TABLE 5.
GRADE evidence profiles.
| Outcome (BMD) | No. of participants (No. Of studies) | Certainty assessment | Absolute effect (95% CI) | Certainty | |||||
|---|---|---|---|---|---|---|---|---|---|
| Study design | Risk of bias a | Inconsistency | Indirectness | Imprecision | Other considerations | ||||
| HM + DMARDs versus DMARDs alone in the lumbar spine | 1,042 (13) | High | Not serious | Not serious | Not serious | Not serious | None | MD 0.04 higher (0.03–0.05 higher) | ⊕⊕⊕⊕ High |
| HM + BPs versus BPs alone in the lumbar spine | 110 (2) | High | Not serious | Not serious | Not serious | Not serious | None | MD 0.13 higher (0.12–0.15 higher) | ⊕⊕⊕⊕ High |
| HM + DMARDs versus DMARDs in the femoral neck | 876 (11) | High | Not serious | Not serious | Not serious | Not serious | Publication bias strongly suspected b | MD 0.03 higher (0.02–0.05 higher) | ⊕⊕⊕○ Moderate |
| HM + BPs versus BPs alone in the femoral neck | 110 (2) | High | Not serious | Very serious c | Not serious | Serious d | None | MD 0.17 higher (0.02–0.30 higher) | ⊕○○○ Very low |
| XLGB + DMARDs versus DMARDs alone | 196 (3) | High | Not serious | Not serious | Not serious | Not serious | Publication bias strongly suspected b | MD 0.04 higher (0.03–0.04 higher) | ⊕⊕⊕○ Moderate |
| HBK + DMARDs versus DMARDs alone | 264 (3) | High | Not serious | Not serious | Not serious | Not serious | None | MD 0.02 higher (0.00–0.03 higher) | ⊕⊕⊕⊕ High |
BMD: bone mineral density, BP: bisphosphonate, CI: confidence interval, DMARDs: disease-modifying anti-rheumatic drugs, GRADE: grading of recommendations assessment, development, and evaluation, HBK: Hanbikang-tang, HM: herbal medicine, MD: mean difference, XLGB: Xianlinggubao-capsule.
a
All studies were evaluated as “Not serious” because BMD, is an objective indicator; therefore, it is difficult to say that bias would have affected the results.
b
Funnel plot is asymmetric.
c
High heterogeneity.
d
95% CI, wideness.