Department of General Practice and Primary Health Care, University of Auckland, Auckland, New Zealand.
J Clin Epidemiol. 2013 Oct;66(10):1135-43. doi: 10.1016/j.jclinepi.2013.05.001. Epub 2013 Jul 13.
Self-reported health and disease status is a common method used in epidemiologic studies and surveys involving younger populations, but its reliability in octogenarians is unknown. The objective of the present study was to examine the level of agreement between self-reports and medical records on specific cardiovascular diagnoses.
The Life and Living to Advanced Age: a Cohort Study in New Zealand recruited 937 octogenarians in New Zealand. Standardized questionnaires were administered to determine sociodemographic status and medical history. Diagnoses of myocardial infarction (MI), stroke, congestive heart failure (CHF), and hypertension were compared between self-reports and two combined sources of medical records.
There were moderate levels of agreement between self-reports and medical records for MI, stroke, and hypertension (κ = 0.43-0.45) and low levels for CHF (κ = 0.19). The proportion of discordance for MI, stroke, CHF, and hypertension was 16%, 12%, 22%, and 27%, respectively. Adjusting for socioeconomic-demographic status and cognitive function, the number of comorbidities is highly associated with agreement between self-reports and medical records (P < 0.01). Gender, socioeconomic status, and cognitive function were also related to agreement between self-reports and medical records, but the strength of association was ethnic specific.
Self-reported information on specific cardiovascular conditions has only modest reliability in octogenarians and is associated with number of comorbidities.
自我报告的健康和疾病状况是涉及年轻人群的流行病学研究和调查中常用的方法，但在 80 岁以上人群中其可靠性尚不清楚。本研究的目的是检查自我报告与病历在特定心血管诊断上的一致性程度。
新西兰的“生活与长寿：新西兰队列研究”招募了 937 名 80 岁以上的新西兰人。进行了标准化问卷调查以确定社会人口统计学状况和病史。比较了自我报告和两种合并的病历来源中心肌梗死（MI）、中风、充血性心力衰竭（CHF）和高血压的诊断。
自我报告与病历在 MI、中风和高血压方面具有中度一致性（κ=0.43-0.45），而 CHF 的一致性较低（κ=0.19）。MI、中风、CHF 和高血压的不一致比例分别为 16%、12%、22%和 27%。调整社会经济人口统计学状况和认知功能后，合并症的数量与自我报告和病历之间的一致性高度相关（P<0.01）。性别、社会经济地位和认知功能也与自我报告和病历之间的一致性有关，但关联强度具有种族特异性。
80 岁以上人群中特定心血管疾病的自我报告信息可靠性仅适中，与合并症数量有关。
