the medical model of patient care
The issue of what constitutes the best framework for evaluating and treating mental health patients has occupied clinicians for decades. Of course, everyone is concerned about the patient as an individual person, but many clinicians point out the value of the medical model in patient care. Just as many, perhaps more, grow hot under to collar at the mention of the term, “medical model.” They shouldn’t, and I’d like to spend a little time explaining why.
I suspect that the problem lies partly in the assumption that anything medical necessarily involves medicines — that is, the giving of pills, potions, nostrums, and other trappings of the medical profession. And to be fair, that has been the way the psychiatric profession has been trending over the past several decades, towards the increasing use of psychotropic drugs (and ECT, and other physical therapies) in the pursuit of mental health, away from the provision of psychotherapy.
But that is not what "medical model" meant to me as a young clinician, and perhaps to most clinicians even today who were reared in the tradition. For us, the medical model means this: that the treatment of our patients depends on the use of a scientific approach of observation of the natural history of mental disorders and systematic, controlled follow-up study to see what works. The medical model, therefore, is a method of clinical investigation.
Here’s another source of confusion. For many clinicians, it is implicit in the phrase medical model that there must be a medical causation of the patient’s ills, and that in turn requires that we must be able to demonstrate an anatomical or physiological basis for the disorder. The analogy seems clear: tuberculosis can be diagnosed from a set of symptoms (coughing up blood, weakness, shortness of breath, chronic wasting) that are caused by infection from a tiny organism called Mycobacterium tuberculosis. The cause-effect relationship could not be clearer, yet it wasn’t established until 1882. And before then, there had been some pretty interesting theories as to the cause of TB (or consumption, as it was often called then).(One had to do with vampires that gradually sucked away the life forces of the relatives of a patient who had died of the disease.
For clinicians who insist that a physical or physiological cause be identified before we can refer to a psychiatric condition as a disease, I would point out that scientifically, we mental health clinicians are still at the stage of pulmonary physicians before the discovery of M. tuberculosis. But even then, the absence of an identifiable pathological process didn’t mean that one didn’t exist; it only meant that, at that stage of the science, we had not yet found it. It’s the same place we occupied 150 years ago when GPI was the scourge of mental patients.
GPI, or general paresis of the insane (also called paralytic dementia), was first identified as a separate disorder during the first half of the Nineteenth Century. What is it? Or, more precisely, was it, because it is so seldom encountered anywhere in the developed world today. GPI affected younger patients, and most often struck men. The initial symptoms of GPI were often weakness, headache, and insomnia—it was mistaken in its early stages for what used to be called neurasthenia. As the disease progressed (we’re talking about a period of months or years here), the patient’s judgment became impaired; trouble with concentration and memory led to eventual complete dementia. On the road to that end, such symptoms as depression, mania, delusions (perhaps of power or great wealth) led to treatment by psychiatrists in asylums. Eventually, the patient became paralyzed, unable to speak or move beyond the bedstead. Up to 12% of admissions to an asylum for men in New York were attributed to GPI.
In the early days, weakness of character was one favored explanation for GPI. But gradually throughout the late 1800s, the idea took hold that it was the downstream consequence of syphilis. The cause wasn’t nailed down until 1913, when two researchers demonstrated the spirochete Treponema pallidum in the brains of people who had died of GPI.
So, in our ancient history we have an example of a psychiatric illness that was eventually proven to be a disease by anyone’s standards. But, where are we now? Well, we aren’t left with a lot of other examples as neat as that of GPI. What we in the mental health professions do have now is a way forward.
That way is the essence of the medical model in psychiatry. That is, by identifying symptoms and signs and other features that a group of patients holds in common, we can separate out groups of patients who may have other features in common — say, outcome, response to treatment, and, eventually, etiology.
That approach was formalized in a seminal 1972 paper by Samuel B Guze and Eli Robins, in which they identified 5 steps in the characterization of disease.
1. Clearly describe the condition, including signs and symptoms as well as demographic features.
2. Laboratory studies to identify reliable features: imaging, chemistry, anatomical, and psychological testing are all included.
3. Delimitation from other disorders through careful inclusion/exclusion criteria.
4. Follow-up studies to determine that the condition does not change over time; that is, that no different disorder (either mental or physical) appears and that the patient continues to be symptomatic.
