05/08/2010 05:12 am ET | Updated Nov 17, 2011

Real Health Care Reform: Limit Excessive Care and Remove Nonclinical Factors From Medical Decisions

The decisions doctors make - what tests to order, what procedures to recommend, and what medications to prescribe - are often influenced by factors other than good medical care. Patient expectations and defensive medicine sway these decisions, as do the economic interests of hospitals, doctors, medical equipment manufacturers, and pharmaceutical companies. If we were able somehow to control medical decisions so that these nonclinical factors played only a minimal role, we could both improve care and cut costs. It may sound simple, but limiting excessive care and basing medical treatment on science tempered by compassion has proven very difficult to accomplish. So difficult, in fact, that Congress in its ongoing discussions on health care have chosen largely to confine the debate on how to finance our overpriced, underperforming medical system, and have ignored the more politically difficult question of how to improve it.

Recently, several studies have been published which can serve as models to show that improving quality and controlling costs can be simultaneously accomplished. In December, 2009 two articles were published in the Archives of Internal Medicine discussing the increasing use of CT scans and the possible harmful effects of the radiation exposure that result from that test. CT scans are a computerized series of x-rays that provide detailed views of different parts of the body. These scans have been very helpful at diagnosing and monitoring diseases and represent such a significant medical advance that the inventors of the CT scan were awarded the Nobel Prize for medicine in 1979. Over the past three decades, their use has been increasing yearly. In 1980 about 3 million CT scans were done in the US but by 2007, 72 million CT scans were done.

Most CT scans are performed for valid reasons but millions are carried out for only marginal indications. These additional scans are not just costly, but they present potential medical problems. Each CT scan, x-ray, mammogram, PET scan, and radionucleotide cardiac stress test exposes a person to radiation. Radiation from diagnostic tests can cause cancer with the risk of developing cancer dependent on the cumulative lifetime radiation exposure. Each and every additional test increases the possibility of a fatal disease. Investigators from the National Cancer Institute estimate that CT scans done in 2007 alone would cause 29,000 malignancies and 15,000 cancer deaths! If there were some way that we could decrease the number of CT scans (and all other diagnostic radiation exposure) without compromising quality we could both improve overall medical results and also cut costs.

In the January 2010 edition of Connecticut Medicine the use of CT scans in the emergency room of Norwalk Hospital to diagnose a pulmonary embolus (PE), or a clot in the lung, was reviewed. PEs are potentially fatal and are responsible for 50,000 deaths annually in the US. It is a diagnosis physicians do not want to miss for when found, effective, life-saving treatment is available. The experience in the ER in Norwalk Hospital in their approach to diagnosing PEs likely represents the pattern in ERs across the whole country. Although there are a variety of ways to diagnose pulmonary emboli, from 2000 to 2005, CT scans became essentially the sole test employed in the Norwalk emergency room entirely replacing all other tests. That short time span saw a 14-fold increase in CT use in the evaluation of patients for PE. Of the scans that were done, ninety three percent were normal and only seven percent were positive for the clots. For the overwhelming majority of patients with normal CTs, these extra scans were not only unnecessary, expensive, and avoidable but, because of the added radiation, presented a real risk for the future development of cancer. Again, this type of suboptimal and expensive medical care is not unique to a single hospital but represents nationwide trends.

There are well-documented ways, however, to avoid the excess CT scans and still not miss the diagnosis of a PE. Predictive models have been published in the medical literature, using clinical information including the patient's medical history and physical examination, which can determine a patient's likelihood of having a pulmonary embolus with a high degree of certainty. Those ER patients with a low likelihood could have a simple, inexpensive, risk-free blood test that would exclude a pulmonary embolus. Those with a greater likelihood could have a V/Q scan, a less expensive test than a CT that also exposes the patient to less radiation. CT scans would be reserved to the occasional patient for whom the V/Q scan was not diagnostic. This approach could both save money and improve care.

So why isn't this tactic already in place? Because the forces that drive American medicine demand otherwise! Doctors order the CT as added protection against malpractice suits. Radiologists and hospitals encourage their use because of the higher reimbursement rates third parties pay for the CT scans. And patients like the more advanced, more involved test. It is time that the less expensive approach with fewer dangerous side effects becomes the standard way to diagnose PEs. Pressure needs to be exerted to limit CT scans unless really indicated and doctors who follow well accepted predictive models should be protected against malpractice suits.

In other areas of health care where such similar models are already available but are similarly ignored, cost-effective medicine should become the accepted way to determine practice policies. Where not available, cost-effective models need to be studied and created. Adopting accountability in health care is an important means to control America's rising health costs while improving the quality delivered.