS, a formal comparison using the Delong test revealed no statistically significant difference between the 2 (z = –0.48, P = .63). The sensitivity and specificity of TurPAS and EARP were 94.9% and 25% (positive predictive value [PPV]:38.1%, negative predictive value [NPV]: 90.9%) and 94.9% and 25% (PPV:38.1%, NPV: 90.9%), respectively for ≥ 1 positive items. ≥3 positive items provide 82% sensitivity and 70% specificity with a false-positive rate of 42% and a false-negative rate of 11% (PPV: 57.1%, NPV: 89%, likelihood ratio: 2.7, diagnostic accuracy: 74%) for TurPAS ≥ 3 positive items also provide 89.7% sensitivity and 58.8% specificity with a false-positive rate of 47.8% and a false-negative rate of 7.8% (PPV: 51.5%, NPV: 92.2%, likelihood ratio: 2.5, diagnostic accuracy: 68.9%) for EARP. Diagnostic performances of TurPAS and EARP questionnaires according to item count were described in Table 3. The sensitivity and specificity of item 1 (question; Q1) were 94.9% and 36.3%, 46.2% and 93.8% for item 2, (Q2), 35.9% and 80% for item 3, (Q3), 66.7% and 57.5% for item 4, (Q4), 84.6% and 56.3% for item 5, (Q5), 35.9% and 100% for item 6 (Q6) in TurPAS questionnaire (Table 4).

Figure 1.

ROC analysis of TurPAS and EARP for diagnosis of Psoriatic arthritis. EARP = early arthritis for psoriatic patients, ROC = receiver operating characteristic, TurPAS = Turkish psoriatic arthritis screening questionnaire.

Table 3.

Diagnostic performance of screening tool according to item count in patients with psoriatic arthritis.

Item counts	Sensitivity (%)	Specificity (%)	PPV (%)	NPV (%)	Accuracy (%)
≥1 positive items
TurPAS	94.9	25	38.1	90.9	47.9
EARP	94.9	25	38.1	90.9	47.9
≥2 positive items
TurPAS	88.7	47.5	45.5	90.5	61.3
EARP	92.3	40	42.9	91.4	57.1
≥3 positive items
TurPAS	82	70	57.1	89	74
EARP	89.7	58.8	51.5	92.2	68.9
≥4 positive items
TurPAS	61.5	91.3	77.4	83	81.5
EARP	74.4	86.3	72.5	87.3	82.4
≥5 positive items
TurPAS	30.8	98.8	92.3	74.5	76.5
EARP	74.4	86.3	72.5	87.3	82.4
≥6 positive items
TurPAS	2.6	100	100	67.8	68
EARP	64.1	92.5	80.7	84	83.2

EARP = early psoriatic arthritis screening questionnaire, LR = likelihood ratio, NPV = negative predictive value, PPV = positive predictive value, TurPAS = Turkish psoriatic arthritis screening questionnaire.

Table 4.

Diagnostic performance of items of TurPAS.

Items	Sensitivity (%)	Specificity (%)	PPV (%)	NPV (%)	Accuracy (%)	PTP (%)
Item 1	94.9	36.3	42	93.5	55.5	42
Item 2	46.2	93.8	78.3	78.1	78.2	79
Item 3	35.9	80	46.7	71.9	65.6	47
Item 4	66.7	57.5	50	80.6	67.2	44
Item 5	84.6	56.3	48.5	88.2	65.6	48
Item 6	35.9	100	100	76.2	79	NA

LR = likelihood ratio, NA = not applicable, NPV = negative predictive value, PPV = positive predictive value, PTP = posttest probability, TurPAS = Turkish psoriatic arthritis screening questionnaire.

4. Discussion

PsA is a heterogeneous disease that may affect various parts of the body, such as skin, nails, enthesis, joints, and axial skeletons. In our study, about one-third of patients with PsO (32.8%) met the CASPAR classification criteria and were diagnosed as PsA. The frequency of PsA may differ between 5% and 40% in several studies, probably due to differences in study design, diagnosis method, and study population.^[15,16] Higher TurPAS and EARP scores in patients with PsA than those without in our study were also expected findings. In the Lopez-Medina study, while polyarticular involvement was the dominant feature of PsA (60%), mono-oligoarthritis was observed in 40%, enthesitis in 45.8%, dactylitis in 37%, and axial involvement in 35.5% in patients with PsA.^[17] In our study, although the oligo-monoarticular subtype was the most frequent involvement, other manifestations were similar to those in the former study. The frequency of subtypes of PsA highly differs across multiple studies and may be associated with differences in the study population and/or sample size.^[18]

It is important to identify the PsA patients from dermatology clinics for early diagnosis and treatment in patients with PsA. Thus, several screening tools such as the Toronto psoriatic arthritis screening questionnaire (TOPAS), psoriasis epidemiology screening tool (PEST), psoriatic arthritis screening and evaluation (PASE), and the psoriasis and arthritis screening questionnaire, early arthritis for psoriatic patients (EARP) were developed and validated, so far.^[19–23] Currently, there is no widely used, easily applicable, time-saving, and high-diagnostic performance screening tool. In Elaine Husni et al study with 69 participants, the sensitivity and specificity of PASE (cutoff ≥ 47/75) were 82% and 73%, respectively.^[19] Ibrahim et al study revealed a higher diagnostic performance of the PEST questionnaire (cutoff ≥ 3/5) with 92% sensitivity and 78% specificity.^[20] Conversely, the sensitivity and specificity of the PEST questionnaire were 52% and 93.4%, respectively in another study.^[10] In a multicenter study, the sensitivity and specificity of the PEST questionnaire were 68% and 71%, the PASE questionnaire was 66% and 57%, and the EARP questionnaire was 87% and 34%, respectively.^[11] Although the sensitivity was similar (85.2%), the specificity (91.6%) was quite lower in the EARP questionnaire (≥3 items) in the Tinazzi et al study compared to our results (89.7% and 58.8%, respectively).^[23] Diagnostic performances of relevant questionnaires are highly variable according to several studies, probably associated with differences in study design, study population, and sample size. Furthermore, different cutoff values in item counts also significantly affect sensitivity and specificity analysis for the screening tools. For instance, while a cutoff value of ≥ 3 positive items provides 82% sensitivity and 70% specificity, 61.5% sensitivity and 91.3% specificity were observed for ≥ 4 positive items in the TurPAS questionnaire in our study.

In our study, TurPAS and EARP scores were higher in patients with PsA than those without, as expected. Furthermore, TurPAS and EARP scores were higher in patients who had PsA compared to patients who had non-PAS diagnoses such as OA, other mechanical conditions, and FM. These findings are important to make differential diagnoses adequately and allow discrimination from potential mimickers such as OA, FM, gout, and others. In the Meling et al study, 18.1% of PsO patients had a PsA diagnosis, and 73.2% of these patients did not have PsA.^[10] In the Husni et al study, a diagnosis of OA was observed in 24 patients (34.8%), which was higher compared to the diagnosis of PsA (24.6%) among 69 PsO patients.^[19] Due to the common similarities between PsA and OA such as the age of incidence, DIP, and vertebral involvement as well as new bone formation on plain radiography, osteoarthritis may frequently cause a diagnostic challenge in patients with PsO. Additionally, there is some evidence of overlapping with the 2 diseases.^[24] FM and mechanical conditions, such as lumber and cervical disk hernia, were other frequent non-PsA diagnoses in our study, which were consistent with previous results.^[25] In our study, the false-positive rate was slightly lower, but the false-negative rate was slightly higher in TurPAS compared to EARP. Both TurPAS and EARP had acceptable false-negative rates but higher false-positive rates, indicating potential mimickers such as OA and FM may negatively influence the results of screening tools for PsA. Our study and previous studies indicate that current screening tools for PsA are far from being ideal.^[11] Thus, screening tool results should be assessed cautiously, and it should be kept in mind that diagnosis of PsA requires detailed clinical, laboratory, and imaging evaluation by an experienced rheumatologist alongside the relevant questionnaire results.

Our study used EARP for TurPAS validation because it is easily applicable, has been validated in large series, and comprehensively evaluates all aspects of PsA. While the sensitivity of TurPAS was higher, its specificity was lower compared to the EARP questionnaire in our study. Moreover, despite these numerical differences in diagnostic metrics, a formal comparison of the ROC curves demonstrated no statistically significant difference in AUC between TurPAS and EARP, supporting their comparable overall performance. The higher sensitivity of TurPAS compared to EARP is a notable finding, given that TurPAS has fewer items than the EARP questionnaire. The most noteworthy reduction in item counts was related to the questioning of enthesitis and morning stiffness for peripheral arthritis. Although EARP included questioning of several enthesitis sites, such as the heel and elbow, TurPAS only included questions about heel pain with no sensitivity decrease. While EARP questioned specific peripheral joint involvements such as hip, ankle, and wrist, TurPAS had only 1 item for peripheral arthritis questioning (Q1). Although there was no decrease in the number of items, the lack of decline in sensitivity seems to be the most remarkable difference of TurPAS compared to EARP. Having lower item counts in TurPAS may make it more easily performed compared to EARP. Although TurPAS showed comparable diagnostic performance in univariable analyses, its independent predictive value was not significant when both tools were included in the same multivariable model. These results indicate that EARP has superior independent diagnostic value when both tools are considered concurrently.

