Therapy Under Analysis

Science 86

View in PDF

By Nikki Meredith


LIFE IS HARD, sometimes too hard to face alone. Every culture in the world recognizes this and includes a form of reprieve and comfort for its members’ off days. In our culture, one such refuge is psychotherapy, wherein a person seeks advice and help to overcome difficulties from someone unknown to him personally, someone trained to give it, usually in exchange for money.

The first experience with “the talking cure” often comes after a crisis. Let’s say your mother died six months ago, and you’ve had a headache ever since. Last weekend you could not get yourself out of bed. You’ve never thought of seeking help from a therapist before, but you begin to think you’re not getting enough from your wife, your friends.

Some people first seek counseling because they are intellectually curious, others because they think their lives might be made richer though they are not particularly troubled. Still others are driven to it when symptoms like chronic pain and insomnia crowd into their lives for reasons they can’t even pinpoint.

People hope such things won’t happen to them. Yet up to 20 percent of the U.S. population face a bout of serious depression in their lifetimes, and depression is only one of many mental illnesses that can leave people functional but in pain.

One hundred years ago Freud started the therapy business by opening his first office in Vienna. Today there are more than 160,000 professional therapists in the United States alone. Lots of customers come away satisfied, but only in the past three decades has science tried to figure out just what happened to them, or to those who weren’t helped. In the pages Science 86 presents an overview of the research into what psychotherapy can – and can’t yet – accomplish. We follow this review with a consumer’s guide to some of the most popular and highly regarded kinds of therapies, a discussion of the drug treatments that often accompany talking cures, and an insider’s critique of the therapy industry.


PSYCHOTHERAPY HAS BECOME such an established accessory to contemporary American life it’s easy to forget that it was not always so. As recently as 1957 only 13 percent of the population had sought some kind of psychological counseling in their lifetimes. That number is now almost 30 percent – or 80 million people – at a cost of over $4 billion annually.

This increase signifies a substantial change in popular attitudes toward psychotherapy, once considered the exclusive province of the very rich and the very disturbed. The new class of mental health patients has been created in part by the well-documented Sturm und Drang of social change, such as the breakdown of families, and in part by a standard of living free enough from physical hardship to accommodate a quest for emotional fulfillment. Thus, the great majority of those seeing therapists these days are not afflicted with severe and intractable mental illness but are more likely to suffer from problems associated with “normal” living.

Nonetheless, most who find their way into a therapist’s office are truly unhappy, beleaguered by depression, anxiety phobias, or some other distress from the long list besetting the human race. The variety of people and maladies for which they seek relief has spawned a profusion of new treatment techniques. There are more than 250 brands of therapy now on the market, including not only offshoots of traditional individual, group, and family therapies but a multitude of others. The labels suggest that it is possible to convert almost any activity into therapy: work therapy, jogging therapy, breathing therapy, pleasant experiences therapy, soap opera therapy, and, for those not interested in such here-and-now pursuits, past-lives therapy.

While each treatment form has its champions, the therapies considered most credible by professionals are supported by well-developed theoretical formulations and have some sort of traditional pedigree. Freudian psychoanalysis, the antecedent of them all, is the best known but, in fact, one of the least practiced. The analyst and his couch, so often parodied in the media, is in reality accessible only to those sufficiently healthy to withstand the reliving and “working through” of early and often painful stages of their lives and sufficiently wealthy to pay for four or five sessions a week for the two to 15 years that it might take.

Most conventional therapies fall into two classes: psychodynamic – the most widely dispensed form of treatment – and behavioral. Psychodynamic therapies, though much less intense than psychoanalysis, are based on the same principles and therefore delve into such things as unconscious motivation. Behavioral therapies are derived from learning theory and focus on retraining behavior. Instead of looking for hidden causes, behaviorists guide their patients in changing their everyday actions and thoughts.

In recent years, abbreviated versions of the major therapies have gained in popularity. Brief therapy ranges from a single session to as many as 20, but generally the goal is to provide support, minimize weaknesses, and reinforce psychological defenses with a minimum expenditure of time and money. (One of the first practitioners of brief therapy was Freud. Composer Gustav Mahler, suffering from impotence with his wife, was treated successfully by Freud in a single four-hour session.)

