Why Evidence-Based Medicine and Government Guidelines are Impractical

     There have been many articles written recently about how the application of the results of "evidence-based" medicine to patients will both save money and make them healthier. We are also bombarded with many guidelines from the government as well as from various national medical groups as to how to diagnose and/or treat various diseases. I will not discuss the obvious problem that arises when different medical groups produce opposing guidelines (viz. the AMA, American Urological Assn., American College of Physicians, National Public Health Service, American Cancer Society) as to the proper timing of or the utility of PSA testing, or mammograms, or colonoscopies, or-------.

     One problem that the public does not appreciate is the impact these "suggestions" and "guidelines" have on the work time of physicians. Time is our least fungible resource. But every time you deviate (for good reason) from a recommended HMO or government guideline, you end up having to explain yourself in writing, either  in the chart, or in response to a letter from an HMO or Medicare, and this takes up time. Also, sometimes it is necessary to deviate from the guidelines to shield  you from a malpractice suit or to reduce your chances of losing one. The fact that a test you did not do was not indicated in the guidelines will not help you if a malpractice lawyer can convince a jury that application of the possible results of the test might have prevented permanent harm to your patient. And please remember that a doctor much prefers to avoid being sued than to be "right" and win the suit, because defending a suit takes time, energy, and time away from the office.

     When we look at evidence-based results, the first question a doctor asks is if his/her current patient fits the characteristics of the study group as to age, sex, race, economic standard of living and geography. For instance a recent article published in the New England Journal of Medicine on the use of cytisine for smoking cessation did have an equal number of men and women, and married and unmarried. However, all the participants were white, lived in Europe, were between 40 and 60 years of age, and smoked between 10 and 30 cigarettes a day, and your patient is a 25  year old Mexican immigrant who smokes 40 cigarettes a day, and wants to stop smoking. You don't even know the proper dosing regimen because you don't have any data on how rapidly or slowly Mexicans clear cytisine from their blood compared to white Europeans.

     A recent study in Lancet showed that treatment with Augmentin was "not inferior" to surgery for uncomplicated acute appendicitis. The study group was equally divided between men and women and covered ages 18 to 68.  But all were white, all lived in Europe, no mention is made of marital status or use of cigarettes, or even the other medicines they were taking, no one was obese, and no mention is made of prior abdominal surgery.

     Let us say you have a new patient with reflux esophagitis. There are very clear guidelines about treatment and the sequence of adding more drugs for control and which tests and procedures to do. (Although there is no mention of urecholine, probably because it has been generic for so long that no one even thinks about it.) But your patient is also on medicines for heart failure, diabetes, and hypertension, and no one in the GERD treatment group was on these medicines. So some of the GERD medicines may be absolutely contra-indicated in your patient, but it will take time to explain this. Some doctors just write in the office or hospital chart: "This patient refuses to take the following medicines and tests" if the doctor does not think that those particular medicines or tests are either not indicated or might be harmful. This saves him/her from having to answer many letters, and no one from the HMO is permitted to question the patient about this because to do so would violate HIPAA rules. I might also note that the American Assn of Family Practice, the American College of Physicians, and the American Gastroenterological Assn. all have different recommendations about which procedural (i.e. invasively performed by a gastroenterologist) tests should be done to diagnose and follow GERD.

     There are also other possible developments. A new study might be done after the guidelines were published which either introduced a new and better drug for an old problem (Singulair for asthma, dabigatran for anti-coagulation) which can either add an intermediate step to the treatment step-ladder (Singulair before oral steroids) or are safer than an older drug (dabigatran  instead of Coumadin). Sometimes a guideline has an unintended result: Many years ago Medicare decided that no one needed to be transfused just one unit of blood. They started to sanction doctors who did so. Immediately many patients who a doctor thought should have one unit of blood was transfused  two units, thereby doubling the patient's risk of a reaction to the transfusion, e.g. an infection with a disease not screened  for in the transfused blood, such as Chagas' Disease.

     Then we have the problem when the government recommendations directly contradict the clinical evidence. What comes immediately to mind are the many studies that have shown that the antibodies induced by the flu vaccine are useful defensively for only about 6 months. There was even a recent study published in the New England Journal of Medicine that showed that an injection of 4 times the usual dose of flu vaccine provided useful antibody levels for 12 months. It would therefore follow, from elementary pharmacology, that the cheapest way to provide year-round flu coverage would be to immunize every patient with a single dose every 6 months. Yet no insurance plan, HMO, or Medicare will pay for this. Why not? Doesn't the government believe the results of its own clinical tests? Don't they want to keep people as healthy as possible?

     The real problem is that each and every patient is unique, not only in his/her DNA, drug clearance ability, level of hormones, etc. but that each patient has different underlying diseases, is on different drugs, and takes different over-the-counter supplements. They may be smokers or vegetarians. They may be fat or thin. They may be very old or very young. They may be male or female. They may be pregnant or have the possibility of becoming so. They may be on birth control pills (patients on BCP's are never in a study group to test a new drug). They may be  black or brown or yellow or white. They may be on cancer chemotherapy or on medicines to treat heart failure. They may not want to be on "too many" drugs. Outside of a blood transfusion for severe anemia and supplemental oxygen for hypoxia (and be very careful with the amount of oxygen given to premature babies!) there are few treatment recommendations that can be applied uniformly, especially when it comes to prescription drugs.

     Finally, almost all guidelines apply to a patient with a given disease or symptom. None of them takes into account that a patient may already have four other diseases/conditions  for which he/she is being treated, and that the guidelines to treat this fifth disease will almost certainly conflict with some of the requirements of other guidelines. The typical 60 year-old patient is on 6 or more prescription drugs, and NO guideline has studied a group of patients on the patient's particular group of drugs.

     So it comes down to medicine being an art as well as a science, but you cannot subjugate art to check boxes on a computer's template, or note the reaction of a patient to a question, or tell the computer that the spouse insisted on being  in the room with the patient and that  therefore all the answers may not be true, or that the patient requested that certain answers not be entered into the office chart.