In TurPAS, Q1 has the highest sensitivity, and Q6 has the highest specificity. While Q1 does not include axial involvement, the highest sensitivity of Q1 is a remarkable finding, probably due to the lower frequency of pure axial involvement in our study. The definition of axial PsA and discrimination from Ankylosing spondylitis as well as other potential mimickers such as chronic low back pain/fibromyalgia, vertebral osteoarthritis, and osteitis condensans ilii, is challenging in daily practice. For this reason, the sensitivity and specificity of these questionnaires are quite low in axial PsA compared to other disease manifestations as in our study and the original study of EARP.^[23] However, the diagnostic performance of Q4 in our study was higher compared to the question of axial involvement in EARP (Q3), possibly due to the more detailed questioning in Q4 than in Q3. The finding of the highest specificity of Q6 is expected, which is consistent with prior diagnosis. In our study, the sensitivity of TurPAS remarkably decreased by ≥ 6 items compared to EARP, possibly due to the Q6 itself, which questioned whether or not a prior rheumatological diagnosis. Q6 is the most specific but one of the least sensitive questions in TurPAS, as already seen in Table 4. There is no similar question in the EARP questionnaire, and this issue may be responsible for the marked sensitivity decline in TurPAS compared to EARP.

Our study has some strengths and limitations. This is the first study regarding the validation of TurPAS alongside a comparison with the EARP questionnaire. The relatively small sample size is a limitation. Including patients who had no musculoskeletal complaints may have allowed for a more accurate estimation of the true incidence of PsA among patients with PsO and may have reduced selection bias. The diagnosis of PsA was made by a single experienced rheumatologist based on a standardized diagnostic protocol that incorporated clinical examination, acute-phase reactants, autoantibody testing, and imaging (X-ray and MRI), which is a methodological strength. Additionally, the identification and classification of non-PsA musculoskeletal conditions, along with their comparison to questionnaire scores, provided a broader understanding of diagnostic differentiation.

However, a notable limitation is the cross-sectional design of the study, which restricts the ability to draw conclusions about causality or disease progression. As data were collected at a single time point, the study cannot evaluate the predictive value of TurPAS or EARP for future development of PsA in initially negative patients, nor can it assess long-term patient outcomes. Prospective longitudinal studies are warranted to validate the screening tools in predicting future diagnoses and to better understand their utility in clinical follow-up.

Lastly, the relatively high proportion of patients receiving cs-DMARD and b-DMARD treatments might have attenuated musculoskeletal symptoms, potentially leading to underdiagnosis of PsA.

5. Conclusion

In our study, approximately one-third of patients with PsO were diagnosed as PsA. TurPAS had comparable diagnostic performance to the EARP questionnaire in our study. More data are needed with further studies to clarify the relevance of TurPAS as a screening tool in patients with PsO.

Author contributions

Conceptualization: Şeyma Başar Kilic, Murat Bektaş.

Data curation: Şeyma Başar Kilic, Huzeyfe Kulu, Hüseyin Salman, Murat Bektaş.

Formal analysis: Şeyma Başar Kilic, Murat Bektaş.

Funding acquisition: Şeyma Başar Kilic, Murat Bektaş.

Investigation: Şeyma Başar Kilic, Murat Bektaş.

Methodology: Şeyma Başar Kilic, Murat Bektaş.

Project administration: Şeyma Başar Kilic, Murat Bektaş.

Resources: Şeyma Başar Kilic, Murat Bektaş.

Software: Şeyma Başar Kilic, Murat Bektaş.

Supervision: Şeyma Başar Kilic, Murat Bektaş.

Validation: Şeyma Başar Kilic, Murat Bektaş.

Visualization: Şeyma Başar Kilic, Murat Bektaş.

Writing – original draft: Şeyma Başar Kilic, Murat Bektaş.

Writing – review & editing: Şeyma Başar Kilic, Murat Bektaş.