Introduction:
Post-fall analysis of fall risk factors is an effective intervention in minimizing the fall risk in inpatients. Aim of the research: To identify significant fall risk factors in patients with neurological disease, who fell during hospitalization.

Material and methods:
Forty adult inpatients were included in a retrospective quantitative cross-sectional study, who fell in the neurological department in the period 01/2014–01/2020. The research protocol contains selected fall risk factors, the Morse Fall Scale (MFS), and the Barthel ADL Index. The Random Forest method was used to identify significant fall risk factors. ROC with AUC was used to assess the predictive value of fall risk factors.

Results:
The mean age of the sample was 70.50 ±15.20 years. The most common fall risk factors in the sample included the following: gait, balance and mobility disorders, polymorbidity ≥ 5 medical diagnoses, and polypharmacotherapy ≥ 5 drug classes. The mean MFS score was high (75.8 ±26.6), and the after-fall MFS score increased (97.8 ±22.5). The median score of the Barthel ADL Index indicated partial dependency in groups with and without fall in their history. The Random Forest identified significant falls risk predictors between the groups with and without fall in their history: fear of falling, number of drugs taken in 24 h, level of independence, and fall risk during admission. The predictive value of fall risk factors expressed by ROC curve was AUC = 0.842.

Conclusions:
The identification of significant factors in the after-fall analysis should lead to a revision of work procedures aimed at their prevention.