2012 in Review

The article by Barry et al., “Risk Adjustment in Health Insurance Exchanges for Individuals with Mental Illness” (1), is not light reading for most of our psychiatrist readers. We published this projection of the effects of the Affordable Care Act from a leading group of health care economists because it illustrates the kinds of market decisions that will influence the financing of mental health care in the newly forming state insurance exchanges in the near future. In many states, psychiatrists in their APA district branches may need to advocate for the financing of care, taking into consideration the techniques modeled in this article, specifically reinsurance and risk adjustment to protect the financial viability of insurance plans that serve mentally ill persons. Garfield and Druss (2) point out in an accompanying editorial that these techniques are introduced by the Affordable Care Act to attempt to combat the tendency of plans to restrict mental health care benefits. Plans have a natural tendency to restrict benefits to prevent selection of their insurance plan by the mentally ill patients who need more coverage. The article by Barry et al. illustrates how small differences in these risk adjustment techniques may have profound effects for insurance companies and hence effects on the patients whom they insure.

Much of the clinical practice of psychiatry is directed at target symptoms to be suppressed, especially in the treatment of schizophrenia and related disorders. Although our patients benefit when such interventions are effective, our treatments can be considered truly effective only when they actually make a difference in how our patients experience life in the real world. For this reason, my choice for 2012 is the study by Bowie and colleagues (3) that evaluated real-world behavior in individuals with schizophrenia following the combination of cognitive remediation and functional skills training. Compared with either intervention alone, the combination resulted in improvements in community or household activities and work skills. Although the extent to which these effects will endure long-term and hold true in a broader sample of individuals remains to be determined, as few as three individuals need to be treated in order to see meaningful, real-world improvement in one. Hopefully, these findings and the results of similar studies by others portend a day when, if we paraphrase the classic American political question and ask, “Are you better off today than before you started treatment?” our patients with schizophrenia can answer “Yes!”

Thoma and colleagues (4) studied 120 randomized controlled trials of cognitive-behavioral therapy (CBT) for depression. As they, and we, have come to expect, these trials demonstrated the effectiveness of CBT compared with nontreatment conditions, along with the absence of a significant difference when CBT is compared with other active treatments. So far, nothing new or particularly interesting.

However, they went on to examine the methodological quality of each of these trials. Here an interesting pattern did emerge. First, contrary to their expectations, the quality of these CBT trials did not differ from the quality of a series of dynamic psychotherapy trials reported last year. Second, the quality wasn’t very good. For example, more than half of the studies failed to employ blind raters for outcome, failed to analyze the results in terms of intent to treat, or failed to use appropriate statistical consideration of therapist and site effects.

Third, the quality improved steadily over time, for both CBT and dynamic psychotherapy studies. Eight studies reported since 2005 were better than any study reported in the previous 30 years. This suggests that the research system is working as it should, and generating increasingly reliable and valid data in the process.

Their final finding did not surprise them, but it should trouble them and us as well. They predicted that lower methodologic quality of studies would be correlated with larger effect size and greater variability, and it was.

They speculated that the variability could result from “less tightly controlled experimental conditions,” allowing “more room for a variety of experimenter biases.” Perhaps, although tighter controls and reduced variability might have reduced “noise” and therefore allowed a stronger signal to emerge. However, their results are clear. The association of lower quality with greater effect size could mean that the reported effects result from a mixture of true effects and a kind of projective test of the researchers. As they point out, it also might reflect a publication bias, with lower-quality studies with smaller effect sizes never making it into print. In any event, it suggests that the true effects of psychotherapy (and other therapies) may be smaller than many current estimates indicate.

This is an important study, with results that are relevant for psychiatric and other therapy research in general as well as psychotherapy research in particular.

For years, scientists have viewed research on brain development as essential to discovering the causes of mental illness. However, clinical implications of such research have emerged slowly because of the difficulty of safely assessing the developing child’s brain. My favorite article in this year’s Journal, by Wolff et al. (5), might begin to reverse this trend.