The use of drugs has also increased in recent years, winning adherents among both psychodynamic and behavioral practitioners. There is mounting evidence that some drugs are useful in treating particular kinds of depression and anxieties and especially when used in conjunction with therapy.

As would be expected, the profusion of patients and therapies has been accompanied by a growth in the supply of mental health professionals – in 1975 there were 60,000 therapists in the United States; currently there are 160,000 – and also an increase in the number of disciplines they represent. Psychiatrists, once the dominant force in the field, now must share clients with an ever growing stock of psychologists, social workers, and a mix of other practitioners such as psychiatric nurses and clergymen. In fact, the largest portion of the nation’s therapy is now practiced by non-psychiatrists, who charge significantly less for their services: the median cost for psychiatric professionals in private practice is $90 a session, psychologists take in an average of $65; and the going rate for social workers, who outnumber psychiatrists by almost two to one, is $50.

The competition for patients and for insurance reimbursement has sparked fierce territorial disputes between professional organizations, each claiming its members have unique, superior attributes. In fact, varied as their training may be, the distinctions in actual practice are not as great as the acrimony would suggest. The differences in therapists methods are determined more by their personalities and therapeutic ideologies than by their academic backgrounds. One bona fide difference is between those licensed to prescribe drugs – namely psychiatrists, who are physics – and those who can’t. Some non-medical therapists circumvent this limitation by associating with a psychiatrist who will prescribe for their patients.

The unfettered growth of the psychotherapeutic enterprise faced its first serious challenge in the late 1970s when the Carter administration and Congress considered instituting national health insurance and needed to decide what kinds of therapy to include. With Reagan’s election, the prospect of national health insurance with or without psychiatric benefits faded along with the government’s interest in regulation, but the continued rise in health costs has had its own regulatory effect.

Psychotherapy has been one of the fastest growing segments of total health costs, and therefore has been targeted for substantial cutbacks from the insurance industry. Many companies have limited expenditures by requiring practitioners to justify extended treatment to a panel of professionals and by putting arbitrary ceilings on benefits. Federal employees, for example, once had excellent Blue Cross mental health coverage that paid 80 percent of any sort of treatment up to $50,000. Now their best Blue Cross coverage pays 80 percent of only 50 visits a year.

There is an assumption that mental cutbacks are less harmful than in other medical services because there is less scientific evidence to support them. “People in science would like to believe evidence drives public policy, but it doesn’t always,” says Gerald Klerman, former administrator of the federal Alcohol, Drug Abuse, and Mental Health Administration. One example of this is the fate of prison therapy programs. Conventional wisdom has it that psychological rehabilitation efforts in some prisons were discontinued because they didn’t work. However, it appears that no treatment program for adult offenders had been fairly tested, according to the National Academy of Sciences.

While science may not drive public policy, policy seems to be driving science. For as unwelcome as mental health cutbacks and the prospect of government regulation have been to practitioners, both have encouraged research. There are signs that the enormous gap between treatment and scientific study has begun to narrow. “We now have the methodology to conduct credible psychotherapy studies,” says John Docherty, former director of research at the National Institute for Mental Health. “From a research standpoint, the field is the healthiest it’s ever been.”

The erratic history of systematic research got off to an explosive start in 1952 when British psychologist Hans J. Eysenck published a review comparing the improvement rates of a group of “untreated” neurotics with the improvement rates of groups that had been treated with psychoanalytic or eclectic (mixed treatment) psychotherapy. He reported that 72 percent of the “untreated” group improved, while only 64 percent of the group treated by eclectic therapy and 44 percent of the psychoanalytic patients got better.

At the time, psychoanalysts were quite popular and were cranking out reports of spectacular successes. Eysenck’s study, though widely criticized, served to challenge this long-enjoyed complacency and systematic research became a serious pursuit.

Despite the subsequent invalidation of Eysenck’s work, many of the points for which he was criticized – issues of bias and methodology – have continued to weaken therapy research. He was accused of selecting only studies that would prove his point and of comparing studies in which the key variables were too disparate. For example, there was no way of determining if the illnesses of the treated and untreated groups were comparable in severity. Moreover, the studies Eysenck compared did not use a uniform definition of improvement. Critics also pointed out that many of the “untreated groups” were cared for by general practitioners and thus actually received some therapy, i.e., attention, reassurance, and suggestion. (To this day, the problem of setting up a placebo-free study baffles researchers.)

