Emergency department visits due to severe falls: comparing patient self- reports and general practitioner records: A cross-sectional study.

Background: Older individuals who experience a severe fall are at high risk for long-term consequences and require structured follow-up care including risk assessments and secondary prevention (e.g. exercise). Accurate knowledge of pre-existing diagnoses is essential for tailoring interventions. However, identifying these diagnoses in emergency department (ED) is challenging due to time constraints, high workload, potential recall bias after a traumatic event, and limited electronic data exchange across healthcare settings in Germany. This study analyses the concordance between self-reported diagnoses of adults aged > 60 years presenting to the ED after a fall without hospitalisation and diagnostic information provided by their general practitioners (GPs).

Methods: Data from the SeFallED study conducted in Germany were analysed. To analyse concordance, 28 major diagnostic groups (e.g. heart diseases, cancer, lung diseases) were established. Cohen's Kappa assessed the agreement between self-reported and GP-reported diagnoses. Logistic regression identified associations between population characteristics (e.g., cognition, concerns about falling, age) and discordance between self-reported and GP-reported diagnoses.

Results: A total of 216 participants (mean age 75.2), with an average of five diagnoses per person, were included. Agreement was almost perfect (K = 0.81-1.0) for Parkinson's disease, substantial (K = 0.61-0.80) for diabetes mellitus and cancer, and moderate (K = 0.41-0.60) for heart and lung diseases. Other conditions showed fair, poor, or no agreement. Age, sex, BMI, cognition, concerns about falling, education, and living arrangements were associated with discordance. Only higher concerns about falling were linked to non-agreement for overall diagnoses.

Conclusion: Overall, there was poor agreement between self-reported and GP-reported diagnoses. Special attention should be given to older adults and individuals with high concerns about falling, as these factors predicted discordance. Participants were more likely to report diseases with more noticeable symptoms and frequent monitoring such as Parkinson's disease or diabetes mellitus. To optimise follow-up care, improving the accuracy of diagnostic information is essential, which may be facilitated by harmonizing patient information across electronic systems (e.g., electronic health cards). Improved communication between GPs and patients regarding existing illnesses is also crucial to enhance the accuracy of patient self-reports.

Trial registration: DRKS (Deutsches Register für klinische Studien, DRKS00025949, prospectively registered on 4th November, 2021).