A Mixed Effects Model Analysis for Inpatient Falls Using Health Record Data From 72 Hospitals.

Area of Science:

• Healthcare research
• Patient safety
• Clinical epidemiology

Background:

• Patient falls represent a significant safety concern in healthcare settings.
• The Morse Fall Scale (MFS) is widely used to assess fall risk.
• Understanding factors contributing to falls is crucial for effective prevention.

Purpose of the Study:

• To evaluate the relationship between patient falls and Morse Fall Scale (MFS) items.
• To examine the association of patient demographics, length of stay, and hospital site with fall risk.
• To identify specific risk factors for patient falls within a large health system.

Main Methods:

• Retrospective cohort study of 978,920 admissions across 72 hospitals.
• Logistic regression models were used to analyze MFS items, demographics, and length of stay.
• A mixed-effects logistic regression model incorporated patient-level covariates and hospital site as a random effect.

Main Results:

• A total of 6,531 falls were recorded.
• Increased fall risk was associated with MFS items (fall history, secondary diagnosis, gait, mental status) and demographics (male gender, increased age, longer stay).
• Decreased fall risk was linked to MFS items (ambulatory aids, IV therapy) and Hispanic ethnicity. Male gender and mental status showed an interaction effect, increasing fall odds by 3.16 times.

Conclusions:

• Significant variability in fall risk exists among hospitals, even after controlling for patient characteristics.
• The MFS may require re-evaluation, potentially removing IV lock and adding medication/equipment screening.
• Future research should explore hospital-specific factors and gender/ethnic differences to inform targeted fall prevention interventions.