Table 2.

The Grading of Recommendations, Assessment, Development and Evaluation (GRADE) assessment on the quality of the evidence at the outcome level.

Study information		Certainty assessment					Sample size	Effect	Quality of the evidence
No. of studies	Study design	Risk of bias	Inconsistency	Indirectness	Imprecision	Publication bias		Absolute (95% confidence interval)
Fall prevalence—Alzheimer’s disease
21	RCT	S a	S b	NS	NS	NS	N = 5071	43.55% fall prevalence (38.78 lower to 48.33 higher)	⊕⊕○○
	OBS & NRS	S a	S b	NS	NS	NS			⊕○○○
Fall prevalence—Mild cognitive impairment
12	RCT	S a	S b	NS	NS	NS	N = 2537	35.26% fall prevalence (25.61 lower to 44.91 higher)	⊕⊕○○
	OBS & NRS	S a	S b	NS	NS	NS			⊕○○○
Number of falls—Alzheimer’s disease
9	RCT	S a	S b	NS	NS	NS	N = 358	1.30 number of falls (0.74 lower to 1.86 higher)	⊕⊕○○
	OBS & NRS	S a	S b	NS	NS	NS			⊕○○○
Number of falls—Mild cognitive impairment
5	RCT	S a	S b	NS	S c	NS	N = 321	0.77 number of falls (0.39 lower to 1.16 higher)	⊕○○○
	OBS & NRS	S a	S b	NS	S c	NS			⊕○○○

Quality of the evidence legend: ⊕○○○: Very low; ⊕⊕○○: Low; ⊕⊕⊕○: Moderate; ⊕⊕⊕⊕: High. RCT: Randomized Control Trial. OBS: Observational Study. NRS: Non-randomized Controlled Study. S: Serious. NS: Not Serious. Explanations:

^aMajority of studies relied on self-report data.

^bHigh heterogeneity amongst studies as evidenced by χ² and I².

^cSmall number of studies included.