Beware the second opinion

Machiavelli's Laboratory is a free ebook that I published on April 13, 2010. It is a satiric discourse on scientific ethics, from the perspective of an unethical scientist. Please don't take any of the advice and opinions in the book (or the excerpts featured in this blog) seriously.

In a recent New York Times Article, Stephanie Saul told the story of a woman who had been given the diagnosis of DCIS (ductal carcinoma in situ, or non-invasive breast cancer), who subsequently received treatment (surgery, radiation and drugs), and who was later informed that the diagnosis was incorrect: she never had breast cancer.[1]

The diagnosis of DCIS was rendered by a pathologist who was not a breast cancer specialist, but who had gone to the trouble of soliciting a second opinion, from another pathologist. The consulting pathologist, who worked at a nearby hospital, agreed with the first diagnosis of DCIS.

The misdiagnosis came to everyone's attention when the woman transferred her follow-up treatment to another hospital. Her records and pathology slides were sent to the transfer hospital, where a staff pathologist reviewed the pathology material. He found no evidence of cancer in the original biopsy or in the portion of breast that was subsequently removed at surgery. The slides were sent to several expert consultants, all of whom agreed that no tumor was present.

Not surprisingly, the case became the subject of legal action. The lawyer for the plaintiff indicated that the original pathologist [who misdiagnosed the case] could have easily sent the slides to an expert consultant [not to a non-expert pathologist from a nearby hospital.] It is likely that, should this case ever reach a courtroom, debate will center on whether the first pathologist exercised standard of care by seeking a second opinion from a pathologist who had no special expertise in breast cancer pathology.

Setting aside the specifics of the case described in The New York Times article, the broader question of expert consultants deserves the scrutiny of every evil scientist. As you shall see, the issue of choosing expert consultants has ramifications far beyond the small world of pathology.

First, you must understand that getting a second opinion is a no-win situation. If your consultant agrees with your original opinion, nothing was gained. Everyone will assume that the diagnosis must not have been very challenging. You will learn that your opinion was actually superfluous. Only the consultant's opinion carries any real weight. Eventually, after a number of consultations, your employer may wonder why you are needed at all. Why not just pay the consultant to render diagnoses on all your specimens; in which case, your position can be terminated.

Of course, if the consultant disagrees with your opinion, your life will deteriorate rapidly. Everyone will assume that you are wrong, and that the consultant is correct. Your credibility will be destroyed. You may become the center of attention in a lawsuit. You will become a pariah among your co-workers. You may lose your job. Your permanent record will be blemished. You may become unemployable.

As an evil scientist, with no great competence, and prone to committing errors, you must learn to defend yourself. Follow the following recommendations, without deviation.

1. Pick a consultant that you trust to watch your back. This will almost always be someone who asks you to be his consultant when he needs a second opinion. A reciprocal consultation agreement virtually guarantees a non-threatening opinion.

2. In any field, most tasks are routine. In the specialty of pathology, 90% of cases are trivial and can be rendered correctly, and with complete confidence, by the most marginal pathologist. The remaining 10% of cases can be handled with equivocations ("The biopsy is consistent with but not definitely diagnostic of..."), temporizations ("A final diagnosis will be rendered after special stains and molecular tests are completed."), supplementations ("A few atypical cells are noted, but there is insuffficient material for an unequivocal diagnosis. I suggest re-biopsy."), and consultations ("We are waiting for a consultative report").

3. It is best for the marginal professional to eliminate the concept of a final diagnosis, entirely. Restrict yourself to vague pronouncements. On the official report, in the space normally reserved for the "Diagnosis" field, you must substitute the heading, "Diagnostic Impression." For example, "Diagnostic impression: Neoplastic process suggestive of invasive ductal carcinoma." Astrologers and palmists have perfected this dodge. Remember that the official tree of the evil scientist is "the hedge."

When you've picked your trusted consultant, you will need to be prepared to prove that he is an expert. This really is not very difficult, because nobody really knows how to determine whether any person qualifies as an expert. If you doubt this, try to answer the following:

1. Are experts infallible? Of course not. Experts are human and can make diagnostic errors.

2. Do experts make fewer errors than non-experts? Nobody really knows. In the medical field, there is no mechanism by which the diagnoses of experts can be evaluated.

3. Do experts routinely send their difficult cases to other experts? No. People who consider themselves experts tend to have high confidence in their own diagnostic acumen. Though studies have shown that overconfidence is the most common cause of medical error, the expert assumes that this caveat only applies to non-experts.[2]

4. Do experts concentrate their efforts on a specific area, gaining expertise by seeing a large number of cases in their chosen field? Often, yes. But you must remember that simply seeing a large number of specimens does not prove that you're not an idiot. An incompetent pathologist can consistently misdiagnose a large number of specimens, just as easily as he can misdiagnose a specimen that he rarely encounters.

Basically, an expert is someone who is called an expert by others in his own field, such as yourself.

Given that experts can be mistaken, would it not be advisable to consult with more than one expert on very difficult cases? No, no, no, no. The absolutely worse thing you can do is to send a specimen to two experts. If they disagree, which expert will you believe?

Case in point. In 1993, Reggie Lewis was the 27 year old captain of the Boston Celtics. Mr. Lewis collapsed during a basketball game, and was treated by several teams of expert cardiologists. The team assembled by the New England Baptist Hospital believed that Mr. Lewis had focal cardiomyopathy, a life threatening condition requiring Mr. Lewis to retire from basketball. A second team of experts, assembled at the Brigham and Women's Hospital, disagreed. They rendered a diagnosis of vasovagal fainting, a benign condition. A third team of experts, from St. John's Hospital in Santa Monica, California, was non-committal. The Santa Monica team suggested that Mr. Lewis play basketball, but with a heart monitor attached to his body. With three discordant diagnoses, Mr. Lewis decided to take his chances, continuing his athletic career. Soon thereafter, Lewis died, quite suddenly, from cardiomyopathy, while playing basketball.[3]


References

1. Saul S. Prone to error: earliest steps to find cancer. New York Times July 19, 2010.

2. Berner ES, Graber ML. Overconfidence as a cause of diagnostic error in medicine. American J Med 121:S2-S23, May 2008.

3. Altman LK. After a highly publicized death, second-guessing second opinions. The New York Times August 3, 1993.

- © 2010 Jules Berman

key words: consultations, second opinions, diagnostic disagreement, discordant diagnoses, expert opinions, misdiagnoses, misdiagnosis, pathology error