“Here we are again, Curmudge, back in Kaizen Curmudgeon. A week ago we thought our third discussion of evidence-based medicine would be too technical for this blog and would have to go in Curmudgeon’s Wastebasket.”
“Well, Jaded Julie, I asked Mrs. Curmudgeon to review this. If she can understand it, anyone…”
“Stop right there! One more word and you might not get supper tonight. So back to our topic, how I can learn more about my disease, become a more knowledgeable patient, and achieve a better outcome. I guess you’ve been reading the scientific literature for a long time, Curmudge.”
“Reading a lot and writing a bit for about 50 years, Jaded Julie. But clinical research is different from the industrial type, and the layperson needs some guidance in order to learn from it. When we last talked about this, you were playing the role of a patient with a disease who had learned how to use the major secondary sources of information such as Cochrane Reviews, emedicine, and clinicaltrials.gov. Today we are going deeper into the subject. Oh, and one thing further. If anything in your reading leads you to suspect your diagnosis, share it with your physician. There’s not much value in a patient’s studying a disease that she doesn’t have.”
“So here I am, a patient with the full text of what looks like a valuable article in hand or on my monitor. How do I proceed?”
“The sections of articles in most disciplines follow a similar order dictated by the journal: Abstract, Introduction, Experimental Methods, Results and Discussion or Outcomes, Conclusions, and Literature Cited. What you should learn in each section is described in Gaeta and Nagurney’s Evaluating the Literature. Your first objective is to decide if the article is really pertinent to your interest; that should be evident in the Abstract.”
“But, Curmudge, what if the Abstract is a bunch of medical gobbledygook?”
“It shouldn’t be if the article is truly pertinent. Your previous reading of reviews and other secondary sources should have taught you a lot of essential terms. The Introduction will describe what is unknown about the topic and how the proposed research will fill in some of the gaps in our knowledge. From the Methods section you will learn which of the several types of experimental design was used. Common designs include retrospective (looking backward at patient records) and prospective (following subjects forward in time and collecting data as they are generated).”
“My guess is that the design will tell me a lot about the validity of the results.”
“It should, Julie. Let’s assume that you are looking at a prospective study of a treatment that might cure your disease. The findings should be more meaningful if the treatment groups were randomized with everyone equal at the baseline, double-blind with neither patients nor physicians knowing who received the treatment or a placebo, and accounting for all patients.”
“I’m still with your train of thought, Curmudge. The next stop should be Outcomes. All aboard!”
“Look at the figures and tables of results. You might not know all of the words, but an upward or downward trend means something happened; and it’s often a good something. In my earlier life I was sometimes able to interpret figures even if the captions were in a foreign language.”
“So then, how will I know that the study’s findings were not just by chance and that the treatment might really do me some good?”
“Julie, authors usually use statistics to characterize their results. The so-called p-value provides a useful guide for the reader. The smaller the p-value, the more significant the difference between results from the treated group and the control group. Look for p-values less than 0.05, or even better, less than 0.01.”
“Curmudge, you can’t imagine how delighted I am that you didn’t go any further into statistics.”
“Various organizations have adopted formal guidelines for evaluating the literature, including the evidence grading system from the Institute for Clinical Systems Improvement and the Infectious Disease Society of America—US Public Health Service Grading System for ranking recommendations for clinical guidelines. The latter system uses letter grades (A-E) for strength of recommendation and Roman numerals (I-III) for quality of evidence. You may have encountered these grades in your reading of review articles. In general, if you are interested in a specific intervention, look for prospective, randomized, controlled clinical trials.”
“You have made my brain very crowded, Curmudge, but I believe I can proceed with a little more confidence. Do you have any final words of wisdom?”
“Never final, Julie, but I do have two additional comments and one caveat. First, if you see an abstract to a promising article but can’t obtain the full text, see your medical librarian (Margo Lambert at St. Elizabeth Hospital or Michele Matucheski at Mercy Medical Center). Second, many widely used treatments have never been subjected to a controlled clinical trial and are not, strictly speaking, evidence based. They have been used successfully over time and represent the consensus standard of care; one might consider them to be ‘grandfathered.’ The caveat is that there seems to be nothing in the clinical trial literature about the eventual cost of a particular treatment or medication. So don’t allow your hopes to get too high. What does one do after learning that he can gain an extra four months of life at a cost of $93,000 per year?”
“Wow! That’s a problem in medical ethics that neither of us would touch with a ten-foot pole.”
Affinity’s Kaizen Curmudgeon
Note: An observation on the practice of emergency medicine in a primitive location may be accessed via this link.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment