Discussion

Deviating from other reports this study uses an age-specific and sex-specific description of the incidence. This allows a differentiated view of the frequencies of new cases according to life stage and gender. Current findings indicate that between 5% and 45% of PsO patients develop a PsA over the course of the disease.9 10 As shown in figure 3, the incidence rises continuously with increasing age. Most of the new cases occur between the ages of 50 and 59. Consequently, the risk of developing PsA is highest at this age.

As mentioned before, there is a direct correlation between the prevalence of plaque PsO and the occurrence of PsA. The higher the prevalence of PsO in an observed population, the higher the probability to develop PsA.2 Against this background, the drastic decrease in the incidence of PsA before reaching the age of 60 remains unaccounted for since the prevalence of plaque PsO fails to provide an adequate explanation (figure 1). Therefore, further interpretation of the varying incidence within the different age groups appears necessary.

As noted already, the incidence rises continuously over the 3-year survey period (figure 3). This trend has already been observed for the prevalence of these data sets and was previously reported and discussed by Sewerin et al10 (figure 2). Whereas other studies describe a stable course of the incidence of PsA.3 11 Possible explanatory approaches that may account for the differences are an improvement in diagnostic abilities, changed coding behaviour of ICD codes and an increased awareness of the disease.12 However, a real increase in the incidence of PsA should be considered as well.

Unlike previous studies, a difference in the age-specific incidences between males and females was found in this cohort.2 While both sexes are affected by PsO to a similar degree, women appear to be more likely to develop PsA during the course of their disease. This fact can be especially observed in a postmenopausal age, so that a hormonal cause remains to be discussed.

With an average age-standardised incidence of approximately 17 per 100 000 py, the values of this survey appear to be elevated compared with previous studies. A meta-analysis performed by Scotti et al pointed out the large heterogeneity in this field. Their analysis calculated a random effect pooled PsA incidence rate of 8.3 per 100 000 py. However, the incidences of the included studies ranged from 3.0 to 41.3 cases every 100 000 py.2 While there are limited data on the incidence of PsA in Germany for comparison, recent population-based studies confirm that the incidence is increased in the present study population.3 4

The large heterogeneity of data emphasised by Scotti et al can be explained by different study designs and cohorts. In contrast to other studies, this report stands out by combining a large population-based study and a new approach to incidence estimation. The present data sets offer a great potential for comprehensive analysis of the epidemiology of PsA in Germany. In order to constitute the population as a whole, this study included children and adolescents in addition to the adult population. This study design represents a unique feature in the field of PsA epidemiology.

However, the results of the report and the comparability with other studies should be viewed with caution due to different diagnostic criteria. While this study uses ICD-10-coded diagnoses to establish the diagnosis, other studies make use of the Classification Criteria for Psoriatic Arthritis (CASPAR criteria) or refer to a concurrent joint diagnosis of arthritis and PsO.2 Since any officially registered physician in Germany can make diagnoses of chronic conditions, general practitioners as well as specialists, the data are based on both primary and secondary care records. A classification as ascertained cases is required in the coding process. An estimation of completeness of the coded diagnoses has not been undertaken as the sensitivity of claims data is considered very good.4 Evidence for this assumption is provided by a paper that examines the accuracy of claims data in a similar data record of type 2 diabetes patients.13

The estimated incidence differs from the results published for these data sets in 2019.10 Due to missing values, only an approximate calculation could be performed at that time. Consequently, the approach now chosen is to be regarded as the more valid one, since the methodological weakness has been eliminated.

The difference in the observation period of prevalence and incidence arises from the new approach to incidence estimation. Further explanations can be found in the related paper by Brinks et al.14

As shown in figure 2, an impressive decrease in the age-specific and sex-specific prevalence of PsA was observed after the age of 60. A possible explanation for the change in the prevalence is the correlation between PsA and additional comorbidities, which coincide with an increased morbidity and mortality.15 The extent of the decrease also indicates that the incidence is not high enough to compensate the mortality excess of PsA patients.14 The data we have available does not allow to examine factors that drive excess mortality, for example, possible comorbidities. We can only describe that excess mortality exceeds the incidence, but unfortunately, we cannot provide reasons why this is the case.

Regardless of the observations made, it is important to stress that the total number of patients suffering from PsA is likely to be much higher than the numbers of identified cases.16 A correlation between joint complaints and PsO is not always taken into account.

However, the diagnosis of PsA is of considerable relevance to the prognosis and burden of disease. Accounting for the frequency and severity of PsA, the socioeconomic aspect should not be ignored either. For this reason, intensified early detection and secondary prophylaxis should be given greater priority in the care of PsO patients.