Diagnosestellung und Diagnosecodierung von Demenzen im Spiegel der Abrechnungsdaten der gesetzlichen Krankenversicherung

Abstract

Studies and care concepts regarding dementia on the basis of diagnosis codes in ambulatory claims data should be interpreted with great caution.

Dementia in health claims data: The influence of different case definitions on incidence and prevalence estimates

“…For example, the AD proportion in the German outpatient specialist sector was estimated to be 25%-40% in a survey study (Lohmann et al, 2017). Even lower proportions of around 10% were previously found in health claims data (Albrecht et al, 2018(Albrecht et al, , 2019Kaduszkiewicz et al, 2014). Cho et al (2014), who also examined proportions of dementia diagnoses in healthcare data and distinguished between three dementia categories similarly to ours, found 25% AD in incident cases and 31.9% after 1 year of follow-up.…”

“…For example, the AD proportion in the German outpatient specialist sector was estimated to be 25%–40% in a survey study (Lohmann et al., 2017 ). Even lower proportions of around 10% were previously found in health claims data (Albrecht et al., 2018 , 2019 ; Kaduszkiewicz et al., 2014 ). Cho et al.…”

“…Earlier results from the Rotterdam study suggest that AD accounts for the majority of all dementia cases with an estimated 70%, while approximately 15% suffer from VD (Ott et al., 1995 ). The epidemiology of dementia can be considered well studied, and in recent years, results from field studies have been increasingly supported by findings based on health claims data (Doblhammer, Fink, & Fritze, 2015 ; Doblhammer, Fink, Zylla, & Willekens, 2015 ; Kaduszkiewicz et al., 2014 ; Stock et al., 2018 ). Claims data offer a number of research opportunities which can be realized only at very high costs and efforts in field studies, such as the description of multidisciplinary and multisectoral care, including precise documentations of drug treatment patterns in a large number of patients over a long period of time (Gansen, 2018 ).…”

Claims data-based analysis of the influence of individual and regional characteristics on the utilisation of long-term care by people with dementia in Baden-Wurttemberg, Germany

“…Furthermore, claims-based data and ICD-codes do not allow differentiation between different stages of dementia severity. These limitations have already been discussed sufficiently in the literature [28, 32, 53]. However, the study design using secondary data reduced the risk of selection bias and enabled us to investigate a large study population that included groups that are difficult to access and which could not be achieved using primary collected data.…”

“…Based on the literature, we included individuals with diagnosis of dementia, which was defined using the International Classification of Diseases (ICD) Codes F00, F01, F02, F03, F05.1, G30, G31.0 and G31.82 [3, 32–34]. Several measures were taken in order to enhance the validity of diagnoses.…”