5. Family study reveals that the same or similar conditions are present in close relatives of the patients.
This paper has been cited more than just about any other in the psychiatric literature. The misfortune is that so few psychiatric disorders have met all of the above steps. But that does not mean that we must regard most mental disorders as something other than disease, for it may well be that we have only not yet identified the critical cause and effect relationships. Those may lie buried too deeply in the wiring of the brain or in the intricacies of DNA that we have not yet untangled.
In the beginning…
The first phase (OK, some of the Robins-Guze steps can be taken in different order, but the first one seems pretty set in concrete) is to describe the condition clearly, including demographics and signs/symptoms. Let me tell you a story.
A patient I never saw
The summer after my first year in medical school, I visited a friend at his home near the mental institution where both of his parents worked. One afternoon, walking around the vast, open campus, we fell into conversation with a staff psychiatrist, who told us about his latest, interesting patient.
She was a young woman who had been admitted a few days earlier. While attending college nearby, she had suddenly become agitated, speaking rapidly, and rushing in a frenzy from one activity to another. After she impulsively sold her nearly new Corvette for $500, her friends had brought her for evaluation.
“Five hundred dollars!” exclaimed the psychiatrist. “That kind of thinking, that’s schizophrenia!”
Now, my friend and I had had just enough training in psychiatry to recognize that this young woman’s symptoms and course of illness were far more consistent with an episode of mania than with schizophrenia. We were too young and callow to challenge the diagnosis of the experienced clinician, but as we went on our way, we each expressed the fervent hope that this patient’s care would be less flawed than her assessment.
What was wrong with the assessment of that venerable clinician? Basically, that it rested on a single symptom, rather than the collection of symptoms—a syndrome—that we clinicians have come to rely on for the definition of mental disorders. That clinician regarded the poor judgment of selling a car for less than its worth as reason enough to make a diagnosis. This is what I call “seat of the pants” diagnosis—assessment based on the clinician’s intuitive appreciation for the meaning of what appear to be this disorder’s signal characteristics. That was the chaotic state in which clinicians, especially those in North America, found themselves up to the midpoint of the previous century, and just beyond.
Of course, even before that day (my anecdote dates to the mid-1960s), it was well-known that the syndrome of schizophrenia entailed many more symptoms than that one, but clinicians adhered to careful diagnostic criteria much less that is true today. A cross-national study of diagnostic practices helped reveal just how problematic the process had become, especially in our country. That study found that clinicians in America were much more likely to diagnose a group of psychotic patients as having schizophrenia than were their European counterparts. In fact, about 40% of American patients diagnosed “schizophrenia” were probably bipolar; bipolar disease was, according to one study in 1961, nine times more likely to be diagnosed in the UK as in the US. These findings spurred psychiatrists at Washington University in St. Louis (disclosure: my alma mater) to formulate the guidelines for diagnosing both schizophrenia and bipolar disorder that would in 1980 evolve into DSM-III.
Even without the laboratory confirmation of our diagnosis, the medical model offers important benefits that accrue largely to the patient. That is, it allows a more confident prediction of the future, especially what treatment is likely to be helpful, but also how likely the disorder is to remit in the absence of treatment (the natural history of the disorder), what the likely social concomitants will be, and the risk that close relatives of the patient will come down with the same or similar symptoms. I almost forgot: the medical model also gives us the opportunity to predict what treatments are unlikely to help the patient. Let’s inspect an example from my files.
For decades, the memory of the blown diagnosis of that young woman has haunted me, in part because it is by no means unique in the annals of mental health lore. Here’s another, a patient I treated over 30 years ago.
Gail Downey
"Go ahead, cut!”
Gail lay flat on her hospital bed, staring at the ceiling. Her hair was carefully washed and combed, but her expression was stiff. "I want a lobotomy. I’ll sign the papers. I can’t take this anymore.”
Gail was an attractive 34-year-old divorcee with three children. Over the course of five years she had had multiple depressive episodes but no manias or hypomanias. Her course of illness had been marked by frequent suicide attempts and hospitalizations. In her current episode, which had lasted nearly five weeks, she had felt severely depressed throughout nearly every day. She complained that she lay awake until the early hours of each night; she had no pep, interest, or appetite. She cried frequently, and she was so distracted by her emotional turmoil that her boss had reluctantly let her go. She had been prescribed at least six antidepressants, often in combination. Most of these seemed to help the depression initially, raising her mood enough that she could at least return home. She also had responded positively to each of several courses of ECT. Within a few months of each new treatment she would relapse and return to the hospital, often with a fresh set of stitches in her wrist. While on a brief pass from the present hospitalization, she had swallowed a nearly fatal overdose of chloral hydrate. (“Chloral hydrate?” I can almost hear you ask. “Whoever prescribes that anymore?” Well, someone did, probably in desperation, and Gail nearly expired as a result.)
