Since the money is there, it’s reasonable to ask why states and districts aren’t already using it in this commonsense way. There are two reasons. First, federal funds that are intended to help districts reimagine their systems are instead often used to expand existing programs, typically without adequate consideration of whether those programs are serving schools well. The second reason is even more problematic and will take more political muscle to turn around. In many states, laws hamstring districts by mandating that pay scales be based on a teacher’s length of service or degrees earned—neither of which is strongly linked to student learning—instead of job performance. Where such policies stand in the way of reform, we have to replace them with new laws. Florida, Indiana, and Michigan have changed their laws to require that administrators factor a teacher’s job performance into his or her salary.
I’ve seen performance pay work. In Washington, D.C., when I was chancellor, we launched a system in which highly effective teachers can get paid up to $130,000. As a result, I saw committed educators stay in the classroom, where they are so badly needed, rather than move into more-lucrative fields.
Michelle Rhee, the former Washington, D.C., public schools chancellor, is founder and CEO of StudentsFirst, a national bipartisan education-reform group.
Big Idea in a Small Place
How biomarkers could transform medicine.
By Gail R. Wilensky
When National Journal asked me what I thought was the most promising innovation in health care, I first tried to think, probably like most people, of a new instrument or a handheld device that might help doctors get us well quicker or keep us in our homes longer as we age. But now, I submit something different: the increasing use of biomarkers, usually gene- or protein-based compounds, to indicate the presence or risk of disease or to fight it on a more targeted basis—which means more cost-effectively. These are big issues for an aging, overweight population.
Biomarkers are not new. Physicians already use hemoglobin A1c to measure glucose levels for diabetics and PSAs, or prostate-specific antigens, a protein that (somewhat unreliably) suggests a risk of developing prostate cancer. Other markers can be used to indicate whether a disease exists, how the disease will develop, or the probable effect of a treatment. They may also modernize drug development by helping to predict what kind of patients will benefit from new drug regimens, reducing the time and money it takes for drugs to be tested and approved.
The really promising biomarkers, however, are the ones that are still being developed—to treat cancer, heart disease, and potentially even neurological diseases such as Alzheimer’s.
We have probably heard the most about how biomarkers might help doctors detect cancer earlier and choose the best drug for a particular patient, accounting not only for the tumor type but also his or her own biomarkers. The presence or absence of estrogen receptors or HER-2/neu (human epidermal growth factor receptor 2) as a guide to the appropriate treatment of breast cancer received a lot of attention in recent years, but there are now biomarkers for colorectal, gastric, blood, lymphatic, lung, and skin cancers as well.
Researchers are also investigating ways that biomarkers can diagnose strokes, which would allow for earlier and less invasive treatment. Because heart disease (including strokes) and cancers are the most common causes of death in the United States, these advances have great potential.
Biomarkers offer a lot of promise, but potential barriers exist. The data are complex and frequently require new modeling to help us understand what they mean. But scientists are making progress using enormous data sets. Sequencing the human genome was a $3 billion project that took 10 years. New projects aim to cost $1,000 per genome. The IT requirements to process these data are also formidable. What’s more, different computer systems—between different hospitals, and sometimes even those within a single hospital—can’t talk to each other. Even the Defense Department and the Veterans Affairs Department, the two agencies with the most advanced electronic medical records, can’t yet share inpatient data (although they can now share other types of data). And, of course, researchers—and the companies that sponsor them—often have few incentives to share information.
Finally, for the country to learn how to treat better and to spend smarter, patients, practitioners, and insurers will have to adjust their views about appropriate treatment strategies: Expensive, but frequently less invasive, therapies should be encouraged (i.e., with low co-pays) for those who might experience significant clinical gains and should be discouraged (i.e., with high co-pays) for those with a low likelihood of clinical benefit. Getting agreement on the latter will be tough for the United States—although it’s commonplace for the rest of the world—even without good biomarkers at hand. Good, credible, objective data that support these treatment decisions will be critical. This is the function of the newly created Patient-Centered Outcomes Research Institute that was included in the Affordable Care Act. Increasing the financial support for this center would be especially helpful.
Gail R. Wilensky, Ph.D., is a senior fellow at Project HOPE; a former administrator of HCFA, now CMS; and a former chair of the Medicare Payment Advisory Commission.
This article appears in the July 14, 2012, edition of National Journal.