Table 2.

Psychometric Properties of Risk Assessment Tools Used in Included Studies

Published Tool	Psychometric Performance Source	Tool Description	Acute Care Dataa	U.S. Datab	Reliability Across Studies	Validity Across Studies
ADAPT Fall Assessment Tool15	Individual study15	ADAPT computerized information system, fall risk embedded into routine assessment documentation, allows customized interventions for specific patient risks, risk information integrated into care plan, report sheets, care conferences	Yes	Yes	n/a	Concurrent validity: risk assessment correlates 0.96 with Hendrich scale scores
Berryman Predisposition for Falling scale44 (applied to at-risk patients)	Review, data from1 study60	Assessed domains: age, mental status, length of stay, elimination, falling within the past 6 months, visual impairment, confined to chair, blood pressure	No	Yes	n/a	Face validity: most falls (3 VA patient care units observed for 3 months) were in patients aged ≥70
Hendrich, Hendrich II Fall Risk Model / Assessment4,29,33,36,56	Review, data from 1 study61	Assessed domains: mental state, gait and mobility, fall history, elimination, diagnosis, continence, mood, dizziness, weakness	Yes	n/a	n/a	Predictive validity: sensitivity 0.77, specificity 0.72
I'M SAFE Fall Risk Assessment Tool, Children's Hospital Denver45	Individual study62	Assessed domains: environment, history of falls, intravenous medications, orthopedic and muscular, rehabilitation and occupational and physical therapy, seizures andepilepsy	Yes	Yes	Internal consistency (α) 0.69	n/a
Innes Score; St Francis Memorial Hospital Standard Care Plan for the High-Risk Patient31,32,48	Systematic review, data from 1 study63	Assessed domains: previous trauma, disorientation, impaired judgment, sensory disorientation, muscle weakness, multiple diagnoses, language barrier	n/a	No	n/a	Predictive validity: sensitivity 0.89 (95% CI = 0.78–0.96), specificity 0.74 (95% CI = 0.72–0.75); PPV 0.07 (95% CI = 0.05–0.10), NPV 1.00 (95% CI = 0.99–1.00), OR = 23 (95% CI = 10.1–55.5)
Morse Falls Scale1,2,6,16,28,39	Systematic review, data from 4 studies64	Assessed domains: history of falling, presence of secondary diagnosis, use of ambulatory aids, administration of intravenous therapy, type of gait, mental status	Yes	n/a	n/a	Predictive validity: sensitivity 0.72–0.96, specificity 0.51–0.83
	Systematic review, data from 2 studies63	Score of 45 used as cutoff	Yes	No		Predictive validity: sensitivity 0.73–0.96, specificity 0.54–0.75, PPV 0.04–0.10, NPV 0.99–1.00.
	Systematic review, data from 3 studies65	6 items	Yes	n/a	Interrater agreement 0.96–0.98	Predictive validity: sensitivity 0.72–0.83, specificity 0.51–0.68
Schmid Fall Risk Assessment Tool49,51	Systematic review, data from 1 study63	Assessed domains: gait, confusion, assisted toileting, fall history, anticonvulsants; 5 items; score of 3 used as cutoff	n/a	Yes	n/a	Predictive validity: sensitivity 0.93 (95% CI = 0.80–0.98), specificity 0.78 (95% CI = 0.73–0.83), PPV 0.37 (95% CI = 0.27–0.47), NPV 0.99 (95% CI = 0.96–1.00), OR = 44.3 (95% CI = 13.2–172.4)
	Systematic review, data from 2 studies65	17 items; score 3 used as cutoff	Yes	Yes	Interrater agreement 0.88	Predictive validity: sensitivity 0.91–0.93, specificity 0.25–0.78
Timed Up & Go test17	Systematic review, data from 1 study65	Score 10–12 used as cutoff	No	n/a	Interrater agreement 0.56–0.99	Construct validity: judged as “good”
Unpublished tool, tool shown and risk factors reported3,13,14,18,20,21,23,26,30,34,37,38,43,47,50,53–55,57,58	n/a	n/a	n/a	n/a	n/a	n/a
Tool not described19,40,59	n/a	n/a	n/a	n/a	n/a	n/a
No risk assessment5,7–12,22,24,25,27,35,41,42,46,52	n/a	n/a	n/a	n/a	n/a	n/a

References in this table are found in the online supporting information.

^aTool tested in acute care setting.

^bApplied in U.S. organization.

n/a = not available, not applicable; VA = Veterans Affairs; ADAPT = Assess: Disorientation, Activity, Postmedication, and Toileting.