Table 3.
Evidence quality rated using the GRADE approach.
| Outcomes | No. of studies | Limitations | Inconsistency | Indirectness | Imprecision | Publication bias | Evidence quality | |
|---|---|---|---|---|---|---|---|---|
| DAS28a | 5 | Seriousb,c | Seriousd | Not serious | Not serious | Not found | ⊕⊕°° | Low |
| VASa | 5 | Seriousb | Seriousd | Not serious | Seriouse | Not found | ⊕⊕°° | Low |
| Serum vitamin D level | 3 | Seriousb,c | Not serious | Not serious | Not serious | Not found | ⊕⊕⊕° | Moderate |
| TJCa | 2 | Seriousb | Not serious | Not serious | Not serious | Not assessedf | ⊕⊕⊕° | Moderate |
| SJCa | 2 | Seriousb | Not serious | Not serious | Seriouse | Not assessedf | ⊕°°° | Very low |
| ESRa | 4 | Seriousb | Not serious | Not serious | Not serious | Not found | ⊕⊕⊕° | Moderate |
| CRPa | 4 | Seriousc | Not serious | Not serious | Seriouse | Not found | ⊕⊕°° | Low |
| PTHa | 2 | Seriousc | Seriousd | Not serious | Seriouse | Not found | ⊕⊕°° | Low |
a
DAS28, Disease Activity Score 28; VAS, Patient global pain score; TJC, Tender joint count; SJC, Swollen joint count; ESR, Erythrocyte sedimentation rate; CRP, C-reactive protein; PTH, Parathyroid hormone.
b
Study had reporting bias.
c
Study had attrition bias.
d
Significant heterogeneity was observed in this meta-analysis.
e
Wide confidence intervals, including values in favor of the experimental group and values in favor of the control group.
f
Not assessed because a limited number of studies were included in the meta-analyses on TJC and SJC.