Intended for healthcare professionals

CCBYNC Open access
Research

Association of BCG, DTP, and measles containing vaccines with childhood mortality: systematic review

BMJ 2016; 355 doi: https://doi.org/10.1136/bmj.i5170 (Published 13 October 2016) Cite this as: BMJ 2016;355:i5170

Re: Association of BCG, DTP, and measles containing vaccines with childhood mortality: systematic review

Dr Shann correctly points out that, in our review of epidemiological studies on the association between receipt of DTP vaccine and all-cause mortality [1], our sensitivity analysis removing the Papua New Guinea study presented the result of a fixed-effect meta-analysis giving an estimated RR of 1.36 (95% CI 1.09 to 1.68). We had intended to present results of a random-effects meta-analysis, which gives an RR of 1.53 (95% CI 1.02 to 2.30), compared with the estimated RR of 1.38 (0.92 to 2.08) in the main analysis including all studies. We thank Dr Shann for pointing out this error, and have requested that a correction be published. We note that the substantive conclusions do not differ greatly between the sensitivity analyses based on fixed- and random-effects analyses.

We agree with various rapid responses that there are major concerns about some of the studies, including the Papua New Guinea study, as discussed by Dr Shann and by Aaby et al [2]. As we explained in detail in our responses to the peer reviewers, including Dr Shann, it is important to take a consistent approach to evaluating the risk of bias in all studies, based on information presented by the study authors rather than from commentators (including peer reviewers). The results of our assessments were mostly unaffected by our returning to them carefully in order to respond to the referees’ comments and suggestions. It may well be that there are serious biases in studies that we judged to be at ‘high risk’ rather than ‘very high risk’ of bias based on the information available; our assessments were of risk of bias rather than bias itself. We also agree that direction of bias is critical and would like to have been able to make useful statements about these. As we explain in the paper and in our responses to peer reviewers, we did not have sufficient confidence to make such statements about the effects of all the potential biases at play.

[1] Higgins JPT, Soares-Weiser K, López-López JA, Kakourou A, Chaplin K, Christensen H, Martin NK, Sterne JAC, Reingold AL. Impact of BCG, DTP and measles-containing vaccines on childhood mortality: a systematic review. BMJ 2016; 355: i5170.
[2]. Aaby P, Ravn H, Benn CS. The WHO review of the possible non-specific effects of diphtheria-tetanus-pertussis vaccine. Pediatr Infect Dis J 2016; 35: 1247–57.

Competing interests: No competing interests

27 February 2017
Julian PT Higgins
Professor
Karla Soares-Weiser, Jonathan AC Sterne, Arthur L Reingold
University of Bristol, Bristol, UK