Ma'u Etuini, Cullum Sarah, Mukadam Naaheed, Davis Daniel, Rivera-Rodriguez Claudia, Cheung Gary
Department of Psychological Medicine, University of Auckland, Auckland, New Zealand.
Te Whatu Ora Waikato, Hamilton, New Zealand.
Lancet Reg Health West Pac. 2024 Oct 21;52:101216. doi: 10.1016/j.lanwpc.2024.101216. eCollection 2024 Nov.
Issues of under-diagnosis and under-coding of dementia in routinely collected health data limit their utility for estimating dementia prevalence and incidence in Aotearoa New Zealand (NZ). Capture-recapture techniques can be used to estimate the number of dementia cases missing from health datasets by modelling the relationships and interactions between linked data sources. The aim of this study was to apply this technique to routinely collected and linked health datasets and more accurately estimate the incidence of dementia in NZ.
All incident cases of dementia in the NZ 60+ population were identified in three linked national health data sets-interRAI, Public hospital discharges, and Pharmacy. Capture-recapture analysis fitted eight loglinear models to the data, with the best fitting model used to estimate the number of cases missing from all three datasets, and thereby estimate the 'true' incidence of dementia. Incidence rates were calculated by 5-year age bands, sex and ethnicity.
Modelled estimates indicate 36% of incident cases are not present in any of the datasets. Modelled incidence rates in the 60+ age group were 19.2 (95% CI 17.3-22.0)/1000py, with an incident rate ratio of 1.9 (95% CI 1.9-2.0) per 5-year age band. There was no difference in incidence rates between males and females. Incidence rates in Asian (p < 0.001) but not Māori (p = 0.974) or Pacific peoples (p = 0.110) were significantly lower compared to Europeans, even after inclusion of missing cases.
This is the first study to provide estimates of age 60+ dementia incidence in NZ and for the four main ethnic groups and suggests over a third of incident dementia cases are undiagnosed. This highlights the need for better access to dementia assessment and diagnosis so that appropriate supports and interventions can be put in place to improve outcomes for people living with dementia and their families.
Nil.
常规收集的健康数据中痴呆症诊断不足和编码不足的问题限制了它们在估计新西兰（NZ）痴呆症患病率和发病率方面的效用。捕获-再捕获技术可用于通过对链接数据源之间的关系和相互作用进行建模，来估计健康数据集中遗漏的痴呆症病例数。本研究的目的是将该技术应用于常规收集和链接的健康数据集，更准确地估计新西兰痴呆症的发病率。
在三个链接的国家健康数据集——interRAI、公立医院出院记录和药房记录中，识别出新西兰60岁及以上人群中所有痴呆症的新发病例。捕获-再捕获分析对数据拟合了八个对数线性模型，使用最佳拟合模型来估计所有三个数据集中遗漏的病例数，从而估计痴呆症的“真实”发病率。发病率按5岁年龄组、性别和种族计算。
模型估计表明，36%的新发病例在任何数据集中都未出现。60岁及以上年龄组的模型发病率为19.2（95%CI 17.3-22.0）/1000人年，每5岁年龄组的发病率比为1.9（95%CI 1.9-2.0）。男性和女性的发病率没有差异。即使纳入遗漏病例后，亚洲人的发病率（p<0.001）与欧洲人相比仍显著较低，但毛利人（p = 0.974）和太平洋岛民（p = 0.110）并非如此。
这是第一项提供新西兰60岁及以上痴呆症发病率以及四个主要种族群体发病率估计的研究，表明超过三分之一的痴呆症新发病例未被诊断出来。这凸显了更好地获得痴呆症评估和诊断的必要性，以便能够实施适当的支持和干预措施，改善痴呆症患者及其家庭的结局。
无。