After her parents divorced, nine-year-old Gail had been reared by her mother. Since the age of 13, Gail had been arrested three or four times for taking small items such as pantyhose or a lipstick from department stores. Each of these incidents had occurred while she was under particular stress, usually because a job or personal relationship was going sour. She always noted increasing tension before taking these items, and felt nearly explosive joy each time she left the store with her trophy in the pocket of her overcoat. As a juvenile, whenever she was caught she had been remanded to the custody of her mother; once she had paid a fine. The most recent episode had occurred just before this hospitalization. This time, the charges had been dropped because of her repeated suicide attempts.
Gail’s medical history read like a catalog of symptoms. It included urinary retention, a lump in her throat that seemed about to strangle her, chest pains, severe menstrual cramps, vomiting spells, chronic diarrhea, heart palpitations, migraine headaches (a neurologist said they were "not typical”), and even a brief episode of blindness (from which she had recovered without treatment). At the time of the divorce, Gail’s husband had confided that she had been "frigid” and often complained of pain during intercourse. Starting in her teens, she had taken medicine or consulted a physician for more than 30 of these symptoms. The doctors had never found much wrong with her physically; they had either given her tranquilizers or referred her to a succession of psychiatrists.
After several years Gail had been evicted from her apartment, and her husband had obtained custody of their three children. The only nonmedical person she ever talked to was her mother. Now she was demanding an operation that would permanently sever some of the connections within her brain.
Evaluation of Gail Downey
Gail had more than enough mood symptoms (low mood, loss of pleasure, insomnia, anorexia, suicide ideas, loss of energy, trouble thinking) to qualify her current depression as a major depressive episode. She did not abuse substances; the exclusions for general medical conditions and substance use are discussed below. Any patient who presents with severe depression should be evaluated for major depressive disorder, which is potentially life-threatening and often responds quickly to the appropriate therapy. Gail had had numerous episodes of depression, no manias or hypomanias, and no psychotic symptoms; she had also apparently recovered for at least two months between episodes. She would therefore qualify for a diagnosis of major depressive disorder, recurrent. The persistent suicide attempts would mark it as severe without psychotic features. And in that diagnosis lay the seeds of her misery, for the diagnosis caused her clinician to treat her over and over again for a mood disorder when other aspects of her history should provoke suspicion that there is some other underlying condition.
Since her teens Gail had had a variety of somatic symptoms, at least some of which (like the migraines) were atypical. She had well over eight of the DSM–IV somatization disorder symptoms, which were distributed appropriately for that diagnosis. (The DSM-5 criteria for somatic symptom disorder are so loose and permissive that people far less affected than Gail could qualify. We’ll not consider the DSM-5 criteria further here.) Among the medical and neurological disorders to consider would be multiple sclerosis, spinal cord tumors, and diseases of the heart and lungs. The fact that Gail had been unsuccessfully treated by so many physicians would reduce the likelihood that she instead had a series of general medical conditions. The vignette provides no evidence that Gail consciously feigned her symptoms for gain (malingering) or because she enjoyed being the focus of medical attention (factitious disorder).
No additional diagnosis is needed for Gail’s anorexia; any problem with maintaining body weight was not due to refusal of food, but to her lack of appetite. Her insomnia could be given a separate diagnosis had it been serious enough to warrant independent clinical evaluation; it wasn’t. Similarly, her sexual dysfunction would not be independently coded (even if the vignette gave enough specifics as to its exact nature), because it is readily explained as a symptom of somatization disorder.
Finally, Gail’s history revealed a pattern of repeated shoplifting characterized by tension and release. Nothing else in her history would lead us to believe that she had antisocial personality disorder, and she had certainly never had had a manic episode. So, she also qualifies for the diagnosis of kleptomania.
BTW, many somatization disorder patients also have panic attacks and other symptoms of a variety of anxiety disorders. As with depressive disease, there is the danger that clinicians will make an incomplete diagnosis of an anxiety disorder and ignore the underlying somatization disorder.
Gail thus had three codable diagnoses.
