Reflux and AutoimmunityResponse: Common Links Among Patients With Pulmonary Fibrosis?

We thank Drs. Netzer and Omron for their interest in our article. We agree that observational data such as we used are tricky and cannot consider all potential confounders. Confounding by indication is always the weakness of this kind of work. However, we tried to avoid this, and could not identify plausible causes for this kind of confounding. Dr. Netzer points out that we did not look at markers of cardiovascular disease, such as cardiac drugs, and she is right. However, in another study we have looked at cardiovascular morbidity as a function of respiratory drugs and found that inhaled steroids tended to be protective, but this effect was independent of interaction with cardiac drugs. Dr. Omron is not convinced that users of inhaled steroids were as sick as people who did not receive these drugs. We would argue that the evidence favors them being at least as sick; we regard frequent physician visits and multiple drugs as evidence of perceived severity, not the reverse. Further, we studied dispensed drugs only; we had no data regarding drugs that were prescribed but not dispensed. The fact that the steroid effect was most notable soon after dispensation we take as evidence favoring drug use.