Patients with lupus (or any other medical illness) expect their doctors to be certain about their diagnosis, their treatment, and their futures. These are reasonable expectations, if one’s only experience with illness has been with short-term medical events. There is nothing uncertain about a fractured leg, a heart attack, or a biopsy that shows a cancer.
Lupus is different. It is a long-term problem; and certainty may be harder to find. But even in the face of uncertainty, the most important facts about lupus are known, choices among treatments are not flips of a coin, and there can always be a plan to go forward.
Here are a few questions my patients often ask; they are examples of the uncertainty of long-term illness.
The line that separates lupus from not lupus is sometimes very unclear. Yes, many people do have typical, easily diagnosed diseases, but many others do not. Some people, for instance, are found by chance to have abnormal laboratory tests, but they have no symptoms. They may never become ill. Only if they develop symptoms do we say they have lupus. Other people have symptoms, but the laboratory tests for lupus are normal or only suggest the presence of the disease. Perhaps someday the tests will become positive, but perhaps they will not. The symptoms may change to suggest a different diagnosis, or perhaps symptoms will disappear. Only time will tell.
Other people have symptoms and tests that simultaneously show several different but related diseases—rheumatoid arthritis and lupus together, for example. Should the doctor diagnose two diseases or just one, albeit in an unusual form? Often the best way to make a diagnosis is to wait to see what happens. An honest doctor might be unable to answer the question, “Do I have lupus?” with a definite yes or no.
Being uncertain is not the same as not knowing what is wrong. Symptoms and laboratory tests have to pass a threshold of abnormality to give a clear diagnosis. Sometimes the patient has symptoms, but the symptoms are not severe enough to pass the threshold. Some symptoms could indicate several diagnoses. Many rashes occur in the “butterfly” area of the face but only some are lupus rashes. Arthritis occurs in most rheumatic diseases. Doctors use “criteria” to diagnose lupus, but those criteria describe only typical lupus, not lupus symptoms that are rare.
Uncertainty about diagnosis occurs because the rules for diagnosis are inflexible, and diseases do not always follow those rules. Whether symptoms fit criteria or not, the doctor can still act. The patient’s question should not be, “Do I have lupus?” but “Do you have a treatment plan?”
Treatment trials enroll only patients who have uniform, common symptoms. Otherwise it would be hard to recruit enough patients to make sense of the results. Lupus treatment trials therefore mostly focus only on kidney, joint or skin disease and ignore lupus symptoms that are rare. Treatment trials also exclude pregnant patients and those with infection, prior cancer, or other issues that might confuse the results. Nonetheless, patients with these problems exist. Doctors have to treat them despite the absence of information from trials, so they make informed guesses according to their prior experiences with similar patients. Doctors must be more flexible than usual to change course if the treatment does not work or causes a complication.
There may be no actual “best medicine” but instead there may be a number of options. The patient’s best choice may be to consider many opinions—especially and importantly her own—on the pros and cons of different options before deciding what to do. 
Lupus is often unpredictable. Some patients have a single episode of the disease and spend the rest of their lives well. Others are always ill. While doctors do know something about what causes flare-ups—heavy sun exposure, for instance, or allergic reactions to drugs—they cannot predict most flare-ups. I do not know any way to reduce that uncertainty. Perhaps the best way for patients to cope with unpredictability is to live their lives as normally as possible, assuming that they will continue well, but keeping a plan to withdraw from a commitment if a flare-up occurs.
Doctors’ goals and patients’ goals have different timelines. Patients want to be immediately free of pain. Doctors have to think ahead twenty, thirty, even forty years: Will the prescription I write today have side-effects that, two decades from now, will make my patient’s life worse? Doctors will never be certain of the answer because even in long-term illness, treatment studies usually answer questions only about the next one or two or five years.
For long-term prediction, doctors rely on what they know from treatment of other diseases, from predictions based on what happened to similar patients decades before, and from guesses about what the disease and the drugs can do over time, but they cannot predict how medicine will change. A doctor in the 1960s could not have imagined AIDS, heart-lung transplants, the role of genes, or treatment based on using the body’s bacteria or “biologic” drugs. How we treat patients today will have little to do with how we treat them a quarter- or a half-century from now. While predictions are just guesses, a doctor should not knowingly prescribe something for immediate relief that he or she knows will cause harm to the patient in two or three or four decades.
Although long-term decisions for lupus involve uncertainty, uncertainty should not be a cause for frustration or despair. Rather, it should be a call for open-mindedness—on the part of the doctor and on the part of the patient—to the possibilities that may emerge. Uncertainty should be an opportunity for patients to participate as equals with their doctors in conversations about their goals.
 It is difficult to describe uncertainty in 1000 words. I deal with the topic much more fully in a book for general readers, Uncertain in Googlestan: Doctors, Patients, & Medical Ambiguity, which will be published in late 2014.
 Such circumstances are not rare. In my practice, 42% of patients do not fit treatment guidelines, and so we devise individualized treatment plans.