In health care as in other fields, we can be misled into believing that a particular treatment strategy is right, when the evidence for that notion was either shaky to begin with or better evidence suggests it was erroneous. A recent issue of JAMA features two studies demonstrating that conventional medical wisdom should constantly be tested and questioned. An accompanying editorial reminds us of the dangers of statistics, which have become quite complex and often blind us from the characteristics of the raw data; characteristics which themselves are quite informative. (JAMA Abstracts)
The first study related to lung transplant outcomes in relation to various factors, including volumes of procedures performed at the transplant facility. It has been an article of faith for many medical procedures that centers with higher volumes have better outcomes. The study, however, found that volume of procedures accounts for only about 15% of the difference in outcomes across patients and centers. Assessing over 15,000 lung transplants across 61 centers on a primary outcome of survival time, the researchers concluded that choice of center was associated with significant variability in survival, but that volume alone only explained a small part of the variability, around as much as the recipient’s functional status. Some low and medium volume centers had very good outcomes. So the facility used to obtain a transplant may make a difference, but it probably isn’t largely due to volume.
The second study relates to blood pressure control among patients with diabetes and coronary artery disease. Based on what they believed to be reliable trial results, groups such as the American Diabetes Association and the American Heart Association have recommended very aggressive control of blood pressure in this population, seeking to keep systolic pressure under 130 mm and diastolic under 80, and suggesting that even further lowering could be beneficial without risks. The authors conducted a retrospective review of data from large blood pressure trials which included substantial populations of diabetes patients and found that such aggressive control of blood pressure was not associated with better outcomes and in fact likely heightened the overall risk for death, potentially due to adverse effects from the drugs used to obtain such tight control. The authors suggested that having such a low target for blood pressure was not appropriate or necessary.
One of the biggest cautions suggested by these studies is the danger of relying too heavily on guidelines or evidence-based medicine. What appears to be a solid basis for medical care at one point in time unfortunately often turns out to be wrong upon the accumulation of more research data.