Table 3 Subgroup analyses of PRP compared with HA for WOMAC total scores at 12 months
From: Platelet-rich plasma versus hyaluronic acid in the treatment of knee osteoarthritis: a meta-analysis
Subgroup | No. of trials | Standard mean difference (95% CI) | P value | I2 (%) | Test of interaction, P |
|---|---|---|---|---|---|
Total | 8 | − 1.47 (− 2.23, − 0.70) | 0.000 | 95.3 |  |
Number of PRP injections | |||||
  1 | 4 | − 1.15 (− 2.14, − 0.85) | 0.000 | 69.3 | 0.006 |
  ≥2 | 4 | − 1.68 (− 2.35, − 0.72) | 0.000 | 98.2 | |
PRP spinning approach | |||||
  Single | 3 | − 1.13 (− 2.22, − 0.85) | 0.000 | 0.0 | 0.028 |
  Double | 5 | − 1.62 (− 2.34, − 1.11) | 0.000 | 88.4 | |
PRP classification | |||||
  LPP | 3 | − 2.89 (− 6.55, − 0.68) | 0.000 | 82.4 | 0.002 |
  LRP | 5 | 0.86 (0.15, 1.56) | 0.000 | 83.5 | |
Fresh or frozen PRP | |||||
  Fresh | 5 | 0.78 (0.54,1.01) | 0.000 | 57.8 | 0.105 |
  Frozen | 3 | 0.50 (0.33,0.67) | 0.000 | 88.1 | |
Activator used | |||||
  Yes | 4 | − 2.78 (− 4.32, − 1.15) |  | 82.1 | 0.031 |
  No | 4 | 0.58 (0.12, 1.58) |  | 93.4 | |
Risk of bias | |||||
  Low | 6 | − 1.23 (− 2.08, − 0.35) | 0.002 | 85.9 | 0.254 |
  Unclear/high | 2 | − 1.33 (− 2.24, − 0.46) | 0.004 | 89.4 | |
Molecular weight | |||||
  High | 4 | − 1.90 (− 4.53, 0.73) | 0.157 | 97.9 | 0.003 |
  Low | 4 | − 1.31 (− 1.97, − 0.64) | 0.000 | 92.2 | |
Structure | |||||
  Cross-linked | 4 | − 2.06 (− 3.20, − 0.92) | 0.352 | 96.2 | 0.000 |
  Not cross-linked or unclear | 4 | − 0.57 (− 1.78, 0.63) | 0.000 | 95.0 | |