Short note for paper published in Issue 2, Volume 9: ‘Characteristics and risk factors for accidental injury in Canada 1986-1996 …’

1. It has been perceived that the tone of the article may lead readers to believe that the data represent all injuries in Canada. The data only represent those collected within CAIRE, which is limited to the participating hospitals. The article should have said that the participating hospitals were in six provinces rather than seven. 2. The nature of the data with respect to the inconsistent data collection over the 10-year-period and the limited number records selected for inclusions due to the number of participating hospitals (at three large children’s hospitals and four general hospitals) and selected consumer products were briefly discussed in the methods section. However, due to the importance of these limitations, they could also have been identified in the discussion and the link between these limitations and the findings would have been clearer to the reader. 3. It has been suggested that the methods used for the logistic regressions were confusing. 4. In Figure 2 and Table 2 the age groups used are of varying widths and for consistency, it would have been appropriate to use the same age groups. Using the crude numbers rather than the injury rates may give the reader a misleading comparison. 5. The ‘injury rate’ as presented in this paper is based solely on the limited data included in this study and is not population-based. 6. The statement that injury claims more children’s lives than birth defects, cancer and infectious diseases combined may be correct depending on how ‘children’ are defined. It is true for 1–14-year-olds, and 0–20-year-olds but not for 0–14-year-olds. 7. The authors state that CHIRPP ‘found that 75% of injuries involving children aged 5–14 years were caused by bicycle injuries.’ Both the information and the references are wrong. References were given to an article on sports injuries in CHIRPP (reference 9), which specifically excluded cycling injuries, and to an article on basketball injuries in CHIRPP (reference 10). No data on bicycling injuries are presented in either paper. A correct statement would have been that about 6% of all CHIRPP records for children aged 5–14 report injuries are associated with bicycling. (Children aged 5–14-years do, however, account for about 75% of CHIRPP records of injuries associated with bicycles. This high percentage is not surprising given that 75% of CHIRPP data are from children’s hospitals.) 8. In Table 3 the 4 digit numbers should have been labelled more precisely as ‘ICD-9 external cause codes.’ The values should have been presented as E883.9-E886.9 (rather than 8839–8869). This specific group of E-codes is said to represent ‘all kinds of accidental falls,’ however, typically all unintentional falls includes the ICD-9 external cause codes E880.0-E888. 9. In Table 5 the difference between the two categories (‘other and unspecified places,’ ‘unspecified places’) is tenuous and probably they should have been combined in the analysis. Since the information comes from medical records, the high percentage (about 62%) of missing place values is not surprising.