Ruby, I agree with you on your issues with the methodology that was used for this essay on AJWRB. The study that was used showed a range from actually 0 increase in morbid and mortality issues, for certain classifications of procedures, all the way up to 1.5% increase for cardiovascular procedures.
So yes a conservative number of 1% for cardiovascular procedures is correct. But then you say that because the Blood Bank Association says that 4 million people in the US need a blood
And yes, the math does show that for every 1000 people in the US 150 of them need a blood transfusion a year.
Though you then try and correlate a 1% increase in mortality for all procedures that blood could possibly be used. That is not a fair assessment of the original author's intention. While yes for certain procedures the mortality would be higher without a blood transfusion that is not all of the procedures that were even listed in the original study. The author even states this in his conclusion:
"The impact on the overall morbidity and mortality is not known; such a study would take decades of observation with large numbers of randomized patients and is unlikely to ever be done. Patients now often question the risk."