Wolff and colleagues repeatedly assessed 92 infants at risk for autism by using diffusion tensor imaging, which indexes the integrity of the brain’s white matter tracts. The investigators then used these data to compare brain development in the 28 infants who ultimately met criteria for autism with the 64 who did not. Findings suggest that signs of autism manifest in the brain before clinical symptoms are readily apparent. Moreover, these signs continue to change throughout infancy, raising hopes of interrupting this evolving developmental cascade and possibly even preventing the onset of autism. Not only do such findings generate tremendous scientific interest, but they allow clinicians and neuroscientists to join in their search for signs of abnormal infant development. This joint search provides a nidus around which attempts to develop novel treatment might converge.

The Treatment in Psychiatry feature titled “The Changing ‘Weightscape’ of Bulimia Nervosa” by Bulik and colleagues (6) provides an important opportunity to examine the effect of the obesity epidemic on clinical practice. In this case-based article, Bulik et al. highlight the particularly complex influence of obesity on the treatment of bulimia nervosa, a condition where preoccupations with weight fluctuations are intrinsic to the disorder. The case involves a young woman who has a body mass index (BMI) of 38; hence comorbid obesity is a key issue in her care. Rather than the standard cognitive-behavioral therapy (CBT) that may typically be used for bulimia nervosa, she instead participated in a modified CBT program that allowed her to set personalized goals that included a physical exercise regimen. This is a superb example of reaching beyond the traditional symptom-focused approach to address important health concerns and comorbid obesity. It has been well recognized that chronically ill patients requiring antipsychotics are at great risk for obesity, yet the public health problem has now extended well beyond this subgroup. Among 161 outpatients with bipolar disorder, Fiedorowicz et al. reported the mean BMI to be 30.8, i.e., obese (7). In a population sample, obesity has also been associated with an approximately 25% increase in the odds of mood and anxiety disorders (8). The general rate of obesity among young adults is presently 35.7%, and the rate of childhood obesity has tripled since 1980 (9). Consequently, a compelling case may be made for psychiatric care to uniformly address exercise and dietary goals to achieve optimal outcomes for our patients, the youngest of whom are at greatest risk for deleterious consequences of excess weight over their lifetimes.

Pregnancy is a time when no one takes chances and no one takes medicines that are not essential. Therefore, the question of whether and when, if at all, to take antidepressant medications during pregnancy has loomed large as the proportion of childbearing-aged women taking antidepressants has increased. This is why the article by Nulman et al. (10) is so contributory. These scientists asked, What is the effect of antidepressant drugs and/or maternal depression on long-term child neurodevelopment? They studied the question by doing an analysis of children in Canada (between the ages of 3 years and 6 years, 11 months) whose mothers took antidepressants during pregnancy (venlafaxine or SSRIs), had untreated depression during pregnancy, or were healthy with no treatment. The children exposed to maternal depression had lower IQs and more problem behaviors than the healthy group, but these outcomes correlated with the extent of the maternal depression, not with the dose or duration of the antidepressant medication. While an untoward outcome was related to depression during pregnancy, it was related not to the antidepressant medication but to the severity of the depression itself. Treatment of depression really matters, and children can be adversely affected—but by the disease, not the antidepressant.

It has been a successful year for The Residents’ Journal. With so many outstanding pieces submitted, the task of selecting my “Editor’s choice” was not an easy one. In “‘Bath Salts’: Emergence of an Epidemic” (11), Loeffler introduces readers to bath salts, a dangerous group of structurally similar designer drugs that are undetectable in standard urine drug screens and until recently have been legally sold. Bath salts are synthetic cathinones, compounds that bind to norepinephrine, dopamine, and serotonin monoamine transporters. Loeffler describes two classes of bath salts: 1) mephedrone (4-methylmethcathinone), a whitish-yellow powder that produces a high that may consist of euphoria, intensified sensations, “feelings of closeness, sociability, and talkativeness,” and 2) MDPV, which produces a high that lasts nearly twice as long as that of mephedrone and also produces feelings of increased stimulation and euphoria. Mephedrone has been linked with a broad range of adverse effects, from psychosis to death. MDPV has been associated with severe panic attacks, psychosis (including paranoia), and violent behaviors. The treatment recommendations for bath salt intoxication are limited to primarily the use of benzodiazepines, intravenous fluids, and supportive measures. Loeffler’s comprehensive overview stands out for its timeliness, relevance, and broaching of an important subject that was novel to many of our readers. Loeffler equips trainees with the crucial information they need in order to stay abreast of this important matter.