The pattern established by Eysenck and his critics continued for the next few decades. Every study was followed by an attack on the findings and research methodology, which was then followed by another study with different findings which was also attacked, and so on. The result was a collection of studies, whose findings canceled each other out.

The impasse was broken in 1980, when psychologists Mary Lee Smith, Gene Glass, and Thomas Miller published the results of an analysis of 475 studies revealing that the average patient who received therapy was better off at the end of treatment than were 80 to 85 percent of comparable patients who did not receive such treatment. Having considered only studies meeting minimum standards of controlled trial research – those including a control group – the researchers concluded: “Psychotherapy benefits people of all ages as reliably as schooling educates them, medicine cures them, or business turns a profit.”

Although the Smith’s study is considered the most comprehensive and bias free ever done, it created an impasse of its own. All the therapies examined, psychodynamic or behavioral, got comparable results for the treatment of all disorders. This despite dramatic differences in philosophy and procedure.

While upsetting the proponents of various therapies, the findings confirmed the ideas of those in the field who believe that it is the general rather than the specific aspects of therapy that produce change. Psychiatrist Jerome Frank, one of the most respected spokesmen for this point of view, says that all therapies share features that are effective at treating a syndrome shared by all patients: demoralization. Regardless of their complaints, he says, patients feel helpless, unable to cope, depressed, guilty, and worthless.

The elements contained in every therapy that are effective in treating this condition, says Frank, include a special relationship in which the therapist expresses concern and engenders trust; a special setting – the therapist’s office – that is seen as sanctuary; and a conceptual framework that, in addition to providing an explanation for the patient’s behavior, offers hope that the treatment will relieve the suffering. And all therapies produce a degree of emotional arousal and an increase in patients’ awareness of alternatives.

From all indications, most practitioners are not very flexible. Researchers have found that few therapists, regardless of their treatment philosophies, vary their techniques to meet the needs of individual patients. But it may be that specific treatments are crucial. “A major theme of research now is the development of specificity,” says Docherty. “It is the increasing focus on specific factors at is leading to definitive answers.”

The National Institute of Mental Health is now funding several projects that seek to compare specific treatments for specific disorders. Although the results are not yet available, one of the most ambitious is a rigorously controlled study on the treatment of depression.  The study compares the success rates of cognitive therapy and interpersonal therapy, both of which have done quite well in preliminary trials in the treatment of depression, a malady that afflicts more than eight million adults annually. Cognitive therapy teaches patients to modify thoughts that produce feelings of unworthiness, frustration, and hopelessness. Interpersonal therapy, on the other hand, uses more traditional techniques and focuses on relationships and social functioning. The improvement rates of the two therapies will be compared to a group receiving antidepressant drugs.

Identical trials are being conducted at three research sites – University of Oklahoma, University of Pittsburgh, and George Washington University – thus getting a much larger sample of patients than is generally feasible and also providing simultaneous replication.

Researchers in this collaborative study have gone to great lengths to standardize treatment because in many earlier studies it has been almost impossible to define exactly what therapy was administered. Even among therapists within the same school there is so much variability in the methods used that it is often said there are as many therapies as there are therapists. To overcome this, professional therapists hired for the project were trained to use the same methods – a process that is now more consistent because of the recent development of treatment manuals – and then monitored by supervisors. Further, each therapy session was recorded so it would be possible to analyze the extent to which the actual therapy conformed to the program.

Measuring outcome has been another dilemma. Many studies use very different measures of treatment success from the objective, such as re-hospitalization, to the subjective, such as the patient’s sense of well-being. The collaborative study uses a much more complex and sophisticated system. Multiple tests measuring a variety of factors such as symptoms and social functioning are completed at various junctures before during, and after treatment. These evaluations are done by the patients themselves, “significant others,” therapists, and independent evaluators who are blind to the type of therapy administered.

