Table 3

Modified risk of bias assessment of the 14 included studies using a modified* version of the8 Hoy 2012 tool8

Study country (Reference)RepresentationSamplingRandom selectionCase definitionReliability of tool/database validatedMethod of data collection *Prevalence periodNumerators and denominatorsOverall risk of bias
Australia and New Zealand9 HighHighLowLowHighLowLowLowModerate
Canada11 HighLowLowLowHighHighLowLowModerate
Canada10 HighHighLowLowHighLowLowLowModerate
Finland19 LowLowLowLowHighLowLowLowLow
Germany14 HighHighLowLowHighHighHighHighHigh
Italy18 HighHighLowLowHighLowLowHighModerate
Korea15 HighHighLowLowHighHighLowLowModerate
Japan11 HighHighLowLowHighLowLowLowModerate
Poland16 HighHighLowLowHighLowLowLowModerate
Spain17 HighHighLowLowHighLowLowHighModerate
Sweden20 LowLowLowHighHighHighLowLowModerate
Switzerland22 HighHighLowLowHighLowLowLowModerate
USA12 HighHighLowLowHighLowLowLowModerate
USA13 HighHighLowLowHighHighHighLowHigh
  • Excluded the assessment of ‘non-response bias’ and ‘data collection from subjects’ as these are not applicable.

  • *Retrospective studies were included in high risk.