Beyond fall risk assessment: A case–control study in an Urban Medical Center

Abstract Aims and objectives To examine the characteristics of patients that fell and compare them with patients that did not fall and to seek differences between the two groups that might help better predict falls in future patients. Background It has been estimated that between 700,000 and one million inpatient falls occur yearly in hospitals in the United States, which results in an increase in healthcare costs of over $19 billion dollars per year. Design This was a case–control study employing a retrospective analysis of inpatient electronic health records. It includes records from 160 patients who experienced a fall after the implementation of the Johns Hopkins Fall Risk Assessment Tool, and 160 records of patient with similar fall risk scores that did not fall. Methods All fall and nonfall patient data for the database were obtained by one research team member, while systematic random selection of nonfall patient records was performed by three research team members as described below. Each patient was assigned a unique study code number which was entered into the research database. The final sample size was 302 patients. Results Patients who did not receive lorazepam within 12 hr of the fall risk assessment were less likely to fall than patients who did receive lorazepam. A statistical relationship was also found between toileting at the time of the fall and age. Conclusions Better stratification of patient populations combined with astute nursing awareness may result in a further reduction in falls. Relevance to clinical practice The results indicate that the nursing assessment with respect to falls is critical to identifying fall‐prone individuals who may score as a low‐to‐moderate fall risk. In addition, the administration of lorazepam should cue the nurse that fall precautions be implemented regardless of scored risk.