Plasma amino acids in preterm infants fed on human milk or formula.

'Clinical sepsis' was considered to be present if clinical and/or haematological signs were recorded. If, in addition to these, a positive blood culture was recorded, the sepsis was defined as 'bacteriologically confirmed'. Central catheters were used only sparingly in the Netherlands, and total parenteral nutrition was administered mostly by peripheral infusion. Blood cultures were seldom taken routinely, even in infants treated by total parenteral nutrition. Only clinical evidence of sepsis justifies such a procedure in these very small infants. Dr Puntis misread the part about causative organisms. Far from considering these benign, we stated that in infants treated by total parenteral nutrition the causative organisms were 'fairly benign' namely, compared with infants not treated by total parenteral nutrition; this point is illustrated even better by our original unpublished table, categorising the causative organisms by duration of total parenteral nutrition (see table).