某醫學中心癌登資料庫原發性上泌尿道上皮侵襲癌ICD診斷碼之驗證研究及病人存活分析

Abstract

研究目的 1.驗證林口長庚癌症登記資料庫中原發性上泌尿道上皮侵襲癌ICD診斷碼之正確性。 2.分析已驗證為原發性上泌尿道上皮侵襲癌病人之存活情形及影響因素。 方法 申請並通過林口長庚醫院機構倫理審查委員會核准後，取得林口長庚醫院癌症登記資料庫中1997-2001年之原發性上泌尿道上皮侵襲癌之病人資料檔案紀錄，經由詳細之病歷回顧，來驗證資料庫中所登錄之ICD編碼是否真為原發性上泌尿道上皮侵襲癌。已驗證真為原發性上泌尿道上皮侵襲癌病人進一步紀錄其人口學及腫瘤特徵，並比較這些變數在病人存活之差異，以及何種變數為病人存活之顯著因子。 結果 1170位病人之原發性上泌尿道上皮侵襲癌之登錄紀錄，其中1132位病人確為原發性上泌尿道上皮侵襲癌病人，陽性預測率為96.8 %；僅38位病人，屬誤登為原發性上泌尿道上皮侵襲癌，其中包含16位病人有下泌尿道上皮癌、2位病人有原發性上泌尿道上皮原位癌、2位病人有復發性上泌尿道上皮癌、3位病人有不明部位之泌尿道上皮癌以及15位病人有非泌尿道上皮癌之其他泌尿疾病。上泌尿道癌上皮侵襲癌病人全死因存活影響因素之多變量分析顯示診斷時年齡層愈高、男性病人、同時侵犯腎臟及輸尿管者、未接受手術及腫瘤期別分類為區域或轉移癌者，不利於病人之全死因存活；而癌症別存活影響因素之多變量分析顯示診斷時年齡層愈高、男性病人、病人診斷於1997-2003年、未接受手術及腫瘤期別分類為區域或轉移癌者，不利於病人之癌症別存活。 結論 林口長庚癌症登記資料庫對原發性上泌尿道上皮侵襲癌具極高之陽性預測率，為可信之次級資料庫，但仍需在登錄上詳細確認診斷，避免將其他泌尿疾病誤登為原發性上泌尿道上皮侵襲癌，且需注意泌尿道上皮癌是否為復發性、位於下泌尿道或原位癌，以減少誤登。上泌尿道癌上皮侵襲癌病人，如為女性、腫瘤期別分類為局部/區域癌者或診斷時年齡層低者，皆應接受手術治療，可預期有極佳之存活。

Objectives 1. To validate the diagnosis of ICD codes for primary upper urinary tract urothelial invasive carcinomas (UTUC) of cancer registry database of Linkou Chang Gung Memorial Hospital 2. To analyze survival of validated primary UTUC patients and its predictive factors Methods Primary UTUC records in cancer registry database of Linkou Chang Gung Memorial from 1997 to 2011 were obtained after institutional review board approval. The medical chart of each patient was reviewed for validation for primary UTUC of ICD codes in cancer registry database. Demographic features and tumor characteristics were recorded for validated primary UTUC patients. Multivariate analyses using Cox proportional hazard model was performed for predictive factors of survival for primary UTUC patients. Results There are 1170 primary UTUC patients in cancer registry database of Linkou Chang Gung Memorial Hospital from 1997 to 2011 who have sufficient information in medical records for validation. Of the 1170 patients, 1132 patients have been validated as having primary UTUC, and the positive predictive value of this cancer registry database for primary UTUC is 96.8 %. There are 38 patients with false positive diagnoses of primary UTUC, including 16 patients of lower urinary tract urothelial carcinomas (UC), 2 patients of upper urinary tract carcinoma in situ, 2 patients of ureteral stump cancer, 3 patients of UC with unclear location, and 15 patients of urinary tract diseases other than UC. Multivariate analyses of overall survival of validated primary UTUC patients show older age group at diagnosis, male, synchronous UC in the renal pelvis and ureter, no surgery, stage grouping as regional or distant cancer are significant factors for death due to all causes. Multivariate analyses of cancer specific survival of validated primary UTUC patients show older age group at diagnosis, male, cancer diagnosed at 2004-2011, no surgery, and stage grouping as regional or distant cancer are significant predictors for death due to UC. Conclusion Cancer registry database of Linkou Chang Gung Memorial Hospital is an accurate secondary database for primary UTUC by its high positive predictive value. In order to decrease false positive diagnosis in cancer registry database for primary UTUC, other urinary tract diseases other than UC should be validated for their diagnoses and UC should be checked for whether it is primary or recurrent, located at upper or lower urinary tract and for carcinoma in situ or invasive carcinoma. For UTUC patients of females, staging grouping as localized or regional cancer, and younger age group at diagnoses, they should undergo surgery and a high probability for survival could be thereafter expected.