Sample selection

Provincial staff drew eligible patients from provincial cancer registries based on eligibility criteria and linking to treatment data to confirm that treatment had occurred and end of treatment dates (defined for each disease type). The number of surveys was calculated for the adult sample such that 95% confidence intervals, for a percentage assumed to be 50%, would have width no more than ± 5% by disease site and province and assumed a response rate of 30%. Estimates of the number of eligible survivors for each province were based on disease site prevalence at the national level and incidence by province. As smaller provinces were not likely to have enough survivors within disease sites to achieve desired confidence interval precision, all eligible survivors were surveyed. For larger provinces and disease sites where the number of eligible survivors was anticipated to be greater than the required number, a random sample within the cancer disease site was chosen. The sample would be sufficient to achieve precision at least ± 3% by disease site for all provinces combined. A total of 40,790 survey packages were mailed out across the 10 provinces.