期刊
JOURNAL OF CLINICAL EPIDEMIOLOGY
卷 53, 期 2, 页码 183-194出版社
ELSEVIER SCIENCE INC
DOI: 10.1016/S0895-4356(99)00136-5
关键词
validation; physician claims; outcome; injury; elderly
The sensitivity of using physician claims data for injury ascertainment was examined in a cohort of 1,181 elderly who were treated in the emergency department of one of 10 hospitals for injuries in 1993-1994. The clinical record of the type and date of injury was compared with diagnostic and procedure codes in the Quebec health insurance agency records of physician billing claims for the same patients. The proportion of patients correctly classified by claims data was determined for the exact date of injury and for a time window around the date of injury. The most common injuries were fractures (55.4%) and lacerations (19.3%), and 78.9% of injuries were fall related. Overall, the combination of treatment procedure codes and diagnostic codes provided the most sensitive measure of injury occurrence; a sensitivity of 67.3% for the exact date and 81.3% for an expanded data window (95.6% of injuries were within - 1 day to + 3 days of the injury date). Sensitivity varied by injury type form a low of 14% for abrasions to a high of 97.2% for hip fractures. The combination of diagnostic and procedure codes in physician claims is a sensitive indicator of some common injuries that would not be documented in hospitalization databases. (C) 2000 Elsevier Science Inc. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据