LAUSR

We read the letter by Stritzke and Shah [1] and appreciate their comments. It was stated that “... the re-analysis display some serious misinterpretation/recalculation errors that have made the results very inaccurate...”. We must emphasize that we extracted data accurately according to our rigorous and pre-defined exposure definition; however, our exposure definition differed from that used by Stritzke et al. [2], ultimately leading to a different interpretation of the same data set. The difference lies mainly in the definition of exposure in controls (cell B). As outlined in our paper, we created 2 × 2 tables as follows: – Cell A = number of infants with necrotising enterocolitis (NEC) who received a recent packed red blood cell (PRBC) transfusion (within 48 h) – Cell B = number of infants without NEC who received at least one PRBC transfusion during the study period – Cell C = number of infants with NEC who were either never exposed or exposed >48 h prior to the diagnosis of NEC – Cell D = number of infants without NEC who were never exposed to a PRBC transfusion In adherence to these methods, data were, in fact, correctly extracted from Stritzke et al. [2] as follows: – Cell A = 144 infants with NEC who received a recent PRBC transfusion (Results; P2; line 7) – Cell B = 1091 infants without NEC who received at least one PRBC transfusion (Results; P2; line 6) – Cell C = 783 infants with NEC who were either never exposed or exposed >48 h prior to the diagnosis of NEC [927 − 144 (Table 1)] – Cell D = 1690 infants without NEC who were never exposed to a PRBC transfusion [2781 controls without NEC (Abstract, Results, line 1) − 1091 infants without NEC who received at least one PRBC transfusion (Results; P2; line 6)].