Validation of nursing documentation regarding in-hospital falls: a cohort study

In-hospital fall incidents are common and sensitive to nursing care. It is therefore important to have easy access to valid patient data to evaluate and follow-up nursing care. The aim of the study was to validate the nursing documentation, using a specific term in the registered nurses´ (RNs´) discharge note, regarding inpatient falls according to the outcome of a digitalized data extraction tool and the discharge note itself.

At a teaching hospital, 31,571 episodes of care were eligible for inclusion in this retrospective cohort study. A stratified sampling including five groups was used, two with random sampling and three with total sampling. In total, 1232 episodes of care were reviewed in the electronic patient record using a study-specific protocol. Descriptive statistics were used.

In total, 590 episodes of care in the study cohort included 714 falls. When adjusted for the stratified sampling the cumulative incidence for the study population was 1.9%. The positive predictive value in total for the data extraction tool regarding the presence of any fall, in comparison with the record review, was 87.4%. Discrepancies found were, for example, that the RNs, at discharge, stated that the patient had fallen but no documented evidence of that could be detected during admission. It could also be the opposite, that the RNs correctly had documented that no fall had occurred, but the data extraction tool made an incorrect selection. When the latter had been withdrawn, the positive predictive value was 91.5%. Information about minor injuries due to the fall was less accurate. In the group where RNs had stated that the patient had fallen without injury, minor injuries had actually occurred in 28.3% of the episodes of care.

The use of a specific term regarding fall in the RNs´ discharge note seems to be a valid and reliable data measurement and can be used continuously to evaluate and follow-up nursing care.