Book Excerpt

You Can’t Trust What You Don’t See

Perhaps the best thing about being a doctor is seeing a suffering patient get well under your care. One of the most frustrating things about being a doctor is caring for a patient and not seeing him or her get any better. Often, patients do great. Sometimes, they don’t. Sometimes, the treatment just doesn’t work nearly as well as expected. And sometimes even when the treatment does well at first, after a while it doesn’t. In fact, one of the basic principles of my specialty, dermatology, is that after a while, the most commonly used medications—topical cortisone medicines— simply stop working.

The discovery of cortisone in 1948 revolutionized medicine and secured the 1950 Nobel Prize in Medicine for the discoverers, Philip Hench, Edward Kendall, and Tadeusz Reichstein. Kendall, a chemist, described the impact of the new medication: “On September 21, 1948, cortisone was administered to a patient who had rheumatoid arthritis. During the preceding five years she had continued to become worse and at the time that cortisone was given she was confined to her bed and endured much pain. Within a week she walked out of the hospital in a gay mood and went on a shopping trip for three hours without after effects” (Kendall 1953). As effective as cortisone was, even more powerful derivatives were created soon after. In the 1950s, cortisone formulations were developed that could be rubbed on the skin. These topical medications—medications applied to the skin—dramatically changed how skin diseases were treated.

While topical cortisone medicines provided a tremendous advance in the treatment of skin disease, their tendency to lose effectiveness over time limited their usefulness. This was a particularly difficult problem for people with chronic, incurable conditions like the one I specialize in treating: psoriasis. Psoriasis is a condition that causes red, scaly spots on the body. It is usually treated with topical cortisone medicines. The eventual loss of efficacy of topical cortisone treatments is so common and well characterized that it has a name: tachyphylaxis. Teachers of dermatology have, for generations, taught their students that tachyphylaxis can be defined as “the more you use the medicine, the less it works.”

Various treatment approaches were used to minimize the tachyphylaxis problem. Sometimes doctors would prescribe topical cortisone medicines to be used intermittently so the patient’s disease wouldn’t get used to the medicine. Sometimes treatments would be rotated from one approach to another in order to prevent patients from getting used to a particular treatment.
The concept of tachyphylaxis is simple enough. As people continue to use strong cortisone medicines, tolerance develops, and the medicine eventually loses effectiveness. Cortisone drugs work by binding to cortisone receptors in the skin. The loss of effectiveness makes perfect sense if continued exposure to cortisone causes the receptors for cortisone to disappear or to become less sensitive to the drug over time. This receptor desensitization theory provided a good explanation of why people who used the medication for long periods eventually developed tolerance to the drug.

But one thing didn’t quite fit the theory. If a patient became resistant to one cortisone drug, the doctor could prescribe a cortisone of a different brand—yet still of the same potency level—and the new drug would work just fine. The doctor didn’t necessarily need to prescribe a higher-strength cortisone. If patients had become resistant to topical cortisone because their cortisone receptors were reduced in number or in function, changing the brand of the cortisone should not have overcome the tolerance that had developed.

A small, simple research study turned theories about the loss of efficacy of topical treatments upside down. In the study, patients with psoriasis were given a gel to apply (6 percent salicylic acid gel, a safe medication commonly used to remove the thick scales that cover psoriasis spots) (Carroll 2004). The patients were told to use the medication twice a day. They were told that their use of the medication would be monitored, and they were asked to complete a daily diary of their use of the medicine. They also were asked to bring the medication back at return visits so the medication could be weighed. They were not told, however, that there were computer chips in the medication bottle caps recording their use of the medicine. Those computer chips didn’t just record the number of times the medication bottles were opened; the monitors recorded the day and time each time the bottle was opened or closed to see when patients actually used the medication.

The study found that the patients overreported their use of the medication (and that’s being generous). Some patients who hardly used the medicine at all recorded using it almost exactly as had been directed. One of the most important and interesting findings was that the use of the medication dropped steadily over time. Although the patients claimed they had been using the medication regularly as instructed, use of the medication dropped by about 20 percent every five weeks. The study lasted a total of eight weeks. We can’t know for sure what would have happened to the use of the medication after the eight-week study, but if that decrease in the rate of use continued, patients could be expected to stop using the medication entirely in twenty-five weeks or in about six months.

Dermatologists prescribe patients medications but don’t get to see what patients do with the medicines. Doctors prescribe in one compartment—their offices; patients use the medications in another compartment—at home. It had long been assumed that because patients’ skin disease bothers them that they would use their medications. It turns out many patients don’t use their topical medications as directed, and over the long run, their use of medication steadily drops. But dermatologists didn’t know this. Dermatology textbooks didn’t say anything about patients not using their medications.

Theories about how medications lose effectiveness over time changed completely. Dermatologists originally thought tachyphylaxis was “the more you use the medication, the less it works.” But tachyphylaxis was really, “the less you use the medication, the less it works.” Generations of brilliant professors of dermatology had been teaching their students the wrong thing, and generations of students of dermatology had accepted the traditional concepts. Dermatologists had no way of knowing that patients weren’t using their medicines until the development of electronic monitors in medication bottle caps.

The electronic evidence of patients’ use (or nonuse) of their medication didn’t just change theories about tachyphylaxis. Theories about all sorts of other dermatologic phenomena went kaput (Ali 2007). There was the phenomenon of children with eczema—an itchy, dry skin condition—that wouldn’t get better despite all attempts at outpatient management (Krejci-Manwaring 2007). These patients wouldn’t improve despite the use of potent topical or even oral medications. If the frustration with treatment and the severity of disease mounted high enough, these patients would be admitted to the hospital for treatment of the intractable skin disease. They would be treated with rather modest topical treatments, and they would clear up dramatically in just two or three days! For a long time dermatologists taught that these children probably did better in the hospital because they were away from the stress of the home environment, as if trying to sleep in a hospital bed is less stressful for a child than sleeping in his or her own bed at home. Now we know that hospitalized children with eczema improve, and they improve fast because in the hospital someone ensures that the medication is applied as prescribed.

There had been so much dogma about skin diseases and their treatment; some of which was patently ridiculous, and it was built up like a house of cards. All these ideas depended on the assumption that patients use their medications as directed. But you can’t depend on assumptions that you can’t test. A whole host of theories can be wrong, collapsing under the weight of a single inaccurate assumption. Learned teachers—even though they were brilliant, caring, and honest—could be completely mistaken.

We ought to be circumspect when we try to draw conclusions about things of which we don’t have firsthand knowledge. The immortal words of Donald Rumsfeld come to mind (Seely 2003):

As we know, there are known knowns. There are things we know we know. We also know there are known unknowns. That is to say, we know there are some things we do not know. But there are also unknown unknowns, the ones we don’t know we don’t know.

We recognize that known unknowns are a problem. Rumsfeld cautions us that there may also be unknown unknowns that may cause us to stumble. But perhaps our biggest problem may be that the things we know we know just aren’t true.