The background rates in this manuscript resulted from unvalidated code extractions. The figures need further validation and substantiation following scrutinisation of the code extraction scripts. Part of the present heterogeneity may be due to differences in health care system, coding habits, type of coding library (READ and ICD9-CM are the most granular), financial incentives for coding and database type, but there may also be errors. A major concern in the provided rates is the inclusion of false positives in the numerator, may have occurred especially where less specific codes were included in the search criteria. This may have led to an overestimation of some of the rates. The here reported rates represent a first step toward a concerted initiative in a pan-European vaccine safety monitoring framework.

 

Comparison of our rates with those reported in the literature is difficult because of the heterogeneity in rates across the countries. For some but not all ESIs incidence rates have been reported in the literature.

For AIH a Norwegian 22 and Spanish study23 have reported incidence rates of 1.92/100,000 person years and 0.83/100,000 person years respectively. A study based in Sweden has reported an incidence of 0.85/100,000 person years24.

For anaphylaxis rates have been reported before, though subdivisions differed with ours 25, 26. In line with previous reports both ANAI and ANAK showed a decreasing incidence rate beyond the age of 60 and displayed sex differences26, 27. To allow more accurate comparison with previous reports it may be worthwhile to combine ANAI and ANAK into one ESI.

The incidence of BP has been reported to be 20.2/100,000 person years28. Slightly higher rates, which are more in line with the rates of validated BP in country 5, have been reported by an American group (~ 25/100,000 person years)29.
The large number of different codes needed to record all convulsions account for some of the variation between countries in the incidence rate of convulsions.There is no proper literature reference for convulsions either since incidence rates quoted in the literature suggest a wide range of values30 .

Some misclassification of DMD, ON and TM may have occurred, since these ESI share common symptoms or are part of the same disease entity. To avoid overlap between ESI the code lists were made mutually exclusive as much as possible (i.e. codes should occur in one ESI only). By separating them out, the individual rates of DMD, ON and TM may be lower than previously reported.

As with anaphylaxis, encephalitis was divided into encephalitis of viral origin and encephalitis of non-viral origin. Reviews of the literature have estimated the incidence of encephalitis to be ~6/100,000 person years31. The incidence rate of validated encephalitis in country 5 was much lower (respectively 0.43/100,000 PY and 1.17/100,000 PY). Together, viral and non-viral encephalitis in country 5 would have an incidence rate of 1.60/100,000 person years (0.43 plus 1.17).

The incidence of GBS has been widely reported to range between 1.1 and 1.832, 33. The variation in incidence rates of GBS was relatively small across countries which may have to do with the availability of specific codes in all libraries and the specific disease entity. Rates of GBS showed a typical age distribution in all countries, with a peak incidence between age 60 and 8032. The slightly higher rates observed in males has also been reported in the past33. The rate of validated GBS in country 5 was 1.71 per 100,000 person years, which is in line with previous reports. The rate of GBS across all countries in this study (except country 3) varied between 1.49 and 3.41 per 100,000 person years. Some false positive misclassification may have occurred.

Incidence rates of ON from the literature lie around 2 per 100,000/year33-36. Possible overlap with DMD may have occurred since history of DMD was not an exclusion criterion in our case definition. Although some countries showed slightly different age distribution, the peak incidence was mostly at 30-39 years of age. In line with the literature a higher risk of ON was observed for women compared to men37, 38. The rate of validated ON in country 5 was 1.82 per 100,000 person years, which well agrees with previous reports.
Being a composite of different diseases, the rates of thrombocytopenia may vary because of cumulative effects of the individual variations. The general trend was a steep increase in incidence beyond the age of 50. In the higher age categories, men were more likely to develop THRO than women.

TM is an extremely rare event reported to occur in about 0.83/100,000 person years39, 0.46/100,000 person years 40 and 0.15/100,000 person years41. One study reported an incidence of 1.1/100,000 person years42, which included MS related TM. Like ON, TM may have been underreported because it may be coded as MS (ie. DMD). The age distribution pattern of the incidence rate of TM was very similar to that of DMD.
Like thrombocytopenia, vasculitis in this report refers to multiple diseases. Incidence rates for vasculitis in the literature usually concern the individual types of vasculitis. The pattern observed in most countries is that the incidence rate has a J-shape age distribution with a slight tip in the youngest age group and a steep increase beyond the age of 50 with a female predominance.