And, in her story, we see a pretty clear example of how a full, accurate diagnosis made all the difference in treatment. Once the diagnosis of somatization disorder was confirmed, treatment took a different tack. We gradually withdrew all her medications and focused on a form of behavior therapy that emphasized her competence. As she came to see that she could cope with her problems (unruly kids, a recalcitrant ex-husband, lack of a job), her depression lifted. With psychotherapeutic support, she came to regard herself as a capable citizen rather than a helpless victim. Within a few weeks she was released from the hospital; later, she obtained a good job and regained custody of her kids. A couple of years later, I learned that she was to be married. She never did have that lobotomy.
The Differential Diagnosis
There’s another principle that is extremely important when considering a diagnosis for any patient—the differential diagnosis. That is a list of possible diagnoses we should consider whenever we encounter a new patient (and sometimes, when we are reviewing the facts in the case of someone who is well known to us). The order in which we arrange the diagnoses we consider is also an important part of the differential diagnosis (which I’ll sometimes abbreviate as “the differential”).
When I think about a patient, I consider a lot of different possibilities. Someone with mood symptoms could have major depressive disorder, or a bipolar depression, or dysthymia, or cyclothymia, to name just a few. And one of the very first possibilities for mental patients is that there may be something causing the symptoms that isn’t what we usually consider a mental disorder. Actually, there are two such possibilities that we should consider for every patient we see: that the symptoms could be due to a physical (medical) condition, and that the symptoms could be due to the substance use (medications, alcohol, or street drugs). This is another aspect of the medical model: the use of a differential diagnosis. Here’s a pretty good example:
Clara
Fourteen-year-old Clara had been overweight for a long time, but she seemed to have been progressing normally, through both school and puberty. That is, until one evening when she suddenly began to speak rapidly and loudly, and she insisted that her mother speak with her about “crucial issues.” At first, her mood was elevated, but she became cross when her mother finally said she wanted to go to bed. Within hours, her agitation worsened so much that she required admission to a mental hospital.
Clara stood only 5 feet 2 inches and weighed 195 pounds, and she was round-faced and puffy. She had a body mass index of more than 35—well exceeding the level considered obese. Her blood pressure was consistently above 250/120. When she undressed in the hospital, the staff could see that the skin of her abdomen bore reddened stretch marks (called striae) caused by her weight gain.
Every day for the next week, Clara’s mood was elevated and she needed little sleep. Even when interrupted, she wouldn’t stop talking longer than a few moments. She kept repeating that she was Jesus; she’d discovered a solution to every problem in the universe—sin, AIDS, and the nuclear disaster in Japan. She had flight of ideas, and she admitted that her thoughts were racing. It was impossible to interrupt her for longer than a moment or two; in fact, her physicians could barely focus her attention at all. At one point, she used the commode with others present, immodest behavior that was completely uncharacteristic for her. Clara had no personal past history of depression or mania, and her family history was negative for mood disorder.
What sort of a differential diagnosis would we consider for Clara? Of course, bipolar I disorder would be in the mix, and so would a schizoaffective disorder. But first, I’d want to consider manic symptoms caused by either a medical condition or by the ingestion of a substance. The order of the differential would go something like this:
Evaluation of Clara
For the week Clara was acutely ill, she was in turns euphoric and irritable and had increased activity, both required for DSM-5 mania criterion A. (Note that, although she had several other symptoms of mania—she spoke rapidly, needed little sleep, was grandiose, even delusional in that she thought she was Jesus—a full symptom list isn’t required for the diagnosis of an induced bipolar condition.) Although from her inability to connect with other people we might infer that she was distractible, there isn’t enough detail here to diagnose delirium (D). As far as the severity of her symptoms, she suffered from all three consequences mentioned in criterion E: psychosis, hospitalization, and impaired functioning. It’s an easy conclusion that she fully meets the requirements for bipolar I disorder.
And that probably would have been her diagnosis, if her doctors hadn’t arranged the differential as above. Because in the event, Clara was found to have an abnormal serum cortisol level. An endocrinologist recommended an MRI, which revealed a pituitary adenoma. After it was surgically removed, she no longer required psychotropic medications. She became euthymic and returned to school.
Of course, after a successful operation that yields the desired outcome, it’s pretty easy to attribute mood symptoms to a tumor. The trick is to make the connection before the elapse of many months or years. Clara’s tender age at onset, her appearance (typical “moon” face, marked overweight, classical abdominal striations) were diagnostic giveaways. Other patients have been less fortunate.
In Summary…
That’s the essence of my argument in favor of a medical model for the evaluation of mental health disorders. If you have questions or comments, you can email me at the address below.
