[Geriatric multimorbidity in claims data - part 2 : diagnoses of hospitals and diagnoses from physicians in the ambulatory setting].

Background: Due to demographics, characteristic multimorbidity in geriatric patients is resulting in increased social, medical, and healthcare challenges. Geriatric multimorbidity (GM) can be defined as the simultaneous occurrence of at least two diseases that require medical care with an interdisciplinary focus on independence in activities of daily living. Typical conditions of GM are, e.g., incontinence, cognitive impairment, frailty, and decubitus.
Material and Methods: Part 2 of this study is based on claims data of 240,502 AOK insurants (AOK is one of the major health insurance companies of the German statutory health insurance system) aged ≥ 60 years with at least one admission to a hospital with a geriatric ward. Geriatric conditions (GCs) were ascertained in two ways: diagnoses from physicians in the ambulant care setting and diagnoses in a hospital setting in 2008. A total of 15 GC were assessed using diagnoses based on ICD-10 codes (as per suggestion from scientific geriatric societies). An insurant was defined as a person with GM, if he/she had at least two GCs.
Results: The proportion of GCs in ambulant or inpatient diagnoses of 240,502 insurants varied significantly in most cases. For specific GCs, considerably higher proportions of ambulant diagnoses (e.g., pain, impairment of vision, or hearing) or for inpatient diagnoses (e.g., electrolyte or fluid metabolism disorders, malnutrition, incontinence) were identified. Only on rare occasions were small differences observed comparing the proportions of specific GCs in the diagnoses of the two different care sectors. This finding reduces considerably the accordance between the two care sectors with reference to the presence or absence of a GC for ambulant or inpatient diagnoses. The main agreement was with the non-coding of specific GCs, not with ambulant or inpatient diagnoses. Insurants with a geriatric hospital admission or certain care level (level ≥ 1) generally had higher proportions for specific GCs for inpatient and ambulant diagnoses than non-geriatric treated insurants or insurants without a certain care level. Of the geriatric treated insurants and those with certain care levels, 90% were characterized by the presence of GM for both ambulant or inpatient diagnoses. This percentage is remarkably higher than for patients who featured no geriatric treatment or had no certain care level.
Conclusion: The inclusion of ambulant diagnoses in addition to inpatient diagnosis offers comprehensive possibilities to identify insurants with GM in claims data. The contribution of the diagnoses of both care sectors for the identification of GC and GM varies with regard to attribute and insurant orientation. Furthermore, significant attribute-oriented overlap of insurants claiming geriatric treatments and insurants with certain care levels became visible, which can open new possibilities for simpler identification of a portion of patients with GM.