Depression: Simple questions, complex answers
From the April ACP Observer, copyright © 2007 by the American College of Physicians.
By Jessica Berthold
At first, the outlook seemed hopeful for the 45-year-old depressed patient of Dr. Harrington, ACP Member, an internist and geriatrician at the MetroHealth Medical Center in Cleveland. After taking an antidepressant for a few weeks, the patient appeared more animated in a follow-up visit and said she felt better. Dr. Harrington increased the dose, then saw her in another month. Again, the patient had improved, though she reported some non-specific side effects, like a slight headache and malaise.
Noting that the side effects were vague, Dr. Harrington suggested increasing the dose. The patient balked, fearing her side effects would worsen. Dr. Harrington then switched her to a different antidepressant, and the same pattern emerged. "We'd get the dose going and it'd seem to be working, and then she'd develop some sort of side effect after a month," Dr. Harrington said. "She didn't buy into the theory that you just give it a few more weeks and you'll feel better and those depression symptoms will improve."
Primary care doctors often face challenges like these—and more—in treating depression, from knowing when to switch medications or doses, to following up with patients about side effects. Indeed, the majority of depression sufferers are initially seen by primary care doctors—and about 10% of the U.S. population suffers from a depressive illness in any given year, according to the National Institute of Mental Health.
"We've seen a dramatic increase in depression diagnoses in primary care in the last 15 years, as better training and friendlier medications make physicians more comfortable with making the diagnosis," said Wayne Katon, professor and vice chair in the department of psychiatry and behavioral sciences at the University of Washington Medical School in Seattle.
Recent research, published in the November 2006 American Journal of Psychiatry and funded by NIMH, offers new insight for internists on how to treat depression. The Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial found that patients who didn't improve much on a medication at first might eventually respond with more time or a higher dosage, or if switched to a different treatment entirely. Indeed, two-thirds of patients eventually remitted if enough time passed, or if up to four treatments were tried—including drugs, cognitive therapy, or a combination of the two.
On the downside, the study also found that the odds of remission grew slimmer, and the chance of relapse higher, with each new treatment—suggesting there may be considerable advantages to finding the right treatment the first time around.
But that's not always easy. The path to improvement is rarely a straight line, many physicians say, and every roadblock lowers a patient's chances of reaching the desired destination.
Rules get murky after diagnosis
"If someone is sitting across from you and smiling at you, he or she could easily be depressed," said William R. Greer, FACP, a general internist in solo private practice in a suburb of Philadelphia. "Everyone knows that, but you sometimes forget it when you are getting through your day trying to find lung cancer."
"If someone is sitting across from you and smiling at you, he or she could easily be depressed. Everyone knows that, but you sometimes forget it when you are getting through your day trying to find lung cancer."
—William R. Greer, FACP
In an attempt to cast a wide net, Dr. Greer said he asks every patient a few direct questions about depression at his or her annual physical. For patients who visit with specific complaints, he asks an open-ended question about how they feel. "Asking the question takes almost no time, and a lot of patients do reveal depressive symptoms this way," Dr. Greer said.
Physicians can choose from a number of proven diagnostic instruments, such as the Primary Care Evaluation of Mental Disorders, or PRIME-MD. The criteria for Major Depressive Episode based on DSM-IV states that major depression is confirmed if a patient meets at least five out of nine symptoms—such as insomnia, significant unexplained weight loss or gain, or diminished interest in activities—nearly every day for at least two weeks.
The most seriously ill patients, such as those who appear to be suicidal, should be referred immediately for psychiatric care. But in many cases, it is the primary care physician who must decide on how to initiate treatment. "I usually start by asking if they have a preference," said Dr. Greer, "medication, therapy or both together."
For mild or moderate depression, all three are equally effective, the American Psychiatric Association says, but patients with severe or recurrent major depression do best combining therapy and drugs. Cognitive-behavioral, behavioral, interpersonal and brief dynamic therapies work equally well, according to the APA.
Choosing a medication can be challenging because it's not an exact science.
While there's fairly good evidence to indicate that drugs are effective, choosing a medication is not an exact science, said A. John Rush, MD, lead author of the STAR*D study and professor of psychiatry and clinical sciences at the University of Texas Southwestern Medical Center in Dallas. "There are no real good indicators for which drug fits which patient; there's no lab or genetic test."
The STAR*D study confirmed previous findings that no one antidepressant is more broadly effective than another at treating depression. Most physicians prefer to start with selective serotonin reuptake inhibitors (SSRIs), or "second generation" agents, which came on the market after tricyclics and MAOIs (monoamine oxidase inhibitors), because they have been shown to have fewer side effects and carry less risk of lethal overdose than tricyclics.