I suspect that the problem lies partly in the assumption that anything medical necessarily involves medicines — that is, the giving of pills, potions, nostrums, and other trappings of the medical profession. And to be fair, that has been the way the psychiatric profession has been trending over the past several decades, towards the increasing use of psychotropic drugs (and ECT, and other physical therapies) in the pursuit of mental health, away from the provision of psychotherapy.
But that is not what "medical model" meant to me as a young clinician, and perhaps to most clinicians even today who were reared in the tradition. For us, the medical model means this: that the treatment of our patients depends on the use of a scientific approach of observation of the natural history of mental disorders and systematic, controlled follow-up study to see what works. The medical model, therefore, is a method of clinical investigation.
Here’s another source of confusion. For many clinicians, it is implicit in the phrase medical model that there must be a medical causation of the patient’s ills, and that in turn requires that we must be able to demonstrate an anatomical or physiological basis for the disorder. The analogy seems clear: tuberculosis can be diagnosed from a set of symptoms (coughing up blood, weakness, shortness of breath, chronic wasting) that are caused by infection from a tiny organism called Mycobacterium tuberculosis. The cause-effect relationship could not be clearer, yet it wasn’t established until 1882. And before then, there had been some pretty interesting theories as to the cause of TB (or consumption, as it was often called then).(One had to do with vampires that gradually sucked away the life forces of the relatives of a patient who had died of the disease.
For clinicians who insist that a physical or physiological cause be identified before we can refer to a psychiatric condition as a disease, I would point out that scientifically, we mental health clinicians are still at the stage of pulmonary physicians before the discovery of M. tuberculosis. But even then, the absence of an identifiable pathological process didn’t mean that one didn’t exist; it only meant that, at that stage of the science, we had not yet found it. It’s the same place we occupied 150 years ago when GPI was the scourge of mental patients.
GPI, or general paresis of the insane (also called paralytic dementia), was first identified as a separate disorder during the first half of the Nineteenth Century. What is it? Or, more precisely, was it, because it is so seldom encountered anywhere in the developed world today. GPI affected younger patients, and most often struck men. The initial symptoms of GPI were often weakness, headache, and insomnia—it was mistaken in its early stages for what used to be called neurasthenia. As the disease progressed (we’re talking about a period of months or years here), the patient’s judgment became impaired; trouble with concentration and memory led to eventual complete dementia. On the road to that end, such symptoms as depression, mania, delusions (perhaps of power or great wealth) led to treatment by psychiatrists in asylums. Eventually, the patient became paralyzed, unable to speak or move beyond the bedstead. Up to 12% of admissions to an asylum for men in New York were attributed to GPI.
In the early days, weakness of character was one favored explanation for GPI. But gradually throughout the late 1800s, the idea took hold that it was the downstream consequence of syphilis. The cause wasn’t nailed down until 1913, when two researchers demonstrated the spirochete Treponema pallidum in the brains of people who had died of GPI.
So, in our ancient history we have an example of a psychiatric illness that was eventually proven to be a disease by anyone’s standards. But, where are we now? Well, we aren’t left with a lot of other examples as neat as that of GPI. What we in the mental health professions do have now is a way forward.
That way is the essence of the medical model in psychiatry. That is, by identifying symptoms and signs and other features that a group of patients holds in common, we can separate out groups of patients who may have other features in common — say, outcome, response to treatment, and, eventually, etiology.
That approach was formalized in a seminal 1972 paper by Samuel B Guze and Eli Robins, in which they identified 5 steps in the characterization of disease.
1. Clearly describe the condition, including signs and symptoms as well as demographic features.
2. Laboratory studies to identify reliable features: imaging, chemistry, anatomical, and psychological testing are all included.
3. Delimitation from other disorders through careful inclusion/exclusion criteria.
4. Follow-up studies to determine that the condition does not change over time; that is, that no different disorder (either mental or physical) appears and that the patient continues to be symptomatic.
5. Family study reveals that the same or similar conditions are present in close relatives of the patients.
This paper has been cited more than just about any other in the psychiatric literature. The misfortune is that so few psychiatric disorders have met all of the above steps. But that does not mean that we must regard most mental disorders as something other than disease, for it may well be that we have only not yet identified the critical cause and effect relationships. Those may lie buried too deeply in the wiring of the brain or in the intricacies of DNA that we have not yet untangled.
In the beginning…
The first phase (OK, some of the Robins-Guze steps can be taken in different order, but the first one seems pretty set in concrete) is to describe the condition clearly, including demographics and signs/symptoms. Let me tell you a story.
