This Article Full Text (PDF) Services Download to citation manager PubMed PubMed Citation

PREDICTIVE ACCURACY OF FALLS RISK SCREENING TOOLS

Royal Brisbane and Women's Hospital Herston, Queensland, Australia
Institute of Health and Biomedical Innovation Queensland University of Technology Kelvin Grove, Queensland, Australia

Address correspondence to Joan Webster, RN, BA, Nursing Director, Research, Royal Brisbane and Women's Hospital, Centre for Clinical Nursing, Bldg. 34, Butterfield St., Herston, Queensland, Australia 4029. E-mail: joan_webster{at}health.qld.gov.au

To the Editor:

It was pleasing to see the careful review of the accuracy of in-hospital falls risk screening tools in your June 2007 issue of the Journal, however, we believe the conclusion that the "STRATIFY may still be useful in particular settings"(1) is misleading. We are currently conducting a prospective cohort study that includes all patients aged 65 years and older who are admitted to a large acute care facility. Patients are initially screened for their falls risk by research personnel using the STRATIFY falls risk tool. The nurse caring for the patient, who is blinded to the STRATIFY score, is then asked if s/he believes the patient is at risk of falling. We currently have follow-up data for 389 patients. Using the standard cut-off score of two, the sensitivity of the STRATIFY is 58.8% and the specificity 82.9%. However, the positive predictive value is only 23.8%. This means that 76.2% of those who are predicted as being at high risk of falling did not fall. When nurses' clinical judgement is assessed, the sensitivity is higher (94.1%) but the specificity is low—36.6%. The false-positive rate for nurses is 87.5%. We currently do not routinely screen for falls risk but, if we did, considerable resources may be wasted on further assessment and interventions for patients who have only a small likelihood of falling. Additionally, classifying patients as low risk could lead to a false sense of security; over 40% of patients scoring less than two in our study did indeed sustain a fall.

Reference

1. Haines TP, Hill K, Walsh W, Osborne R. Design-related bias in hospital fall risk screening tool predictive accuracy evaluations: systematic review and meta-analysis. J Gerontol A Biol Sci Med Sci. 2007;62A:664-672.[Abstract/Free Full Text]

This Article Full Text (PDF) Services Download to citation manager PubMed PubMed Citation