一所教學醫院的死因診斷形式錯誤及準確度分析

本研究分析一所教學醫院民國82年1月1日至83年6月30日597張死亡診斷書有關死因診斷的形式錯誤及準確度。僅337張(56.5%)死因診斷形式正確；57張(9.6%)只填死亡機轉未填死因；58張(9.7%)填了數個死因但是因果順序不明；55張(9.2%)填入不明確單一死因；90張(15.1%)填入明確單一死因但是表達錯誤。若以是否影響原死因來分類，三分之一的死診形式錯誤會影響原死因的選擇。影響因素有：住院醫師比主治大夫較多錯誤(P=0.011)，內科醫師比非內科醫師較多錯誤(P=0.038)，死者年齡較大者較多錯誤(P=0.001)，死者在本院住院次數小於等於一次者較多錯誤(P=0.002)。以ICD-9基本分類表二位碼評估死因診斷準確度，82%的死因診斷是一致的。以ICD-9三位碼評估死因診斷準確度，72%的死因診斷是一致的。影響因素有：住院醫師比主治大夫較多不一致(P=0.020)，死者年齡較大者較多不一致(P=0.006)，死者在本院住院次數小於等於一次者較多不一致(P=0.026)。對主要死因進行排序，修正復排序與原本排序主要差別為早產低出生體重及意外事故兩個死因。以ICD-9三位碼來比較，惡性腫瘤排序影響不大（kappa值馱0.94），但是對腦血管疾病的排序有相當的影響（kappa值0.34）。原本有25張死因診斷填寫腦中風，回溯病歷資料復廢現其中23個案在病歷上都有明確記錄診斷是腦內出血或腦動脈阻塞。由病歷資料可得知研究樣本中有26個案是損傷或中毒，可是只有七位在死診上有提及外因。譬如研究樣本中有七位是因為跌倒或車禍造成顱內出血在該院接受手術，結果只有一張死因診斷提到外因。四張填腦內出血，兩張填肺炎。一位個案因分娩產後出血化亡，可是在死診上皆未提及，這會造成孕產婦死亡率的低估。本研究結果對編纂教村教育醫師正確填寫死因診新有相當大助益，對死因統計偏差的估計與調整也有所參考。

關鍵字

死亡診斷書；死因診斷；國際疾病分類；準確度

並列摘要

Five hundred and ninety-seven death certificates issued by a teaching hospital during January, 1993 and June, 1994 were reviewed. Only 337(56.5%) of them were formatively correct in cause-of-death diagnoses; 57(9.6%) had mechanism (s) of death only; 58(9.7%) wrote multiple causes of death, but the sequences were not clear; 55(9.2%) gave single cause of death, but not specific; 90(15.1%) gave single cause of death, but had error in expression. Residences compared with attending physicians (p=0.011), interalists compared with non-internalists (p=0.038), elder patients compared with younger patients (p=0.001), patients admitted to this hospital once and less than once compared with those admitted more than once (p=0.002) had higher percentage of making formative errors. Using ICD-9 basic table 2-digit coding, 49 1(82%) cause-of-death diagnoses were concordant with evaluaters'. Using ICD-9 3-digit coding, 432(72%) were concordant. Residences compared with attending physicians (p=0.020), elder patients compared with younger patients (p=0.006), patients admitted to this hospital once and less than once compared with those admitted more than once (p=0.026) had higher percentage of invalid. The formative errors and invalidity in cause-of-death diagnoses in this hospital did not change the ranking of leading causes of death except deaths due to premature and low-birth-weight and deaths due to external causes. The ranking of 3-digit malignant tumors had high concordance (kappa=0.94), but did bias the ranking of cerebrovascular diseases (kappa=0.34). 23 of 25 certificates coded stroke as cause of death could be further specified into cerebral infarction or intracerebral hemorrhage. According to the chart review, there were 26 injuries and poisonings, but only 7 of them mentioned external causes in certificates. There was a delivery-related death, but was not mentioned in the certificate. The results of this study could provide useful information for teaching clinicians how to writie correct cause-of-death diagnoses and give some hints in bias-adjusting when using cause-of-death statistics.

並列關鍵字

death certificate ； cause-of-death diagnosis ； ICD ； validity

國際替代計量

一所教學醫院的死因診斷形式錯誤及準確度分析

全文下載

主題瀏覽