A patient I never saw
The summer after my first year in medical school, I visited a friend at his home near the mental institution where both of his parents worked. One afternoon, walking around the vast, open campus, we fell into conversation with a staff psychiatrist, who told us about his latest, interesting patient.
She was a young woman who had been admitted a few days earlier. While attending college nearby, she had suddenly become agitated, speaking rapidly, and rushing in a frenzy from one activity to another. After she impulsively sold her nearly new Corvette for $500, her friends had brought her for evaluation.
“Five hundred dollars!” exclaimed the psychiatrist. “That kind of thinking, that’s schizophrenia!”
Now, my friend and I had had just enough training in psychiatry to recognize that this young woman’s symptoms and course of illness were far more consistent with an episode of mania than with schizophrenia. We were too young and callow to challenge the diagnosis of the experienced clinician, but as we went on our way, we each expressed the fervent hope that this patient’s care would be less flawed than her assessment.
What was wrong with the assessment of that venerable clinician? Basically, that it rested on a single symptom, rather than the collection of symptoms—a syndrome—that we clinicians have come to rely on for the definition of mental disorders. That clinician regarded the poor judgment of selling a car for less than its worth as reason enough to make a diagnosis. This is what I call “seat of the pants” diagnosis—assessment based on the clinician’s intuitive appreciation for the meaning of what appear to be this disorder’s signal characteristics. That was the chaotic state in which clinicians, especially those in North America, found themselves up to the midpoint of the previous century, and just beyond.
Of course, even before that day (my anecdote dates to the mid-1960s), it was well-known that the syndrome of schizophrenia entailed many more symptoms than that one, but clinicians adhered to careful diagnostic criteria much less that is true today. A cross-national study of diagnostic practices helped reveal just how problematic the process had become, especially in our country. That study found that clinicians in America were much more likely to diagnose a group of psychotic patients as having schizophrenia than were their European counterparts. In fact, about 40% of American patients diagnosed “schizophrenia” were probably bipolar; bipolar disease was, according to one study in 1961, nine times more likely to be diagnosed in the UK as in the US. These findings spurred psychiatrists at Washington University in St. Louis (disclosure: my alma mater) to formulate the guidelines for diagnosing both schizophrenia and bipolar disorder that would in 1980 evolve into DSM-III.
Even without the laboratory confirmation of our diagnosis, the medical model offers important benefits that accrue largely to the patient. That is, it allows a more confident prediction of the future, especially what treatment is likely to be helpful, but also how likely the disorder is to remit in the absence of treatment (the natural history of the disorder), what the likely social concomitants will be, and the risk that close relatives of the patient will come down with the same or similar symptoms. I almost forgot: the medical model also gives us the opportunity to predict what treatments are unlikely to help the patient. Let’s inspect an example from my files.
For decades, the memory of the blown diagnosis of that young woman has haunted me, in part because it is by no means unique in the annals of mental health lore. Here’s another, a patient I treated over 30 years ago.
Gail Downey
"Go ahead, cut!”
Gail lay flat on her hospital bed, staring at the ceiling. Her hair was carefully washed and combed, but her expression was stiff. "I want a lobotomy. I’ll sign the papers. I can’t take this anymore.”
Gail was an attractive 34-year-old divorcee with three children. Over the course of five years she had had multiple depressive episodes but no manias or hypomanias. Her course of illness had been marked by frequent suicide attempts and hospitalizations. In her current episode, which had lasted nearly five weeks, she had felt severely depressed throughout nearly every day. She complained that she lay awake until the early hours of each night; she had no pep, interest, or appetite. She cried frequently, and she was so distracted by her emotional turmoil that her boss had reluctantly let her go. She had been prescribed at least six antidepressants, often in combination. Most of these seemed to help the depression initially, raising her mood enough that she could at least return home. She also had responded positively to each of several courses of ECT. Within a few months of each new treatment she would relapse and return to the hospital, often with a fresh set of stitches in her wrist. While on a brief pass from the present hospitalization, she had swallowed a nearly fatal overdose of chloral hydrate. (“Chloral hydrate?” I can almost hear you ask. “Whoever prescribes that anymore?” Well, someone did, probably in desperation, and Gail nearly expired as a result.)
