[FALL PREDICTION AND FALL SEVERITY IN REHABILITATION HOSPITAL]

We tested whether the Farmer questionnaire is valid for fall risk assessment in Hebrew. We tested whether NDNQI (National Database of Nursing Quality Indicators) is valid for fall severity evaluation in Hebrew. Finally, we tested whether the Farmer and NDNQI are correlated.

Patients in rehabilitation hospitals are exposed to fall-risking conditions. Falls with severe outcomes can extend the hospitalization, and increase the workload on health systems. Fall risk assessment at the beginning of hospitalization is crucial for making supportive and preventive adjustments. The Israel Ministry of Health obliges using fall risk assessment at hospitalization. Nonetheless, fall risk assessment has not been validated in Hebrew, and has not been tested for prediction power of fall severity outcome.

Farmer measurement was validated in 1187 patients retrospectively, out of whom 288 had fallen during hospitalization. Twenty-five fall cases with varying severities were ranked by 47 staff members for their fall severity score. Non-parametric Spearman's correlation was tested between Farmer and NDNQI measurements.

Mean Farmer value of the falling group was larger than the mean Farmer value of the non-falling group (F=9.5, pv=0.002). Variability between raters was smaller than variability between conditions in NDNQI (ICC(2,1)=0.75). Farmer index was not correlated with NDNQI score (ρ=0.092, pv=0.118).

Farmer measurement is a valid tool for fall risk assessment in Hebrew. NDNQI is a valid tool for evaluation of fall severity. Farmer index is not predictive of fall severity.

There is a need for predictive measures of fall severity outcomes. We recommend using fall severity scores for ranking the intervention's success.