Saturday, July 5, 2014

Using guidelines? Think carefully

  1. Guidelines are meant to help the doctor deal with an individual patient but they are based on information obtained from groups of patients. Description of groups has to rely on statistics using averages to describe various characteristics and results.
  2. When you apply the results of various medical interventions obtained in a group (frequently in the thousands) to your individual patient, you are presuming that your patient is similar to the patients in the group so that you expect him to respond to the treatment in a similar way to that of the group. But the response of different patients in the group is not the same. You do not know the characteristics of various individuals in the group and how each one responded. What you may know is the averages of their characteristics and the averages of their responses. Those who wrote the guidelines for you based their guidelines on these.
  3. The critical question then is whether your patient is sufficiently similar to the average of the group (or the groups) studied so that you can reasonably apply the guidelines to him.
  4. This can best be explained by an example. Guidelines on the treatment of non valvular atrial fibrillation using CHAD or CHADS2 score lump together patients with permanent atrial fibrillation and patients with paroxysmal atrial fibrillation. Treating the two types similarly is based on studies that showed similar prognosis regarding their liability to develop strokes. The cause of this is not clear but it may be related to the fact that the patient with paroxysmal fibrillation is more prone to develop a stroke  when fibrillation reverts to sinus rhythm and this may offset his decreased risk when he is in sinus rhythm. The vital question is this: are we justified to treat a patient who has two or three attacks a year each lasting several minutes in the same way as a patient who has daily (or every few days) attacks each lasting many hours because both patients carry the same label of paroxysmal fibrillation?!
  5. Applying guidelines to your patient without considering his individual characteristics and circumstances is like a tailor who, when asked to make a suit for an Iraqi man, does not take his measures; instead he makes the suit relying on statistical figures describing the physical characteristics of Iraqi men. If the man happened to be much taller (or shorter) than the average Iraqi, it is just bad luck!
  6. This does not mean that we should not use guidelines but that we should take into consideration the individual characteristics and circumstances of the patient we are treating. You should ask yourself whether your patient is sufficiently similar to the average patient with that condition to justify applying the guidelines in his case. If not, you should modify your application of the guidelines according to your judgment of the magnitude of the difference.