Swedish MPA publishes interim results of a cohort study assessing the risk of narcolepsy after vaccination with Pandemic influenza vaccine Pandemrix®Archived

ECDC comment

A four-fold increased incidence of narcolepsy in vaccinated children/adolescents compared to unvaccinated children was observed in this register-based cohort study.

On 29 March, 2011 the Swedish Medicinal Product Agency (MPA) published preliminary results of a cohort study assessing the risk of developing narcolepsy after vaccination with the monovalent A(H1N1) 2009 pandemic influenza vaccine – Pandemrix® [1]. This public health development summarises the findings and should be read along with an earlier version which ECDC published in early March [2].

The Swedish study followed the observations of an increase in the incidence of narcolepsy among children and adolescents in Finland and Sweden during the summer of 2010. The National Swedish authorities initiated this register-based study in the population of four large Swedish counties/regions covering 5.3M people (57% of the Swedish population). Exposure to Pandemrix® was assessed using pre-existing county/regional immunisation registries in the four counties, in which all vaccines administered at well-baby clinics, schools or health care institutions were documented. Pandemrix® was the only pandemic vaccine used in Sweden. Overall, 67.1% of children (born 1990 or later) and 51.0% of adults in this cohort were vaccinated. Cases of narcolepsy were detected using the county health care databases. The diagnostic ICD code G47.7 for narcolepsy was used as the case definition for incident new cases. A date cut-off October 1, 2009 was used to distinguish between old and new diagnoses since immunisation with Pandemrix began about that time. In all, 38 new cases were identified in the age group born 1990 or later in the vaccinated group compared to 6 in the unvaccinated group from October 1, 2009 through December 31, 2010. In the age group born after 1990 (20 years or older) the number of new cases was 26 in the vaccinated and 26 in the unvaccinated group.

Using a person-time analysis, the researchers estimate that the risk of narcolepsy among vaccinated children/adolescents was 4.06 cases / 100 000 person-years compared to 0.97 cases/ 100 000 person-years among unvaccinated yielding a relative risk of 4.19 (95% CI: 1.76-12.1). Among adults the incidence rates did not differ between vaccinated (1.16 / 100 000 person-years) or unvaccinated (0.96/100 000 person-years). The background incidence rates of narcolepsy at the national level have been 1.04 /100 000 among persons over 20 years of age and 0.46 among persons under 20 years of age during 2005-2008.

ECDC comment:

A four-fold increased incidence of narcolepsy in vaccinated children/adolescents compared to unvaccinated children was observed in this register-based cohort study. The results and conclusions are consistent with the interim results of the study in Finland; however the effect estimates are somewhat lower. These results strengthen the possibility of a link between Pandemrix® and narcolepsy, even if it is obvious that both studies share some important design and analysis problems. In the Swedish study– case validation is still ongoing, time of onset of disease is not presented, and assessment of confounding factors is as yet not possible.

As of 22 March, 2011 a total of 229 cases of narcolepsy following vaccination with Pandemrix have been reported to the EMA Eudravigilance database. The final results from this study, the study in Finland and the European ECDC-VAESCO case-control study are expected in the coming six months. If the association is confirmed in the current on-going studies further studies addressing possible risk factors such as environmental, genetic and autoimmune are needed. In addition, a possible biological plausibility needs to be explored in collaboration with the European Narcolepsy Network experts.

ECDC understands that very little Pandemrix® is now being used in Europe currently and emphasises that among licensed vaccines it is, with few exceptions, combination seasonal influenza vaccines (rather than monovalent vaccines) that should be used to prevent seasonal influenza. This follows from the fact that seasonal influenza involves the circulation of more than one influenza virus while by definition monovalent vaccines will only prevent one of these.


1.  Swedish Medicines Agency press release, 29 March 20102.

2. ECDC public health development Reports of cases of narcolepsy and cataplexy in children and adolescents in Finland, Iceland, and Sweden, March 2nd 2011