Should thoracic paravertebral blocks be used to prevent chronic postsurgical pain after breast cancer surgery? a systematic analysis of evidence in light of IMMPACT recommendations
Hussain N., Shastri U., McCartney CJL., Gilron I., Fillingim RB., Clarke H., Katz J., Juni P., Laupacis A., Wijeysundera D., Abdallah FW.
The role of thoracic paravertebral block (PVB) in preventing chronic postsurgical pain (CPSP) after breast cancer surgery (BCS) has gained interest, but existing evidence is conflicting, and its methodological quality is unclear. This meta-analysis evaluates efficacy of PVB, compared with Control group, in preventing CPSP after BCS, in light of the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT) recommendations. Electronic databases were searched for randomized trials comparing PVB with Control group for CPSP prevention after BCS. Eligible trials were assessed for adherence to IMMPACT recommendations. The primary outcomes were CPSP at 3 and 6 months, whereas secondary outcomes were PVB-related complications. Data were pooled and analyzed using random-effects modelling. Trial sequential analysis was used to evaluate evidence conclusiveness. Data from 9 studies (604 patients) were analyzed. The median (range) of IMMPACT recommendations met in these trials was 9 (5, 15) of 21. Paravertebral block was not different from Control group in preventing CPSP at 3 months, but was protective at 6 months, with relative risk reduction (95% confidence interval) of 54% (0.24-0.88) (P 5 0.02). Meta-regression suggested that the relative risk of CPSP was lower when single-injection (R2 5 1.00, P < 0.001) and multilevel (R2 5 0.71, P 5 0.01) PVB were used. Trial sequential analysis revealed that 6-month analysis was underpowered by at least 312 patients. Evidence quality was moderate according to the GRADE system. Evidence suggests that multilevel single-injection PVB may be protective against CPSP at 6 months after BCS, but methodological limitations are present. Larger trials observing IMMPACT recommendations are needed to confirm this treatment effect and its magnitude.