Beyond that, there are a few other key factors to consider when deciding which drug to prescribe:
Co-morbidities. Nearly sixty-two percent of patients in the STAR*D trial had a co-morbid psychiatric disorder—a reasonable estimate of what one sees in real life, said J. Craig Nelson, MD, professor and director of geriatric psychiatry at the University of California-San Francisco, and author of an accompanying editorial in the American Journal of Psychiatry. Many co-morbid disorders, like obsessive-compulsive disorder and attention-deficit hyperactivity disorder, can be treated with antidepressants.
"A drug like bupropion is moderately effective to treat ADHD, and is also effective with depression, so you might pick that for a depressed ADHD patient," Dr. Nelson said.
Co-morbid medical conditions are also a consideration, due to the physical effects of some antidepressants, said Bruce L. Rollman, ACP Member, associate professor of medicine and psychiatry at the University of Pittsburgh School of Medicine, who has done extensive research on treating depression in primary care. "In cardiac patients I'd use an SSRI before an SNRI (serotonin-norepinephrine reuptake inhibitor), because the latter has been found to slightly raise blood pressure," he said. "Tricyclics are contraindicated in cardiac patients as well."
Side Effects. People with insomnia might benefit from a sedating antidepressant, like mirtazapine, while patients who are generally prone to nausea from drugs should avoid antidepressants with that potential effect, several physicians said. Some patients may have preferences, too: one may be more bothered by loss of libido than weight gain, while another may feel the opposite.
Ralph Lanza, ACP Member and general internist in private group practice in Paoli, Pa., recently prescribed bupropion for a 45-year-old female patient, though he usually prescribes SSRIs for younger, first-time antidepressant takers. "She was very overweight, and bupropion tends to cause less weight gain than others," Dr. Lanza said. "She was also on triptans for migraines, and there is a small interaction risk with SSRIs and triptans."
Drug history. Because depression tends to recur, some patients may have been successfully treated in the past. "If they had a previous episode and some drug worked well, I'd go to that one right away, because drugs seem to work again if they worked previously," Dr. Rush said.
Brand preference. Some patients will state preference for a certain brand. Whether they heard about the drug on a television ad or have a friend who takes it, belief in a drug's efficacy is important. "If a sister or brother or spouse did great on Drug X, then all things being equal, Drug X is a good one to choose, because the patient believes in it," Dr. Rush said.
Insurance. Many patients' insurance companies have tiered prescription plans that cover more of the co-pay for a generic than a brand drug, Dr. Lanza noted.
When to switch course
Figuring out whether and when to change doses, or switch medicines, can also be a challenge. Most physicians tend to switch drugs that don't bring significant improvement after six to eight weeks, several doctors said. Yet in the STAR*D study, there were patients who, after showing as little as 30% improvement in two months, continued improving in the third month—suggesting it might be wise to wait on switching in some cases.
The study also found that the more treatments a patient tried, the less likely it was that the treatments would work, particularly after the second try. At the same time, severely depressed patients, and those who had concurrent disorders, required more treatment steps to improve or remit, the study found. Ultimately, "clinicians must decide when remission (given our current treatments) is sufficiently unlikely that subsequent alternative treatments should be considered," the study's authors said.
Kurt Kroenke, MACP, professor of medicine at Regenstrief Institute, Inc. and Indiana University in Indianapolis, said that if a patient has no response at all from a drug within two weeks, he increases the dosage. If he sees a partial response after two weeks, however, he might wait to see if the patient continues to improve over the next 4-6 weeks, and eventually raise the dosage if the patient plateaus.
"The reality is that if someone has had a partial response, it makes sense to increase the dose, but if the person has had no response from increasing, it doesn't make sense to keep going to mega-doses," Dr. Kroenke said.
Measurement tools are the best way to evaluate patient response to a drug, he said. A 9-item version of the Patient Health Questionnaire (PHQ), which Kroenke helped develop, is available online. The STAR*D study used the 16-item Quick Inventory of Depressive Symptomatology-Self Report (QIDS-SR), which is available online.
Clinician ratings or patient self-reports are both fine options to assess a drug's effectiveness, Dr. Rush said. "If you ask patients how they are doing, they may say they aren't much better, because a 30% reduction in symptoms is hard to recognize. But if the scale measures a 30% drop, that indicates they should probably stay on the drug and just increase the dosage," Dr. Rush said.
Clinicians should aim for an assessment that indicates remission, the STAR*D study suggests, as patients who remitted had a better prognosis than those who improved without remission. Patients who try more treatment steps take longer to remit, STAR*D found.
Primary care doctors in the study didn't view using measurements as a burden, Dr. Rush said. "When we put them into practices that hadn't used them before, the physicians said they found it very helpful and easy to do," Dr. Rush said. "They said they'd keep doing it after the study was over."
