A total of six RCTs were included in this review (total n=561). Sample sizes ranged from 40 to 198 patients. Overall quality was considered to be fair: two trials reported concealment of allocation; no trials reported using blinded outcome assessors; losses to follow-up were generally reported in adequate detail; and four trials used intention-to-treat analyses. Funnel plots did not indicate the presence of publication bias, but the number of trials was low.
Seroma formation (six RCTs) was significantly less likely in patients who received volume-controlled drainage compared with short-term or no drainage (RR 0.44, 95% CI 0.24 to 0.80). Significant heterogeneity was present.
No significant differences in rates of wound infection (six RCTs) were found in comparison of volume-controlled and short-term/no drainage. No statistical heterogeneity was noted for this analysis.
Patients who received volume-controlled drainage stayed in hospital significantly longer than those randomised to short-term/no drainage (WMD 1.50, 95% CI 1.23 to 1.76; five RCTs). Significant heterogeneity was present.
It was not possible to pool the data on frequency of seroma aspirations, amount of volume drained or numbers of drain reinsertions due to variation in outcome reporting. Overall the sensitivity analyses did not significantly alter the main results, but the authors reported that volume-controlled drainage only reduced seroma formation in patients who underwent mastectomy and not breast conserving surgery. Statistical heterogeneity was reported for both of these analyses.