Paperonium

Child health · Lancet Child Adolesc Health · 2018

Summary: correct fluids clearly cut child deaths in every group. Its cited figures don’t show that.

Breathing signs1.23×, could be no differenceCirculation signs0.95×, could be no differencePale children1.70×, could be no differenceDehydration signs only1.50×, could be no differenceBrain or nerve signs0.86×, could be no difference0.68 to 2.24×0.531.730.953.020.792.880.511.480.5×0.7×1×1.5×2×3×no differencefewer deaths
Breathing signs1.23×, could beno differenceCirculation signs0.95×, could beno differencePale children1.70×, could beno differenceDehydrationsigns only1.50×, could beno differenceBrain ornerve signs0.86×, could beno difference2.24×0.680.531.730.953.020.792.880.511.480.5×1×2×5×no differencefewer deaths

Times the odds of dying within 2 days, correct fluids versus not, by group

The paper says

For children in hospital, the summary says fluids prescribed as guidelines recommend “significantly” (clearly) cut deaths within 2 days in all five groups it names, like pale children. Its evidence: 1.23, 0.95, 1.70, 1.50 and 0.86 times the odds of dying (under 1 means lower).

Its own data show

The summary prints a range of likely values beside each figure. All five include 1 (no difference), so none shows a clear drop; pale children’s runs from 0.95 to 3.02. Its Table 2 points to fewer deaths with correct fluids in every group, but never reports that figure for each warning sign (such as paleness).

Each of the five divides one group’s figure by another’s. With correct rather than wrong fluids, pale children had 0.50 times the odds of dying (worked out from the table), other children 0.29. 1.70 is about 0.50 ÷ 0.29: it compares the groups; it does not mean more deaths. Summary readers see five unclear figures, three seemingly showing harm.

“The summary” means the paper’s abstract: the short summary printed at its top, which is the part most readers see.

Read the paper →Full technical write-up →

Risk factors for mortality and effect of correct fluid prescription in children with diarrhoea and dehydration without severe acute malnutrition admitted to Kenyan hospitals: an observational, association study · Lancet Child Adolesc Health (2018) · doi:10.1016/S2352-4642(18)30130-5