Home            Past Issues            About IP            About IAP           Author Information            Subscription            Advertisement              Search  

   
correspondence

Indian Pediatr 2009;46: 81

Reply

G Girish and AK Deorari,

Department of Pediatrics,
AIIMS, New Delhi, India.

 

We agree with statistical interpretation of the results made by Dr. Dutta. However, conclusions and interpretation made in our study must be viewed in light of some important facts. In absence of concrete data on duration of phototherapy in Rh hemolytic disease, calculation of sample size was based on our pilot data (unpublished). We enrolled the pre-calculated number of subjects, but because of wider dispersion of phototherapy duration, the primary outcome of our study, we were unable to reject the null hypothesis that 1 g/kg of IVIg is not better than 0.5 g/kg in reducing duration of phototherapy. Post-hoc analysis showed that the study was underpowered to detect a difference of 24 h in the duration of phototherapy- the intended difference. But the study had 80% power to detect a difference of 36 hr in the duration of phototherapy. The trend towards decreased duration of phototherapy observed in 1 g/kg IVIg group should be interpreted with caution in light of the opposite trend of longer hospital stay in the same group and the illness severity of babies. Moreover, we were unable to detect significant difference in other outcomes like number of exchange transfusions, duration of hospital stay, number of packed red blood cell transfusions and peak serum total bilirubin. Our study was not powered to detect change in these outcomes, but detecting statistically meaningful difference in an important outcome like need of exchange transfusion will need huge sample size.

Furthermore, even a small difference will achieve statistical significance if sample size is big enough. Statistical significance testing does not reflect the magnitude of the effect, and the term "statistically significant difference" does not denote that the difference between a test and control group was clinically meaningful with regard to a desired outcome. We agree that reporting of no dose-effect relationship between IVIg and duration of phototherapy may not be correct statistically, but we based our conclusions on utility of clinical benefit than what our study was powered enough to detect.

 

Copyright© 1999 by the Indian Pediatrics (Disclaimer)