The safety of safety research: the case of patient fall research.

Most fall intervention programs consist of 2 components: fall risk prediction instruments to identify the patient who is likely to fall, and fall intervention strategies to prevent the patient from falling or to protect the patient from injury should a fall occur. While critical to the effectiveness of a fall intervention program, many of the fall risk prediction instruments have been criticized for their failure to accurately identify the fall-prone patient. In this article, in the context of the validity assessments conducted on the Morse Fall Scale, the research conducted in the past 2 decades on fall risk assessment is critiqued. Some fall prediction research is based upon invalid assumptions and/or errors in design, both in the development of risk scales and in the evaluation of these instruments. Many of these instruments have been constructed with inappropriate reliance on face validity, have been evaluated inadequately, or have been tested in the clinical setting using an invalid design. Finally, improper use of fall scales in the clinical area may increase the risk of injury to the patient. The author concludes that much nursing research on patient falls does not contribute to improved patient safety.