Table 2. | Quantitative analysis of publication bias evaluation |
| Original meta-analysis | Trim and fill analysis | ||||
|---|---|---|---|---|---|
| Outcome | Mean difference | P-value | Mean difference | Studies trimmed/ total studies |
Egger regression P-level |
| 95% CI | 95% CI | ||||
| PD reduction (total) | 0.42 (0.27; 0.58) | < 0.00001 | 0.42 (0.27; 0.58) | 0/6 | 0.5 |
| PD reduction (moderate) | 1.07 (0.61; 2.8) | 0.002 | 1.07 (0.61; 2.8) | 0/4 | 0.1 |
| PD reduction (sever) | 1.44 (0.55; 2.33) | 0.002 | 1.44 (0.55; 2.33) | 0/4 | 0.1 |
| CAL gain (total) | 1.04 (0.2; 1.88) | 0.01 | 1.04 (0.2; 1.88) | 0/6 | 0.07 |
| CAL gain (moderate) | 1.33 (0.33; 2.34) | 0.009 | 1.33 (0.33; 2.34) | 0/4 | 0.7 |
| CAL gain (sever) | 1.47 (0.26; 2.68) | 0.02 | 1.47 (0.29; 2.68) | 0/4 | 0.01 |
| CAL gain (Sgolastra et al. [27]) | 0.42 (0.23; 0.61) | < 0.0001 | 0.37 (0.19; 0.54) | 1/6 | 0.72 |
PD = probing depth; CAL = clinical attachment level; CI = confidence interval. |
.png)