Editor—Abbasi quotes Richard Smith saying, “Journals are in the debate business, not the truth business.”1 Evidence based medicine (EBM), on the other hand, is supposed to glean facts and is clearly in the truth business. It is disturbing when these truths become debatable. We would like to use your columns to debate the truth business.
We recently completed a small randomised controlled trial of surfactant in premature newborn infants of 27-30 weeks' gestation in India. We found that surfactant did not improve survival. This study apparently contradicted a Cochrane review that says that the meta-analysis supports a decrease in the risk of neonatal mortality (typical relative risk 0.60, 95% confidence interval 0.44 to 0.83; typical risk difference -0.07, -0.12 to -0.03).2
On examination of the details of the Cochrane report we found that the meta-analysis did not actually find better survival in those receiving surfactant (relative risk 0.70, 0.47 to 1.06, for survival up to discharge). Further meta-analysis found better survival when looking at death within 30 days of birth. This was sufficient “evidence” for the reviewer's conclusion, that there “was a decreased risk of mortality.”
How many parents would consider survival for 30 days a crucial end point if their baby did not survive to go home? “What is truth?” Pontius Pilate once asked.3 It seems that evidence based medicine reports are not averse to a bit of spin.
Competing interests: None declared.
References
- 1.Abbasi K. Lifeline. Lancet 2004;364: 675.15325832 [Google Scholar]
- 2.Soll RF. Prophylactic natural surfactant extract for preventing morbidity and mortality in preterm infants (Cochrane Review). In: Cochrane Library. Issue 3. Chichester: John Wiley, 2004. ( [DOI] [PMC free article] [PubMed]; Cochrane Database Syst Rev 2000(2): CD000511.) [DOI] [PMC free article] [PubMed]
- 3.Holy Bible. John 18, 38. (New international version.)