The importance of keeping tabs on patients was a major lesson from STAR*D, Dr. Rush said. An increasing percentage of patients left the study after each treatment trial-from 20.9% after trial 1 to 42.3% after trial 3. "Perhaps patients that drop out are becoming demoralized with each failure and are giving up," Dr. Nelson said.
Other studies have indicated that patients are apt to quit taking their medication within the first three months, Dr. Rush noted. Many also stop within the first three weeks, when side effects first appear, or when they begin feeling better and assume they no longer need the drug, he added.
Some patients get discouraged if the pill doesn't work right away, Dr. Kroenke noted. "I tell my patients from the get-go that there are a number of medications we can try if the first one doesn't work. But I also tell them that you need to try the first pill for a few weeks, and that the side effects often go away after a week or two."
Indeed, when doctors discuss potential side effects with patients, they are less likely to stop taking antidepressants, a Sept. 18, 2002 study in the Journal of the American Medical Association found. Visiting a doctor three or more times in the 12 weeks after starting medication also significantly improves adherence, the study found.
"Following up three times in the first 12 weeks is not a bad metric," Dr. Kroenke said, adding that there is scarce evidence-based research on how often to follow up. "After that, if the patient seems to be responding, I follow up once or twice more in the first year because of the risk of relapse. You can easily have a phone follow-up, and it doesn't need to be the physician calling. It can be a nurse."
Yet for many internists, it's hard to find time to see patients so frequently. "Being in a busy practice, it's kind of tough," said Dr. Lanza. "Sometimes I'll put the onus on them, and tell them to call if they are having difficulty on the medication."
On the whole, said Dr. Lanza, his follow-up routine varies by patient. Patients he's concerned about he brings back in one week, then once a month until the person stabilizes. Patients who seem more stable, and who don't have significant life struggles, he will see a month to six weeks after starting the antidepressants, Dr. Lanza said.
Dr. Greer said he sees all patients after one month. To enforce the follow-up visit, he'll write an initial prescription for only four weeks, or give out only four weeks' worth of drug samples. And if a patient doesn't show up for the visit, he calls.
"It takes about four weeks to see the full effect of an SSRI, so I don't think they need to come back before four weeks," Dr. Greer said. "When I prescribe the medication, I warn about diarrhea or other side effects, and about the possibility they will feel more depressed. They tend to call me on their own if there's a problem with side effects."
The frequency of additional follow-up appointments varies by patient, he said. Those with a great response after one month may not be seen for a year; those who merely improve are seen in another four weeks.
If he feels a patient is too sick to wait a month for a first follow-up, Dr. Greer refers the person to a psychiatrist, and stays in touch with that psychiatrist about his patient's condition. "A phone call to check in can take 20-30 seconds." Dr. Greer said. "I really try to maintain a team effort with the psychiatrist as well as the patient."
As a general rule, primary care doctors should refer to psychiatrists those patients who are suicidal, resist several treatment attempts, have alcohol or substance abuse, have complex psychosocial stress, or appear to have another disease like bipolar disorder, several psychiatrists said.
"I'd say that if you've failed several times with treatments, then it's time—just like with a complex diabetic patient—to get specialists involved," Dr. Kroenke said. "I always make it clear to the depressed patient that I'm not dumping them, but am collaborating with another physician."
In the case of Dr. Harrington's patient, who kept developing side effects more than a month into new treatments, success came only when he referred the patient to a psychiatrist who used a more sophisticated treatment.
"The psychiatrist tried a different medication and also put her on an anti-anxiety drug," said Dr. Harrington. "That did the trick."
Anxiety disorders are common in primary care and undertreated, according to research published in the March 6 issue of the Annals of Internal Medicine.
A study of 965 patients in a primary care setting found that nearly 20% had at least one of four main anxiety disorders and many had more than one. Further, 41% of the patients with an anxiety disorder were not receiving treatment. Anxiety was associated with high levels of depression, functional impairment and high medical use. Researchers validated two simple tests that detected most patients with anxiety, regardless of type.
In the study, 965 randomly sampled patients from 15 U.S. primary care clinics who completed a self-report questionnaire and agreed to a follow-up telephone interview. A clinical exam was followed by a telephone-administered, structured psychiatric interview by a mental health professional who was blinded to the results.
The authors wrote, "Considering the frequency with which depression and anxiety co-occur, a search for one condition should always be accompanied by an assessment of the other."
The May issue of ACP Observer will examine how to differentiate between anxiety and other mental disorders such as major depression, as well as different diagnostic strategies and drug treatment options for the two conditions.
Internist Archives Quick Links
Superior MOC Solutions from ACP
Meet your requirements with our approved activities. See details.
Making the Most of Your ICD-10 Transition
To help you and your practice make a smooth and successful transition to ICD-10 coding, ACP and ICD-10 content developers have created multiple resources available at discounted rates for ACP members.