Patient Characteristics

Eighty-two patients were included in the study. The majority were male (n = 51, 62.2%), and the mean age (standard deviation, SD) at diagnosis was 39.3 (11.7) years.

Most patients had a diagnosis of r-axSpA (67.1%) and were positive for HLA-B27 (84.2%). Nineteen (23.2%) patients had a previous history of uveitis. Arterial hypertension and hypercholesterolemia were the most frequently recorded comorbidities (Table 1).

Table 1 Clinical characteristics of the study population at symptom onset and diagnosis

At diagnosis, 91.5% of patients presented musculoskeletal signs and symptoms, with inflammatory lower back pain (76.0%) and sacroiliac pain (52.0%) being the most frequent ones.

The majority of patients had active disease according to BASDAI and ASDAS indexes (50% and 86.5% of patients, respectively). The mean (SD) scores were 4.0 (2.4) and 2.8 (1.3), respectively (Table 1).

The mean (SD) diagnostic delay was 10.1 (9.3) years, with a median (interquartile range, IQR) of 5.4 (2.3, 17.2) years, and was significantly higher in patients with r-axSpA vs. nr-axSpA (12.6 vs. 4.8; p < 0.001), with a median (IQR) of 14.6 (2.7, 18.5) and 2.7 (1.5, 5.4), respectively. No significant differences were found by sex: the mean diagnostic delay was 10.5 years in women vs. 9.8 years in men (p = 0.954), with a median (IQR) of 5.3 (2.2, 18.3) years, and 5.4 (2.3, 17.2) years, respectively. The most common symptom at disease onset was inflammatory lower back pain (81.7%), followed by sacroiliac syndrome (67.1%) and cervicalgia (19.5%).

HRU, Loss of Work Productivity, and Associated CostsHRU

During the follow-up period, almost all patients (92.7%) utilized primary care services with a mean (SD) number of annual visits per patient of 4.9 (5.2). All patients visited a specialist, rheumatologists being the most consulted specialists every year (100% of patients) with a mean (SD) number of annual visits per patient of 3.1 (1.7), followed by ophthalmologist and gastroenterologist (17.1% and 13.4% during the first year, respectively). Only 15.9% of patients visited rehabilitation and/or physiotherapy services during the first year, decreasing to 4.8% in the third year (Supplementary Table S1 and S2).

Over half of the patients (57.3%) visited the emergency department at least once during their follow-up, with a mean (SD) number of annual visits per patient of 0.9 (0.7). However, few patients required hospitalization throughout the study (11.0%). The mean (SD) hospital length of stay during the study period was 4.8 (6.6) days (Supplementary Table S3).

Almost all patients (97.2%) required some form of testing during the 3-year follow-up period. The most frequent test performed annually was blood analysis (98.7%), followed by X-ray (54.4%) and magnetic resonance imaging (22.2%).

Non-steroidal anti-inflammatory drugs (NSAIDs) were the most commonly prescribed medication class for axSpA in the first 3 years post-diagnosis, with etoricoxib being the most frequently used agent. The use of biological therapies increased during the study period, with up to 39.0% of patients receiving these drugs by the third year (Supplementary Figure S2). In addition, patients received medications for their comorbidities, the most frequent being antihypertensives, antidiabetics, and lipid-lowering agents (Supplementary Table S4).

Loss of Work Productivity

Among the four (4.9%) patients who presented permanent work impairment at the time of diagnosis, only one case was directly associated with the locomotor system (data not shown). Of these, three (75.0%) were completely unable to perform any kind of work, while one (25.0%) was totally impaired only for his usual occupation. Additionally, up to 42.7% of patients experienced temporary work impairment during the follow-up period. The mean (SD) annual number of days of sick leave days due to temporary impairment during the study was 33.8 (65.0) days and decreased over the study period (Supplementary Table S5).

Associated Costs

The mean (SD) annual total cost per patient was 8604.2€ (11,207.5) with a mean (SD) total cost during the study period of 25,812.6€ (33,622.5) (Table 2). Of the total costs, 63.5% and 36.5% corresponded to direct and indirect costs, respectively (Fig. 1).

Table 2 Mean annual and total costs (€) per patientFig. 1

Distribution of direct and indirect costs per patient. axSpA axial spondyloarthritis

Direct costs increased throughout the study period, while indirect costs were higher in the first-year vs. the second and third year, decreasing from 50.1% to 31.3%, and 28.9% of total cost, respectively (Supplementary Table S6 and Figure S3).

