Health knowledge made personal
Join this community!
› Share page:
Search posts:

Is Evidence Based medicine morphing into Algorithm Based medicine?

Posted Feb 03 2010 8:20pm

We all want health care to be based, as much as possible, on a scientific basis. When we initiate a treatment or make a diagnosis, we need evidence that the diagnostic methods are accurate and that the treatment is efficacious. But when “evidence based” medicine goes through the political process it can turn into “algorithm based” medicine. Let me explain.

Algorithm based medicine uses data from large groups and applies that to the individual. The NICE commission of Great Britain does that with their decisions about how much money they are willing to spend to prolong a patient’s life by six months. The recent new mammogram recommendations put forth by the U.S. Preventive Services Task Force do that for American women. Recommendations of this sort portend thinking that may soon be put into law. This approach places economics and so-called efficiency at a higher rank than autonomy and beneficence.

Going even farther than the U.S. PSTF, both the Pelosi and Reid versions of the Health Care Reform act currently before Congress mandate such algorithm decisions. For example, if you are 75 years old, have type 2 diabetes, had prostate cancer surgery six years ago, and have mild hypertension, you may have enough “points” against you that the algorithm will deny you hip replacement surgery on grounds that it is a bad economic investment. Your neighbor, who does not have type 2 diabetes but in all other ways matches you, might qualify.

This form of algorithm-based medicine is presented to the public as a way of containing costs and increasing efficiency. However, a truly intellectually honest way to present the matter is the following: “The algorithm dictates that it is preferable for a certain percentage of women in their 40’s die of breast cancer in order to make the cost of mammograms for women in their 50’s more affordable.” If presented this way the public would have a clearer picture of choices.

Here is another real life example. My fiancé, Beau Briese, a first year Resident in Emergency Medicine at Stanford, is currently working on a series of research projects that involve the PESI score, or the Pulmonary Embolism Severity Index score – an algorithmic method for determining the risk of mortality for patients who have image proven pulmonary embolisms. The score categorizes patients into five classes of risk for in-patient and 30 day mortality based upon 11 characteristics including gender, age, heart rate, and whether or not the patient has altered mental status. The goal of this score is to help physicians determine who needs to be admitted to the hospital and who can be discharged from the emergency room on the day of diagnosis. While I am glad that Beau is working on an algorithm that will help physicians determine an otherwise potentially subjective decision for hospitalization, it is also a method that if implemented strictly as rule rather than guideline may be detrimental to healthcare in general. For example, the PESI score class II has a 0.5% risk of mortality. That means that 1 out of every 200 patients who are sent home with this score will unnecessarily die at home. On the other hand, 199 out of those 200 people will not risk being needlessly hospitalized, which will save the patient time and worry and the hospital money and resources.

Putting distributive justice above autonomy and individual responsibility seems superficially to improve health care efficiency. However its long-term dangers are grave. Should the emphasis on health care be efficiency and equality? Imagine the following scenario.

Identical twin brothers graduate from the same high school and college in the same field. They live in the same neighborhood. One spends his money on fancy cars, fine wines, expensive restaurants, and trips to Las Vegas. He decides he is healthy and does not buy health insurance, using his resources for fun. His brother, who makes the exact same amount of money, buys a “Cadillac” health insurance policy, puts money aside to tide himself over in the event of disability, and drives modest cars. When they turn 40 they both develop the identical cancer. Should their treatment be the same? Should the prudent brother’s foresight gain him nothing? Should the brother who elected not to buy insurance until after he was diagnosed be allowed to have the same new health insurance as the one who bought it before the diagnosis?

The following are links to a description of the tragedy of the commons and defense of the proposed mammography guidelines and feel free to email me if you would like a pdf of the JEM PESI Score article:

Post a comment
Write a comment: