Six RCTs (or randomised trials following cohort accrual) with 1,044 participants were included in the review.
The authors reported substantial inter-rater agreement in the validity assessment (weighted kappa 0.70).
People who received ABMT had better disease-free survival at 48 months than those who received chemotherapy or no further treatment (ratio of disease-free survival probabilities in 6 trials 1.24, 95% confidence interval, CI: 1.06, 1.44, P=0.006; heterogeneity P=0.38). The ratio of overall survival probabilities (from 5 trials) was 1.01 (95% CI: 0.89, 1.15, P=0.86; heterogeneity P=0.33). There was no statistically significant difference in disease-free survival (P=0.38) or overall survival (P=0.33) between the studies.
ABMT was associated with a statistically significantly greater risk of death during first remission (odds ratio from 6 studies 2.63, 95% CI: 1.6, 4.32, P<0.001).
Funnel plots suggested some degree of publication bias, in that the results of small negative trials might not have been published. The authors performed a modelled analysis to correct for this and found that publication bias was unlikely to have made a statistically significant difference to the findings.