The main finding in our study is that a relatively high proportion of patients who were not transferred to an adult clinic had an active disease at the 18-year follow-up study visit, despite previously being in long-term remission off medication. However, their disease activity was mild and had a minimal impact on their daily lives. On the other hand, those who were directly transferred had probably had a more severe disease course at the time of their transition to the adult clinic, and most of them still had either active disease, or ongoing medication, and regular appointments at the adult rheumatology clinic at their 18-year follow-up study visit. These findings could indicate that the appropriate decision concerning transition was made in the past. However, whether young adults with a history of JIA either do not recognize the joint symptoms or do not have the information needed to reach the health care system for evaluation remains to be considered. Patients with JIA may be habituated to their ongoing pain and discomfort caused by the disease, and these experiences may have given them a tenacious character with a higher threshold to seek medical help. However, we did not investigate these factors in our study. As a conclusion, it can be considered that a different kind of transition system might have nonetheless benefitted these young adults.

In our study, 40% of the patients were directly transferred to an adult clinic. Considering the previous findings in this study group, one might expect the transition rate to be higher when it is noted that 49 and 53% of the patients with JIA were not in remission after 8 years and 18 years from diagnosis, respectively [13, 14]. The cumulative transition rate of just above 50 % in our study could also be considered too low in light of the documented chronic nature of JIA [13, 14, 22, 23].

The ‘directly transferred’ and ‘later referred’ groups were similar in many aspects in our study; patients in those groups were older at the disease onset, thus with a shorter follow-up period in the paediatric ward, and they had more polyarthritis and extended oligoarthritis compared to the ‘not transferred’ group. A more detailed analysis of the ‘later referred’ group could reveal aspects of whether they should have been transferred directly as well. Nevertheless, only half of them were still attending an adult rheumatology clinic, so the symptoms leading to a later referral were supposedly mild. Our deduction is that too many patients were incorrectly assessed outside the transition to adult care, as their disease was activated later, and they needed to be referred to an adult clinic. The worrying finding was the relatively large amount (39%) of patients that had an active disease without an appropriate healthcare contact. These concerns could be avoided by transferring all patients, but in most sites the resource constraints cause some limitations.

A recent article revealed detailed information about uveitis in this Nordic study population [24]. It was discovered that, during the 18-year follow-up, 22% had had uveitis at some point. In our current transition study, a relatively high number of patients with JIA-U at some point of the disease course were not transferred directly to the adult clinic. Chronic uveitis is the most common extra-articular manifestation of JIA [25], and it has been shown to increase the disease burden in young patients with JIA [26]. However, we did not focus more closely to the transition in uveitis in our study and more studies are needed to investigate the need for special transition procedures also in JIA-U.

We observed some differences in the transition practices between the different sites in the Nordic countries participating in this study. However, no study site had a special protocol in the transition process at the beginning of the twenty-first century. Since 2010, most of the sites participating in this study have developed their processes for transition, and this work will undoubtedly create better and more correctly focused transitions. In an article by Conti et al. [7], different kinds of transitional programmes around the world were introduced. They report that actual timing of transition varies; apart from the age (usually 16–18 years, up to 21), it can be linked to a patient’s educational stage or to the level of overall disease and treatment awareness, whereas in some centres, it can be assessed quite individually. The transition age of 16 could be seen as logical and congruent with the present JIA classification criteria [15]. Nevertheless, some unfavourable development, a functional decline in performance concerning relational reasoning for example, has been observed in adolescents [27]. During normal brain development however, there is an increased activation of prefrontal brain regions [28], and consequently, favourable progress towards more goal-directed behaviour with better control of emotions, social functioning and learning and less impulsivity and risk-taking behaviour [29]. These matters most certainly have an influence on adolescents’ competence in disease management. Ongoing development may further reduce the success of a transition and must be considered when establishing the appropriate age for transition.

The patients in the ‘directly transferred’ group were most often still using anti-rheumatic medications at the study visit. This is in line with the transition procedures in the Nordic countries, i.e., transition to an adult clinic commonly involves ongoing medication. In the Finnish study by Nousiainen et al. [30], the use of complementary and alternative medicine (CAM) was found to be relatively frequent among adolescents with JIA: 81% reported occasional use. In our study, CAM usage was not very frequent: around 12% of patients reported it, and there were no differences between the three transition groups. The higher CAM use among patients with JIA could be worrying; it might indicate the risk of neglecting the antirheumatic medication [31]. However, in our study patients with active disease did not use more CAMs compared to others.

In our study, we did not find a difference between the socioeconomic status and the level of physical exercise among the three study groups. The use of physiotherapy was highest among the patients transferred directly and referred later, but altogether, a relatively small number of patients—11 and 10%, respectively—received physiotherapy at the 18-year follow-up. There was a difference in physical functioning between the study groups, indicating lower performance in both transition groups. Nevertheless, the number of patients who exercised regularly was pleasingly high; around two-thirds in all groups did some exercise one to three times a week. There were no differences in mental functioning between the groups in our study. Possible concomitant psychiatric disorders are nevertheless important to take into consideration during the transition phase: patients with JIA and psychiatric diagnoses were found to have a lower quality of life, even when the disease was in remission [32].

The strength of this study is the long-term follow-up of this cohort in collaboration between the participating sites. This is the first study reporting transition practices at multiple sites in Nordic countries. However, further studies are needed to make comprehensive conclusions about the transition in each country. The lack of congruent transition protocols in the Nordic countries makes it challenging to draw definitive conclusions. It is also a limitation that data on transition practices during the study period was collected retrospectively. Further research on those patients not directly transferred might yield valuable information for the future development of transition protocols. The need for further studies concerning transition is obvious, especially among patients with JIA-associated uveitis.