Skip to content
JMIR Aging Publication
Publications Science Press

JMIR Aging: Prognostic validity of LINDERA fall severity makes a traffic light for care possible for the first time

Diana Heinrichs
Diana Heinrichs |

In our retrospective cohort study, the prognostic validity of theFall Risk Score (FRS), a multifactorial fall risk scale generated from the mHealth app LINDERA Mobility Analysis, was investigated. The study included 617 older adults (857 observations) in German care facilities.

The FRS was collected at the first assessment (T1) and correlated with the frequency of falls at the follow-up assessment (T2). JMIR Aging, a highly renowned scientific journal, has now published the results.

 

fall-risk-score

Flowchart for participant selection

Study Results:

  • A quadratic regression model showed the best fit to the data and revealed a significantly strong positive correlationbetween the FRS at T1 and fall frequency at T2 (Spearman correlation coefficient = 0.960, P<.001).

  • Threshold values for the FRS were determined: An FRS of 45%, 32%, and 24% corresponded to the expectation of a fall within 6, 12, and 24 months, respectively.

  • The minimum clinically important differences (MCID) for FRS changes were calculated to quantify the clinical relevance of changes in the FRS.

  • Subgroup analyses showed that the FRS demonstrates particularly high prognostic validity in individuals with slower walking speed (<0.6 m/s), walking aid use, and an assessment interval of 120 days.

JMIR Aging Studie

JMIR Aging Study

In our retrospective cohort study, we examined the prognostic validity of the Fall Risk Score (FRS), a multifactorial fall risk scale generated from the mHealth app LINDERA Mobility Analysis. The study included 617 older adults (857 observations) in German care facilities.

The FRS was assessed at the initial evaluation (T1) and correlated with fall frequency at follow-up assessment (T2). JMIR Aging, a highly renowned scientific journal, has now published the results.

 

fall-risk-score
Flussdiagramm zur Teilnehmerauswahl

Study Results:

  • A quadratic regression model showed the best fit to the data and revealed a significantly strong positive correlationbetween the FRS at T1 and fall frequency at T2 (Spearman correlation coefficient = 0.960, P<.001).

  • Threshold values for the FRS were determined: An FRS of 45%, 32%, and 24% corresponded to the expectation of a fall within 6, 12, and 24 months, respectively.

  • The minimum clinically important differences (MCID) for FRS changes were calculated to quantify the clinical relevance of changes in the FRS.

  • Subgroup analyses showed that the FRS demonstrates particularly high prognostic validity in individuals with slower walking speed (<0.6 m/s), walking aid use, and an assessment interval of 120 days.

jmir-aging-studieJMIR Aging Study

Key Features of the LINDERA Mobility Analysis:

  • The app uses computer vision algorithms to extract precise gait parameters from smartphone videos recorded by nursing staff.
  • A standardized questionnaire integrated into the LINDERA fall prevention and mobility analysis app captures additional fall risk factors.
  • The FRS considers both intrinsic and extrinsic risk factors and weights certain factors (e.g., limited mobility, dizziness) more heavily.

Study Limitations:

  • The study focused on a specific population of older adults in residential care facilities.
  • The predictive power of the FRS varied between subgroups, suggesting that further refinements may be necessary for certain populations.

Our study supports the potential benefits of AI-based mHealth applications such as the LINDERA Mobility Analysis for fall risk assessment and prevention in older adults.

The FRS proved to be a promising prognostic tool, particularly for individuals with specific risk factors. The results emphasize the importance of personalized fall prevention and the need for further research to validate and optimize such mHealth tools.

To learn about the relevance of these results in nursing practice, read our press release. More about the results can also be found in our AI podcast on Notebook LM.

Diesen Beitrag teilen