Predictive properties of various assessments for fall risk detection, including AI-supported approaches
Conditions
Interventions
Group 1: The study is divided into different data collection points.
t0 Screening and Information
At the beginning, participants receive verbal information about the study, along with the distributio
Sponsors
Charité Universitätsmedizin Berlin
Eligibility
Sex/Gender
All
Age
18 Years to No maximum
Inclusion criteria
Inclusion criteria: - Capacity to provide Informed Consent - Minimum duration of hospitalization of 48 hours - Mobility score up to Jones-4a
Exclusion criteria
Exclusion criteria: - Age < 18 years - Immobility (Jones-4b) - Legal guardianship
Design outcomes
Primary
| Measure | Time frame |
|---|---|
| The Area Under the Receiver Operating Characteristic Curve (AUROC) of each model serves as a measure of the overall performance of the assessments. Hypotheses for the Primary Endpoint: - Null Hypothesis (H0): There is no difference in the AUROC between all four models. - Alternative Hypothesis (H1): There is a difference in the AUROC between at least two of the four models. | — |
Secondary
| Measure | Time frame |
|---|---|
| 1. Recall: Proportion of correctly identified patients at risk of falling. 2. Precision: Proportion of correct fall risk predictions among patients classified as high-risk. 3. Accuracy: Proportion of overall predictions that were correct. 4. F1 Score: Harmonic mean of Precision and Recall. 5. Fall event 6. Injury aeverity according to Joint Commission Standards 7. Fall prevention measures | — |
Countries
Germany
Contacts
Public ContactRobert Klebbe
Charité Universitätsmedizin Berlin
Outcome results
None listed