September 15, 2014
Patients are given an idea of their diagnosis, treatment and its effectiveness using a wide range of resources like Medical Literature, personal experience and research and intuition. The classic example for this would be the classification of cancer patients into different stages and then giving them choices of treatments which gives them X more months.
But the chairman of the department of quantitative health sciences at Cleveland Clinic’s Lerner Research Institute, Michael Kattan says that this approach doesn’t always take the personal idiosyncrasies of each patient into consideration which are unique to each individual.,. This made the doctors turn to increasingly sophisticated calculators that can combine all these unique characteristics, compare them with the vast databases of similar cases and studies, and use them to predict probable outcomes depending on the treatment a patient chooses.
Patients too are increasingly demanding objective and accurate information when faced with making tough decisions, for example choosing a treatment that improves quality of life against the one that improves longevity.
Dr.Kattan, after facing a similar situation for his lymph node cancer began to shift his attention to health-care prediction, developing clinical risk calculators called “nomograms” that use statistical modeling to tailor predictions to individual patients, using large databases from medical studies. He joined Memorial Sloan Kettering Cancer Center in New York in 1998, developing more statistical models and starting a website, continuing to add new calculators. In 2004, he joined Cleveland Clinic, with the lure of an opportunity to expand big-data modeling beyond cancer to include coronary artery disease, Type 2 diabetes and total joint replacements.
“Risk calculators are vital for improving medical decision-making,” Dr. Kattan says. “Tools like this represent another step toward personalized medicine that will ultimately improve efficiency, outcomes and patient care.”
The intended audience for most nomograms are physicians, who can access them online for free. While some have been adapted for patient use, such as a prostate-cancer-risk assessment available through the ClevelandClinic.org website it doesn’t mean that these calculators can replace doctors.
The most important thing to remember, is the fact that the calculators can only give probabilistic values that are more accurate than the traditional doctors’ predictions but this doesn’t mean that calculators can reliably predict the outcomes for certain.
Dr. Kattan while acknowledging the limitations counters that nomograms are hard to beat for counselling patients on their prognosis.
Physicians who are using the calculators say that these calculators help in the often tough decisions patients must make as the models work by determining points for each of a patient’s individual characteristics, and then map the sum of the points to a predicted probability of outcome.
A semiretired nuclear physicist, Eric Weissman, , had a prostate cancer surgery in 2002, but a decade later, his PSA started to rise again. He consulted with a physician about undergoing radiation therapy.
“As a scientist, I value highly decisions that are made on quantitative, precise information rather than just the art of a medical approach,” Mr. Weissman says. While a doctor’s experience is important, he says, it is “comforting” to draw on statistics from a large pool of patients. He decided to have the radiation and remains cancer free 2½ years later.