Facility vs. community perinatal mortality David Flood, University of Minnesota 7 September 2015 I read with great interest this study by Garces et al. of the Global Network Maternal and Newborn Health Registry, which offers very important statistics on child and maternal health in Chimaltenango, Guatemala. In their conclusion, the authors write that "these data suggest that consideration should be given to a policy of phasing out home births with TBAs." I am less convinced that this conclusion can be drawn from the results. Facility mortality rates did show a statistically significant decline across each of the three metrics (neonatal, stillbirth, perinatal), whereas community rates did not decline. However, no explicit comparison is made between facility vs. community rates for 2013, and, in fact, community rates are slightly lower for two of the three metrics (though without statistical testing shown). Both statements can be true because facility mortality rates were much higher in 2010. There are likely to be many confounders; for example, facility births may be more complicated than community births. And in future years, as the situation in Chimaltenango continues to evolve, new data may indeed prove the authors' correct. My overaching concern is that calling for the end of home births in Guatemala is an extremely delicate issue, one which should be invoked cautiously and judiciously. Competing interests I declare no competing interests.