| Summary: | Neðst á síðunni er hægt að nálgast greinina í heild sinni með því að smella á hlekkinn View/Open Research has indicated low accuracy of information on the underlying cause of death, recorded on death certificates. When compared to autopsy data, discrepancy has been reported for between 25% and 56% of cases. An Icelandic study reported around 33% discrepancy. Here we report an investigation on the accuracy of information on the underlying cause of death on death certificates of Icelandic cancer patients. The aim was to look for differences in accuracy according to various causes of death, specifically between various cancer groups. Death certificates were cqmpared to autopsy reports for those patients from the following two study groups, who had been autopsied. A: All cancer patients diagnosed in 1971 to 1973 (n=326). B: All women diagnosed with breast cancer in 1974 to 1984 (n=118). In group A we found 25% discrepancy. For cancer deaths the discrepancy was 18%, for circulatory diseases it was 42% and for respiratory diseases 73%. Among cancer deaths, breast cancer had high accuracy, whereas it was low in cancers of the digestive tract. For group B, where three fourth of deaths were caused by breast cancer, the total discrepancy was around 9%. Sensitivity of the death certificates for breast cancer was 93% and the positive predictive value 97%. We conclude that the accuracy of death certificates depends on the cause of death and that for breast cancer in Iceland, the accuracy is satisfactory. Rannsóknir benda til þess að upplýsingar um undirrót dánarmeins á dánarvottorðum séu talsvert óáreiðanlegar. Þegar þær hafa verið bornar saman við upplýsingar á krufningarskýrslum hefur komið fram allt frá 25% og upp í 56% misræmi. Í íslenskri rannsókn kom fram um 33% misræmi. Hér er greint frá athugun á áreiðanleika upplýsinga um undirrót dánarmeins á dánarvottorðum íslenskra krabbameinssjúklinga. Tilgangurinn var að athuga hvort áreiðanleikinn væri ólíkur eftir dánarorsökum, og að skoða sérstaklega hina ýmsu flokka ...
|
|---|