After her parents divorced, nine-year-old Gail had been reared by her mother. Since the age of 13, Gail had been arrested three or four times for taking small items such as pantyhose or a lipstick from department stores. Each of these incidents had occurred while she was under particular stress, usually because a job or personal relationship was going sour. She always noted increasing tension before taking these items, and felt nearly explosive joy each time she left the store with her trophy in the pocket of her overcoat. As a juvenile, whenever she was caught she had been remanded to the custody of her mother; once she had paid a fine. The most recent episode had occurred just before this hospitalization. This time, the charges had been dropped because of her repeated suicide attempts.
Gail’s medical history read like a catalog of symptoms. It included urinary retention, a lump in her throat that seemed about to strangle her, chest pains, severe menstrual cramps, vomiting spells, chronic diarrhea, heart palpitations, migraine headaches (a neurologist said they were "not typical”), and even a brief episode of blindness (from which she had recovered without treatment). At the time of the divorce, Gail’s husband had confided that she had been "frigid” and often complained of pain during intercourse. Starting in her teens, she had taken medicine or consulted a physician for more than 30 of these symptoms. The doctors had never found much wrong with her physically; they had either given her tranquilizers or referred her to a succession of psychiatrists.
After several years Gail had been evicted from her apartment, and her husband had obtained custody of their three children. The only nonmedical person she ever talked to was her mother. Now she was demanding an operation that would permanently sever some of the connections within her brain.
Evaluation of Gail Downey
Gail had more than enough mood symptoms (low mood, loss of pleasure, insomnia, anorexia, suicide ideas, loss of energy, trouble thinking) to qualify her current depression as a major depressive episode. She did not abuse substances; the exclusions for general medical conditions and substance use are discussed below. Any patient who presents with severe depression should be evaluated for major depressive disorder, which is potentially life-threatening and often responds quickly to the appropriate therapy. Gail had had numerous episodes of depression, no manias or hypomanias, and no psychotic symptoms; she had also apparently recovered for at least two months between episodes. She would therefore qualify for a diagnosis of major depressive disorder, recurrent. The persistent suicide attempts would mark it as severe without psychotic features. And in that diagnosis lay the seeds of her misery, for the diagnosis caused her clinician to treat her over and over again for a mood disorder when other aspects of her history should provoke suspicion that there is some other underlying condition.
Since her teens Gail had had a variety of somatic symptoms, at least some of which (like the migraines) were atypical. She had well over eight of the DSM–IV somatization disorder symptoms, which were distributed appropriately for that diagnosis. (The DSM-5 criteria for somatic symptom disorder are so loose and permissive that people far less affected than Gail could qualify. We’ll not consider the DSM-5 criteria further here.) Among the medical and neurological disorders to consider would be multiple sclerosis, spinal cord tumors, and diseases of the heart and lungs. The fact that Gail had been unsuccessfully treated by so many physicians would reduce the likelihood that she instead had a series of general medical conditions. The vignette provides no evidence that Gail consciously feigned her symptoms for gain (malingering) or because she enjoyed being the focus of medical attention (factitious disorder).
No additional diagnosis is needed for Gail’s anorexia; any problem with maintaining body weight was not due to refusal of food, but to her lack of appetite. Her insomnia could be given a separate diagnosis had it been serious enough to warrant independent clinical evaluation; it wasn’t. Similarly, her sexual dysfunction would not be independently coded (even if the vignette gave enough specifics as to its exact nature), because it is readily explained as a symptom of somatization disorder.
Finally, Gail’s history revealed a pattern of repeated shoplifting characterized by tension and release. Nothing else in her history would lead us to believe that she had antisocial personality disorder, and she had certainly never had had a manic episode. So, she also qualifies for the diagnosis of kleptomania.
BTW, many somatization disorder patients also have panic attacks and other symptoms of a variety of anxiety disorders. As with depressive disease, there is the danger that clinicians will make an incomplete diagnosis of an anxiety disorder and ignore the underlying somatization disorder.
Gail thus had three codable diagnoses.
- Somatization disorder, severe
- Major depressive disorder, recurrent, severe without psychotic features
- Kleptomania
And, in her story, we see a pretty clear example of how a full, accurate diagnosis made all the difference in treatment. Once the diagnosis of somatization disorder was confirmed, treatment took a different tack. We gradually withdrew all her medications and focused on a form of behavior therapy that emphasized her competence. As she came to see that she could cope with her problems (unruly kids, a recalcitrant ex-husband, lack of a job), her depression lifted. With psychotherapeutic support, she came to regard herself as a capable citizen rather than a helpless victim. Within a few weeks she was released from the hospital; later, she obtained a good job and regained custody of her kids. A couple of years later, I learned that she was to be married. She never did have that lobotomy.