Collaborative study researchers are also able to take advantage of a completely overhauled and more standardized diagnostic system, though new evidence suggests that other patient variables, such as personality type, may be more important than diagnosis in predicting the success of particular therapies. In preliminary research, Ann Simmons and George Murphy of Washington University in St. Louis used a test called the Rosenbaum Learned Resourcefulness Scale to measure patients’ preferred methods of coping. They have had success in using the results to identify which patients will be responsive to cognitive therapy and which will be responsive to drugs.

Some researchers, however, believe that the clinical trial method – even when exhaustive attempts are made at specification and standardization – miss crucial information about the individual characteristics of each patient as well as the nature of the therapist-patient relationship.

John Curtis and George Silberschatz are codirectors of another research project funded by NIMH designed to overcome the limitations of the clinical trial method. Their work, called process research, involves examining, in minute detail, what happens in each session during the course of a subject’s therapy. “Going into the session and examining what happens in this detail is like basic cellular research in medicine where people might spend years studying cells, the nature of cells, and the interaction of cells,” says Silberschatz.

Curtis and Silberschatz are testing the effectiveness of a form of psychodynamic therapy based on a theory called control mastery. Oversimplified, control mastery means that the patient defines goals and is taught to invalidate beliefs that stop him or her from achieving those goals.

Four years into the five-year study, Curtis and Silberschatz have found a high correlation between patient progress and the appropriateness of the therapists’ interpretations. In tracking this, however, they have also witnessed enormous fluctuations in the behavior of therapists, all of whom are highly experienced. “Therapists will be on, then they’ll be off, they’ll be up and they’ll be down,” says Silberschatz. “We’ve seen bad therapies where the patient was not doing well and then, almost by accident, the therapist suddenly got on the right course and the patient got better. Then the therapist reverted to the old pattern and the patient got worse.”

Silberschatz and Curtis are hopeful that the results of the study will help therapists become more consistent. “If we continue to show this high correlation over many, many cases,” says Silberschatz, “it has enormous implications for training.”

While none of the current research has yet resulted in major breakthroughs and there is still little hard evidence, the body of work is beginning to reveal nuggets of information that may one day lead to bigger answers. One of the questions remaining, however, is the extent to which these answers will be listened to by practitioners. In 1984, psychologists R. Bruce Sloane and Fred R. Staples wrote, “There is little evidence that any findings of any outcome study have had much influence on the practice of psychotherapy… ”

Dianna Hartley, a researcher at the University of California Medical Center in San Francisco, believes that this inattention to research is in part due to the limited applicability of what findings there have been, even when they are positive. “If a study is published which says out of a sample of 100, 50 patients got analytic therapy, 50 got behavior therapy, and 70 percent of the patients got better, that doesn’t really tell me much about the patient I’m seeing at three o’clock. The results haven’t been broken down in a way that’s useful to clinicians dealing with individual patients.”

The work will eventually alter the practice of psychotherapy, Docherty guesses. “There really has been a revolution in psychotherapeutic research,” he says, “in its methods and its power to determine clinically relevant findings. But it takes a long time for findings to penetrate practice; that’s true for medicine as well.”

In the meantime, many of the people who are treated do not get better, and one of the criticisms leveled at psychotherapists is that they continue to dispense treatment in the absence of improvement. But historically, medicine has always had the problem of caring for people who have diseases for which there is no known cure.

And as it is most often used in this country, psychotherapy is a last resort. Contrary to myths, therapy is not the first choice of individuals in trouble but usually the last, after they have tried everything else. The average time lapse between the first symptoms of alcoholism and seeking help is five years, for those with panic attacks, 12 years. People suffer with anxiety and depression anywhere from six months to two years before seeking help. It is perhaps the way in which psychotherapy represents a last chance that is its most important contribution. To borrow Jerome Frank’s concept, the very existence of therapy may play an important role in treating the demoralization of society. Perhaps we can never evaluate the symbolic role therapy plays, but the belief that we don’t have to be depressed or anxious or crazy, the hope that there is always a way out, may be invaluable. In their summary of the contribution therapy makes to people’s lives, researchers Smith, Glass and Miller conclude: “Of the levers that can move society forward, psychotherapy is only one. It may not educate so well as schools, it may not produce goods and services so well as management science; it may not cure illnesses so well as medicine; but it reaches a part of life that nothing else touches so well.”