Cost analysis revealed higher total expenses for r-axSpA compared to nr-axSpA, primarily due to increased indirect costs (Supplementary Table S7). Similarly, men incurred higher total and indirect costs than women. However, these differences were not statistically significant.

Association Between Diagnostic Delay and Costs

The primary objective of the study was the analysis of the correlation between diagnostic delay and total costs in the 3 years following diagnosis, which showed a low positive correlation, although it was not statistically significant (Spearman’s coefficient = 0.195, p = 0.08). Similarly, a low positive but not significant correlation was found between diagnostic delay and direct and indirect costs (Spearman’s coefficient = 0.163, p = 0.144 and Spearman’s coefficient = 0.79, p = 0.108, respectively), and total cost in each of the 3 years following diagnosis (p = 0.052; p = 0.101; p = 0.063, respectively) (Supplementary Figure S4).

At the time of data analysis, patients with a shorter diagnostic delay, based on the median, exhibited a significantly lower mean of total cost than those with a longer diagnostic delay (€20,188.7 vs. €31,717.7, p = 0.029). A statistically significant difference was found between groups (p = 0.029) (Table 3). Patients with a diagnostic delay of > 5.4 years had 57.1% higher mean total expenditure and 87.1% higher median total expenditure, mostly due to differences in indirect costs. Differences in the cost of diagnostic delay in each of the follow-up years are shown in Supplementary Table S8.

Table 3 Mean and median annual cost (€) according to the median diagnostic delay (5.4 years)

The secondary goal was to assess the association between diagnostic delay and costs according to patients’ sex and axSpA type. A statistically significant positive correlation between total cost and diagnostic delay was observed for men (p = 0.049) but not for women (p = 0.933). In addition, no significant correlation was observed between total, direct, and indirect costs according to axSpA type and diagnostic delay (all p > 0.05).

Disease Burden

The mean of BASDAI, ASDAS, and BASFI decreased at the 1-year follow-up visit and remained stable throughout the study. Levels of CRP and ESR showed a marked decrease during the first year and remained stable throughout the study. The proportion of patients with active disease decreased at the 1-year follow-up visit and remained stable throughout the study (Fig. 2 and Supplementary Figure S5).

Fig. 2

Number of patients with active disease in each of the follow-up visit of the study. a) activity according to disease activity indices: BASDAI inactive < 4; BASDAI active > 4; ASDAS inactive ≤ 1.3; ASDAS low activity 1.4–2.1; ASDAS high activity 2.2–3.5; ASDAS very high activity > 3.5 and b) activity according to laboratory indices: Negative CRP ≤ 5; Positive CRP > 5; Negative ESR ≤ 20; Positive ESR > 20. ASDAS Axial Spondyloarthritis Disease Activity Score; BASDAI bath ankylosing spondylitis disease activity index; CRP C-reactive protein; ESR erythrocyte sedimentation rate

During follow-up visits, the most common new peripheral signs and symptoms were peripheral arthritis (up to 25.0% of patients), and enthesitis (up to 50.0% of patients). Spinal mobility was similar at diagnosis and at the last follow-up visit (year 3): the mean (SD) scores in Schober’s test were 4.2 (1.2) and 4.1 (1.3), respectively.

Of the 75 patients for whom radiographic progression data were available, 13 patients (17.3%) exhibited progression at the conclusion of the follow-up (year 3). Of these, six patients (46.2%) showed sacroiliac progression (an increase of at least one grade) and five patients (38.5%) experienced an increase in the number of syndesmophytes between diagnosis and year 3, while two patients (15.4%) progressed in both areas.

Association Between Diagnostic Delay and Clinical Variables

No statistically significant correlation was observed between diagnostic delay and the change from baseline of ASDAS, BASDAI, BASFI, CRP and ESR in the 3 years following diagnosis (all p > 0.05). At year 3 of the follow-up period, a statistically significant positive correlation between diagnostic delay and BASDAI was observed (Spearman’s coefficient = 0.2747, p = 0.0294). A similar correlation was found for ASDAS at year 3, although it was not statistically significant (Spearman’s coefficient = 0.211, p = 0.097).

When analyzed based on the median diagnostic delay, a statistically significant higher BASDAI score was observed in patients with a longer vs. shorter diagnostic delay at baseline (4.7 vs. 3.4, p = 0.007) and after 3 years (3.9 vs. 2.9, p = 0.042). The differences in ASDAS did not reach statistical significance at baseline (3.1 vs. 2.5, p = 0.123) nor at year 3 (2.5 vs. 2.1, p = 0.089).