The Differential Diagnosis
There’s another principle that is extremely important when considering a diagnosis for any patient—the differential diagnosis. That is a list of possible diagnoses we should consider whenever we encounter a new patient (and sometimes, when we are reviewing the facts in the case of someone who is well known to us). The order in which we arrange the diagnoses we consider is also an important part of the differential diagnosis (which I’ll sometimes abbreviate as “the differential”).
When I think about a patient, I consider a lot of different possibilities. Someone with mood symptoms could have major depressive disorder, or a bipolar depression, or dysthymia, or cyclothymia, to name just a few. And one of the very first possibilities for mental patients is that there may be something causing the symptoms that isn’t what we usually consider a mental disorder. Actually, there are two such possibilities that we should consider for every patient we see: that the symptoms could be due to a physical (medical) condition, and that the symptoms could be due to the substance use (medications, alcohol, or street drugs). This is another aspect of the medical model: the use of a differential diagnosis. Here’s a pretty good example:
Clara
Fourteen-year-old Clara had been overweight for a long time, but she seemed to have been progressing normally, through both school and puberty. That is, until one evening when she suddenly began to speak rapidly and loudly, and she insisted that her mother speak with her about “crucial issues.” At first, her mood was elevated, but she became cross when her mother finally said she wanted to go to bed. Within hours, her agitation worsened so much that she required admission to a mental hospital.
Clara stood only 5 feet 2 inches and weighed 195 pounds, and she was round-faced and puffy. She had a body mass index of more than 35—well exceeding the level considered obese. Her blood pressure was consistently above 250/120. When she undressed in the hospital, the staff could see that the skin of her abdomen bore reddened stretch marks (called striae) caused by her weight gain.
Every day for the next week, Clara’s mood was elevated and she needed little sleep. Even when interrupted, she wouldn’t stop talking longer than a few moments. She kept repeating that she was Jesus; she’d discovered a solution to every problem in the universe—sin, AIDS, and the nuclear disaster in Japan. She had flight of ideas, and she admitted that her thoughts were racing. It was impossible to interrupt her for longer than a moment or two; in fact, her physicians could barely focus her attention at all. At one point, she used the commode with others present, immodest behavior that was completely uncharacteristic for her. Clara had no personal past history of depression or mania, and her family history was negative for mood disorder.
What sort of a differential diagnosis would we consider for Clara? Of course, bipolar I disorder would be in the mix, and so would a schizoaffective disorder. But first, I’d want to consider manic symptoms caused by either a medical condition or by the ingestion of a substance. The order of the differential would go something like this:
- Manic symptoms due to physical disease
- Manic symptoms due to substance use
- Bipolar I disorder, manic
- Schizoaffective disorder, bipolar type
Evaluation of Clara
For the week Clara was acutely ill, she was in turns euphoric and irritable and had increased activity, both required for DSM-5 mania criterion A. (Note that, although she had several other symptoms of mania—she spoke rapidly, needed little sleep, was grandiose, even delusional in that she thought she was Jesus—a full symptom list isn’t required for the diagnosis of an induced bipolar condition.) Although from her inability to connect with other people we might infer that she was distractible, there isn’t enough detail here to diagnose delirium (D). As far as the severity of her symptoms, she suffered from all three consequences mentioned in criterion E: psychosis, hospitalization, and impaired functioning. It’s an easy conclusion that she fully meets the requirements for bipolar I disorder.
And that probably would have been her diagnosis, if her doctors hadn’t arranged the differential as above. Because in the event, Clara was found to have an abnormal serum cortisol level. An endocrinologist recommended an MRI, which revealed a pituitary adenoma. After it was surgically removed, she no longer required psychotropic medications. She became euthymic and returned to school.
Of course, after a successful operation that yields the desired outcome, it’s pretty easy to attribute mood symptoms to a tumor. The trick is to make the connection before the elapse of many months or years. Clara’s tender age at onset, her appearance (typical “moon” face, marked overweight, classical abdominal striations) were diagnostic giveaways. Other patients have been less fortunate.
- Pituitary adenoma
- Cushing’s syndrome
- Bipolar disorder due to Cushing’s syndrome, with manic-like episode
In Summary…
That’s the essence of my argument in favor of a medical model for the evaluation of mental health disorders. If you have questions or comments, you can email me at the address below.