Skip to main content
. 2021 Nov 26;12:761661. doi: 10.3389/fphar.2021.761661

TABLE 2.

Quality of evidence in the included systematic reviews based on GRADE.

References Outcomes No.of Risk of bias Inconsistency Indirectness Imprecision Publication bias No. of participants Pooled effect size 95% CI Quality of evidence
Intervention Control
Yang et al. (2021) ADL 9 −1 a 0 0 0 −1 b 283 279 MD = −1.30 (-2.27, -0.33) Low
MMSE 10 −1 a −1 c 0 0 −1 b 317 313 MD = 1.17 (0.15, 2.20) Extremely low
ADAS-Cog 4 −1 a −2 d 0 −1 e −1 b 123 121 MD = −4.69 (-16.14, 6.75) Extremely low
TCM-SSS 4 −1 a −2 d 0 −1 e −1 b 108 104 MD = −4.70 (-8.97, -0.43) Extremely low
AR 7 −1 a NR 0 −1 e NR NR NR NR NR Not evaluated
Wang et al. (2020) ADL 35 −2 f −2 d 0 −1 e −1 b 1,006 901 MD = −1.00 (-4.56, 2.56) Extremely low
MMSE 29 −2 f −2 d 0 0 −1 b 1,348 1,229 MD = 1.82 (0.91, 2.72) Extremely low
AR 8 −2 f NA 0 0 −1 b NR NR NR NR Not evaluated
TER 39 −2 f 0 0 0 −1 b 1784 1,656 RR = 1.14 (1.03, 1.25) Extremely low
Li et al. (2019) ADL 5 −1 a −1 c 0 −1 e −1 b 187 169 SMD = 0.27 (-0.41, 0.95) Not evaluated
MMSE 8 −2 f −2 d 0 0 −1 b 247 221 WMD = 1.87 (0.76, 2.98) Extremely low
AR 1 −2 f NA 0 NR NR 24 22 NR NR Not evaluated
TER 6 −2 f 0 0 0 −1 b 222 212 RR = 1.29 (1.16, 1.44) Extremely low
Qin et al. (2019) ADL 6 −1 a −2 d 0 −1 e −1 b 224 180 WMD = −0.52 (-0.76, -0.28) Extremely low
MMSE 8 −1 a −1 c 0 0 −1 b 320 280 WMD = 0.48 (0.31, 0.64) Extremely low
ADAS-Cog 3 −1 a −1 c 0 −1 e −1 b 92 90 WMD = −0.41 (-0.71, -0.11) Extremely low
AR 5 −1 a −1 c 0 NR NR 192 150 NR NR Not evaluated
TER 9 −1 a 0 0 0 −1 b 371 330 OR = 3.37 (2.35, 4.83) Low
Yang et al. (2019) ADL 5 −2 f −1 c 0 −1 e −1 b 165 161 WMD = 0.81 (-2.46, 0.83) Extremely low
MMSE 6 −2 f 0 0 0 −1 b 247 241 WMD = 1.34 (0.74, 1.95) Extremely low
AR 7 −2 f −1 c 0 −1 e NR NR NR NR NR Not evaluated
TRE 7 −2 f 0 0 0 −1 b 205 201 OR = 2.28 (1.47, 3.52) Extremely low
Shi et al. (2019) ADL 4 −2 f −1 c 0 0 −1 b 283 277 MD = −3.40 (-4.92, -1.87) Extremely low
MMSE 5 −2 f −2 d 0 0 −1 b 314 308 MD = 2.69 (2.13, 3.26) Extremely low
AR −2 f −1 c 0 −1 e −1 b NR NR NR NR Not evaluated
TER 3 −2 f 0 0 −1 e −1 b 77 67 RR = 1.27 (1.08, 1.50) Extremely low
Chen et al. (2019) ADL 3 −2 f −1 c 0 −1 e −1 b 149 152 MD = 10.48 (8.33, 12.64) Extremely low
AR 5 −2 f 0 0 0 −1 b 234 237 OR = 0.31 (0.19, 0.49) Extremely low
TER 9 −2 f 0 0 0 −1 b 412 415 OR = 2.07 (1.39, 3.07) Extremely low
Zhang et al. (2019) ADL 16 −1 a 0 0 0 0 NR NR SMD = 0.38 (0.25, 0.49) Moderate
MMSE 23 −1 a −1 c 0 0 0 NR NR SMD = 0.66 (0.44, 0.89) Low
Du et al. (2017) ADL 5 −2 f 0 0 −1 e −1 b 160 156 MD = −3.60 (-4.53, -2.66) Extremely low
MMSE 7 −2 f −1 c 0 0 −1 b 238 238 MD = 2.69 (1.46, 3.92) Extremely low
ADAS-Cog 3 −2 f −1 c 0 −1 e −1 b 88 83 MD = −4.54 (-5.64, -3.43) Extremely low
Ma et al. (2018) CDR 1 −1 a −1 c 0 −1 e −1 b 31 15 MD = −0.40 (-0.69, -0.11) Extremely low
ADL NA −1 a −1 c 0 −1 e −1 b NR NR MD = 0.94 (-1.54, 3.43) Extremely low
MMSE 5 −1 a 0 0 −1 e −1 b 114 109 MD = 0.69 (-0.17, 1.56) Extremely low
AR NA −1 a −1 c 0 −1 e −1 b NR NR NR NR Not evaluated
Xu and Xie, (2015) MMSE 4 −2 f −1 c 0 −1 e −1 b 125 126 MD = 2.87 (0.64, 5.10) Extremely low
TER 7 −2 f 0 0 0 −1 b 278 246 OR = 1.25 (1.14, 1.38) Extremely low
Zeng et al. (2015) MMSE 15 −1 a −1 c 0 −1 e −1 b 717 575 MD = 0.79 (-0.11, 1.69) Extremely low
AR −1 a −1 c 0 −1 e −1 b NR NR NR NR Not evaluated
TER 15 −1 a 0 0 0 −1 b 717 575 OR = 1.09 (0.82, 1.46) Low

NR, Not reported; TCM-SSS, Traditional Chinese Medicine syndrome score scale; ADL, Activies Daily Living; MMSE, Mini-mental State Examination; ADAS-Cog, Alzheimer’s Disease Assessment Scale-Cognitive section; AR, adverse reactions; TER, total effective rate; SMD, standard mean difference; MD, mean difference; OR, odds ratio.

a

The included studies have a high risk of bias in terms of randomization, blinding, allocation concealment, completeness of result data, or selective reporting.

b

Asymmetric funnel plot or less than nine studies.

c

75% ≤ I 2 ≤ 100%.

d

50% ≤ I 2 < 75%.

e

small sample studies accounted for the majority.

f

The included studies have two or more high risks of bias in terms of randomization, blinding, allocation concealment, completeness of result data, or selective reporting.