Testing for bias
In a study published last year in the Journal of Experimental Psychology, two researchers documented just how diagnoses for mental disorders can be swayed by clinicians' theoretical leanings. Experiments conducted with 21 psychologists and psychology graduate students showed that they held complex theories about how symptoms are interrelated. They also regarded certain symptoms as more central to a disease than others.
That runs contrary to the DSM's diagnostic model, which gives all symptoms equal weight.
The experiments strongly indicated that doctors and clinicians supplement the DSM with their own judgment. That in turn can influence their perception of the symptoms, how well they recall a patient's complaints and their final diagnosis.
In one experiment, test subjects were asked to diagnose two hypothetical cases of possible anorexia nervosa. According to the DSM, a diagnosis of anorexia nervosa requires the presence of all four of the following symptoms: refusal to maintain minimal body weight, fear of being fat even when underweight, disturbed experience of body shape and, in women, absence of a menstrual period for at least three months.
Two hypothetical patients were presented as having different sets of three symptoms. In each case, only one of the symptoms was among the diagnostic criteria for anorexia nervosa. For patient A, it was refusal to maintain minimum body weight; for patient B, it was no period for three or more months. Based strictly on adherence to the DSM, both patients had equal probability of being anorexic.
But that's not how the clinicians in the experiment saw it. Instead, they gave higher likelihood of anorexia nervosa to patients who displayed diagnostic symptoms they personally considered particularly important. A clinician who believes lack of menstruation is merely a consequence of a more central symptom — being severely underweight — gave it lower score as an indicator of possible anorexia.
What's more, the clinicians' personal theories led to selective memory. They recalled more easily symptoms they personally considered important than they did symptoms they deemed peripheral. They even falsely remembered central symptoms that had not been given.
The bottom line for patients: Diagnoses for mental disorders can hinge on the doctor as much as the symptoms.
Woo-kyoung Ahn, one of the authors of the study, said the tests provide the first empirical evidence about how theories influence diagnoses. But how does such bias play out in real clinical settings? Ahn theorizes that the impact is even more pronounced.
That's because doctors face many more ambiguities with actual patients, Ahn said, not only in diagnosing but in examining them. For example, a doctor who believes strongly that patients with severe depression feel worthless may unconsciously search for signs to confirm that. Or a doctor may suspect the patient considered suicide even if the patient never brings it up, she said.
Ahn said one way patients could combat such biases in their doctors is to be thorough and explicit in describing their symptoms.
"You might want to be very clear about the fact that you don't have suicidal thoughts," said Ahn, now a professor at Yale University .
Unclear guidelines
Peter Roy-Byrne, chief of psychiatry at Harborview Medical Center , said one big drawback with the DSM is that it offers no clear guidelines about when a symptom really is a symptom. Symptoms of the mania associated with bipolar disorder, for instance, include decreased need for sleep, distractibility and excessive talking. But how much less sleep is too little? And how much talk is too much?
Getting the diagnosis right can be especially tricky when patients are hallucinating, having panic attacks or are high on drugs.
"We have to make these diagnoses sometimes in a period of 30 to 45 minutes," Roy-Byrne said. "Sometimes it becomes like veterinary medicine. Patients can't remember details."
Missed or mistaken diagnoses for mental disorders can mean needless suffering for patients.
A two-year nationwide study led by researchers at Harvard University and published in June in the Journal of the American Medical Association found that in a given year, up to 14 million, or 6.6 percent, of adult Americans experience major depression. But only 20 percent of them get adequate treatment.
David Dunner, a professor of psychiatry and behavioral sciences at the University of Washington and a specialist in mood disorders, said doctors and patients alike too often miss the signs of depression. He estimates that primary-care physicians — who treat the majority of patients with depression — catch only 50 percent of cases. Only half of those patients get treatment and only half of that group get adequate treatment, he said.
Incorrect diagnoses can be dangerous. A bipolar patient who is misdiagnosed with depression could end up on drugs that make him sicker, said Michael First, associate professor of clinical psychiatry at Columbia University .
Bipolar patients need mood stabilizers such as Lithium or Depakote. Treating them solely with antidepressants — which raise the level of serotonin in the body — could trigger manic episodes serious enough to warrant hospitalization, First said.
First is the editor of the current, fourth edition of the DSM, which was published in 1994 (the fully revised fifth edition will not be published until at least 2010). He said the DSM provides a common language that allows doctors to reliably diagnose mental diseases that, so far, lack definitive confirmation tests. First said the DSM is a "man-made construction" that is subject to challenges from its users.
There are ongoing disagreements, for instance, on how long a symptom must last before it becomes diagnostically significant. But overall, the DSM system is "pretty reliable," said First, author of "Am I Okay?: A Layman's Guide to the Psychiatrist's Bible."
More than one 'bible'
Yet is the DSM accurate? That's debatable. Researchers know this much: When it comes to dementia, at least, more people meet the diagnostic definitions for the disorder under the DSM than with diagnostic manuals used elsewhere.
In a 1997 study published in the New England Journal of Medicine, Canadian researchers analyzed the case histories of 1,879 senior citizens to see how many met the criteria for dementia. The researchers compared six manuals, including three different editions of the DSM, two editions of the International Classification of Diseases (ICD, used in most of Europe ) and the Cambridge Examination for Mental Disorders of the Elderly (CAMDEX, used in United Kingdom ).
The percentage of the patients deemed to have dementia ranged from 3.1 percent with the current edition of ICD to 29.1 percent with the 1980 edition of the DSM. Out of 1,879 subjects, only 20 people met the definition for dementia under all six manuals. The study's authors said the conflicting diagnoses had "serious implications for research and treatment, as well as for the right of many older persons to drive, make a will and handle financial affairs."
Ahn, the Yale researcher, said medicine's mixed track record in diagnosing mental disorders supports one truism: "It's important to get a second opinion."
|
