Description of cohorts

The analysis was based on 1,931 RABBIT-SpA patients, 1,794 NDB patients and 1,728 RHADAR patients (Table 1). The mean age in the three cohorts was 55, 56, and 60 years; the percentage of female patients was 59%, 56%, and 55%. Disease duration was longer in the RHADAR cohort compared to RABBIT-SpA and NDB.

Table 1 Sociodemographic characteristicsComparison of the clinical characteristics between the three cohorts

Markers of disease activity showed some differences in the three cohorts (Table 2). That is, DAPSA score, TJC, SJC, and PhGA were highest in RABBIT-SpA, followed by NDB and then by RHADAR. PROs showed a similar pattern, with worst scores in RABBIT-SpA, followed by NDB and RHADAR (Table 2).

Table 2 Clinical and patient-reported characteristicsDMARD treatment

There were differences in the treatments that were prescribed in the three cohorts (Table 3). 76% of the RABBIT-SpA patients were on a b/tsDMARD treatment compared to 55% in NDB and 54% in RHADAR. csDMARD monotherapy was used in 13% in RABBIT-SpA, 29% in NDB, and 40% in RHADAR.

The distribution of the b/tsDMARDs was similar across the three cohorts. TNFi were most often used in all three cohorts; followed by IL-17i, IL-23i, IL-12/23i, and JAKi (Table 3).

Discordance of physician global disease activity compared to patient global disease activity

In RABBIT-SpA, the mean PhGA was 2.5 and the mean PtGA was 4.3; in NDB, the mean PhGA was 1.7 compared to 3.6 (mean PtGA) and in RHADAR, the mean PhGA was 0.85 compared to 2.9 (Table 2). The mean discordance (PtGA-PhGA) was 1.8 in RABBIT-SPA, 1.9 in NDB, and 2 in RHADAR.

Discordance > = 3 was present in 34% (RABBIT-SpA), 35% (NDB), and 37% (RHADAR); discordance of 1–2 in 33% (RABBIT-SpA), 35% (NDB), and 37% (RHADAR). Identical assessments were present in 18% (RABBIT-SpA), 22% (NDB), and 23% (RHADAR). In 15% (RABBIT-SpA), 8% (NDB), and 3% (RHADAR) the physicians rated higher values than the patients.

PtGA and PhGA are shown in Fig. 1. On the x-axis, PtGA values (0–10) are shown, and on the y-axis PhGA values. The boxplots show the distribution of PhGA values per PtGA value. For example, in the RABBIT-SpA cohort (Fig. 1a), in those 233 patients who reported a PtGA of 2, the median of PhGA was 1 with an interquartile range (IQR) of 0–2. The pink line shows the discordance between both medians, in this case, one. In those 166 patients with a PtGA of 8 the median PhGA was 4, the PhGA IQR ranged from 2 to 5 and the discordance was 4.

In all three cohorts there are similar features of the graph: for all PtGA values (0–10), the median PhGA is equal or lower and the discordance (pink line) between patient and physician values is increasing with higher PtGA values.

In the RABBIT-SpA cohort, the discordance was slightly smaller than in the NDB and in the RHADAR cohort. The largest discordance was found in RHADAR - this was also the cohort with the lowest PhGA values.

Fig. 1

Boxplots of patient reported global disease activity (PtGA) and physician reported global disease activity (PhGA). On the x-axis PtGA values (0–10) are shown and on the y-axis PhGA values. The boxplots show the distribution of PhGA values per PtGA value. Boxes range from the 25% percentile to the 75% percentile, the line in the middle of the box shows the median. The pink line shows the discordance between the PtGA median value and the PhGA value A RABBIT-SpA; B NDB; C RHADAR. NDB national data base, RABBIT-SpA German disease register RABBIT-SpA, RHADAR RheumaDatenRhePort

Distribution of discordance between PtGA and PhGA

Figure 2 shows the magnitude of discordance between PtGA and PhGA per value of PtGA and the percentage of patients with the respective value. For example, in the RABBIT-SpA cohort, for those patients who reported a PtGA of 3, 19% had a discordance of > = 3 points, 43% had a discordance of 0–3 points, 21% of the physicians also rated 3, and 17% had higher values in their PhGA than 3.

It is evident that as the PtGA values increase, the discordance also increases. At PtGA 4, a discordance of > = 3 was present in 33% of the RABBIT-SpA patients, 49% of the NDB patients, and 71% of the RHADAR patients. This percentage increased progressively with higher PtGA values across all three cohorts. NDB exhibited higher and RHADAR even higher percentages of discordance > = 3 than RABBIT-SpA. In RHADAR, discordance was observed in the vast majority of patients with PtGA > = 4.

Fig. 2

Distribution and magnitude of discordance between PtGA and PhGA. <0: PhGA higher than PtGA, > 0: PtGA is higher than PhGA. A RABBIT-SpA; B NDB; C RHADAR. NDB national data base, RABBIT-SpA German disease register RABBIT-SpA, RHADAR RheumaDatenRhePort

Association between depressive symptoms and discordance between PtGA and PhGA

The association between the categories of depressive symptoms and the discordance is shown in Fig. 3. On the x-axis, the categories of depressive symptoms are shown. The light brown dot is the mean PhGA and the dark brown dot the mean PtGA; the red line shows the discordance between PhGA and PtGA. In those patients with severe depressive symptoms, the discordance between PtGA and PhGA was larger than in those with no depressive symptoms. The same trend was observed in all three cohorts.

Fig. 3

Categories of depressive symptoms and the discordance between PtGA (dark brown circles) and PhGA (light orange circles). The red lines show the mean difference between PtGA and PhGA for each category of depressive symptoms. WHO-5 > 50: well-being, 29–50: mild depressive symptoms, 13–28: moderate depressive symptoms, 0–13: severe depressive symptoms. A RABBIT-SpA; B NDB; C RHADAR. NDB national data base, RABBIT-SpA German disease register RABBIT-SpA, RHADAR RheumaDatenRhePort, WHO-5 WHO-5 well-being index

To further analyse the relationship between PtGA and PhGA discordance and depressive symptoms, a regression model was applied. Age, sex and depressive symptoms, measured by WHO-5 in RABBIT-SpA and NDB and PHQ-4 in RHADAR, were included into the model in each cohort. Depressive symptoms were associated with discordance also if age and sex were accounted for in all cohorts (supplementary Table 1).