As I recall, I was on Service 2 as a third- or fourth-year resident. I was extremely tired. I was presenting a patient to my attending, who was extremely well-published. He knew the literature incredibly well. (He’s one of these men who can quote an article he read five years ago and tell you what journal it was in and who wrote the article.) After my presentation, he asked me what I wanted to do with the patient. I gave him an answer and then he looked at me and asked what was the data to support my answer. I had no clue. I just knew this is what we had done before. It seemed to work. Here is when he began to lecture on data versus anecdote. He stressed how important it was to know the difference. I think he really needs to lecture to the whole United States. We don’t know the difference between data and anecdote.
Say we take 300 children that all have “colds” and we give half of the children a cough suppressant and the other half a placebo. Our results, if correctly tabulated, would be data. If I told you that I personally have had lots of colds over my life and I use Robitussin because it works better and I get well quicker, this is an anecdote. Anecdotes are nice. They can help make a point. Anecdotes should never carry the same weight as data.
This brings me to an opinion piece in the Wall Street Journal. This is written by a 67-year-old woman who is a breast cancer survivor. She’s had two recurrences. She’s had several rounds of chemotherapy using multiple drugs. At her last recurrence she was found to have several tumors in the liver. She writes:
Thankfully, a clinical trial had recently started and I was selected for a combination of Xeloda and Avastin. Xeloda is a chemotherapy pill that kills cancerous cells. Avastin cuts off blood-flow to tumors. Almost immediately two of my tumors disappeared. The duel-pronged approach appeared to be working.
Nearly three-and-a-half years have passed. Today, I’m not just living life; I’m enjoying it. While I get tired more easily than I used to, I can take part in just about all normal activities. I can go out to dinner with my husband. I can visit with friends.
The Wall Street Journal tells us nothing about this lady. She seems to be a nice lady who can write fairly well. She has metastatic breast cancer which is incurable. The Wall Street Journal decided to print her anecdote but they didn’t print any data to give the reader some perspective. Why? If the Wall Street Journal told you that the drug Avastin was shown in a wonderfully designed trial with over 700 patients with metastatic breast cancer to have no significant effect on overall breast cancer survival, would we read this lady’s anecdote a little bit differently? Of course we would. Had the Wall Street Journal told you that treatment with this drug can cost 50,000-$100,000 a year would that have changed your mind about the anecdote?
This brave woman who is fighting breast cancer is arguing to save the drug Avastin (several trials showing no benefit in breast cancer) which is going on the chopping block in front of the FDA. She is arguing that this drug has saved or prolonged her life. We don’t know this. All we know is that she had a remarkable response to double drug therapy. This is really all we know. It is critical that we make decisions, especially in healthcare, based on data. Remember when we, as a country, went completely crazy over Terri Schiavo? She was one patient with a severe brain injury. There are over 100,000 patients are hospitalized every year with brain injuries. I don’t know the percentage of these patients who are in a persistent vegetative state like Terri Schiavo. The number is clearly in the thousands. Yet, we had a special law drafted and passed in Florida based on her experience and not the collective experience of thousands. Congress was then moved to act based on her experience and not on the overall data. We have to do better.
I’m not saying that we shouldn’t be moved by individual stories. I like hearing the stories. But it is important for us to know the data before we start drafting legislation.