Accuracy of Falls Screening Tools in Adult Patients With Cancer: A Systematic Review.

IMPORTANCE

Patients with cancer experience increased falls risk secondary to oncological treatment and cancer-related sequelae.

OBJECTIVE

Identifying diagnostically and prognostically accurate screening tool(s) for falls risk in populations with cancer is an important issue.

DATA SOURCES: Screening tests were identified in PubMed and CINAHL.

STUDY SELECTION: Two independent reviewers screened citations for inclusion.

DATA EXTRACTION AND SYNTHESIS: Data extraction was performed by 1 reviewer and verified by a second. Tests were investigated for clinical utility, validity, diagnostic accuracy, and predictive capacity. Recommendations for screening measures were formulated using predetermined criteria.

MEASURES: Falls risk screening tools were identified for populations with cancer.

RESULTS

Of 532 articles screened, 24 articles were included. Fifty-five variations of screening measures were identified, of which 47 had sufficient clinical utility. Twenty measures contained data on diagnostic accuracy or predictive capacity. No screening measure met all criteria to be highly recommended for both ruling in and ruling out falls risk currently (diagnostic accuracy) or in the future (predictive capacity). History of falls demonstrated good diagnostic accuracy for ruling in immediate falls risk (specificity 98.9%, positive predictive value 84.6%). A negative falls history was highly indicative of lower future falls risk status (negative predictive value 82.5% to 90.1%). Fear of falling demonstrated accuracy for ruling out immediate risk for falls (negative predictive value 87.0%, sensitivity 88.7%). Strong predictive capacity was demonstrated with the Timed "Up & Go" (TUG) Standard (sensitivity 93% at ≤7.8 s, specificity 95% at ≥11.35 s).

CONCLUSIONS

Based on these results, a history of falls plus either the TUG Standard for those with a history of falls or subjective report of fear of falling for those without a history of falls is recommended for risk screening in populations with cancer.

RELEVANCE: Different screening tools are required for immediate versus future falls risk and are setting dependent.