Von Magnus 1984.

Methods	Study design: interrupted time series. National surveillance data from 1935 to 1980 for cases of confirmed paralytic poliomyelitis. Recollection of paired serum samples for type 1 antibodies Setting: Denmark
Participants	Sample size: 2.7 million people. In connection with the vaccination, paired serum samples from 300 individuals were collected and examined for type 1 antibodies Age: younger than 40 years of age
Interventions	Before: III from 1961‐1967 After: IIIOOO from 1968‐1973 In the years between 1962 and 1967 the polio vaccination program in Denmark consisted of three injections of inactivated poliovirus vaccine (IPV). From 1968 onwards the polio vaccination program changed to include three injections of IPV when children are five, six, and 15 months of age and three vaccinations with tOPV administered at the age of three, four, and five years.
Outcomes	Number of paralytic poliomyelitis cases. Cases that occurred in close relation to the vaccinations were considered as confirmed and the other as doubtful Percentage of persons, by age in years, with antibodies to type 1, 2 and 3 poliovirus, 1954, 1961, and 1973 Geometric mean titter of serum antibodies for all 3 types Acceptance rate Follow‐up: 13 years
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
ITS: protection against secular changes	Low risk	Comment: intervention appears to occur independently of other changes over time
ITS: data were analysed appropriately	High risk	Comment: crude numbers
ITS: reason given for the number of points pre‐ and post‐intervention	Low risk	Comment: rationale for the yearly points is justified in the context of national surveillance system
ITS: shape of the intervention effect was specified	Unclear risk	Comment: data were not analysed as an ITS but we re‐analysed the data.
ITS/UBA: blinded assessment of primary outcome(s)	Low risk	Comment: The outcomes 'paralytic poliomyelitis cases' and 'antibody titres' are objective. The source were national information systems.
ITS/UBA: reliable primary outcome measure(s)	Low risk	Number of paralytic poliomyelitis cases. Cases occurred in close relation to the vaccinations were considered as confirmed and the other as doubtful. Antibody titres is a more objective outcome.
ITS: intervention unlikely to affect data collection	Unclear risk	Comment: sources and methods of data collection were the same before and after the intervention. The outcome variables are objective measures.
ITS: completeness of data set	Unclear risk	Comment: unclear report
Conflict of interest	Low risk	Study supported by the Epidemiology Department and the Enterovirus Department,Statens Seruminstitut, Copenhagen, Denmark.