advertisement-vertical Download Proto magazine app
Social Icons

Second Opinion

Proto readers assess the effects of antidepressants and the transitioning of patients from hospital bed to home.

Summer 2012


As Proto’s article on antidepressants (“Somber Questions,” Spring 2012) relates, clinical psychologist Irving Kirsch analyzed every placebo-controlled clinical trial for six widely prescribed antidepressants, finding negligible benefit and thus concluding that antidepressants do not work. However, our recent study regarding fluoxetine (Prozac) and venlafaxine (Effexor) tells a different story. We analyzed data from all 41 randomized controlled trials of the two drugs in children, adults and the elderly through six weeks of treatment. Our results revealed that for youth, the rate of response (meaning a 50% reduction in depression severity) to fluoxetine was more than five times higher than response to placebo, and the rate of remission was almost three times higher. In adults, response and remission rates with fluoxetine were 64% and 52% higher, respectively. Similar results were obtained for venlafaxine. In the elderly, more modest effects were observed. Furthermore, we found no evidence that a beneficial response was restricted to the severely ill; patients with mild and moderate depression benefited too.

The logical fallacy of Kirsch’s conclusion is that a small average difference implies that there is no benefit to individual patients. In fact, relatively small average differences can and do lead to clinically important differences in rates of response and remission.

Robert D. Gibbons // Professor of Biostatistics, Departments of Medicine, Health Studies and Psychiatry, University of Chicago
J. John Mann // Departments of Molecular Imaging and Neuropathology, New York State Psychiatric Institute; Department of Psychiatry, Columbia University

In “Somber Questions,” psychiatrist Andrew Nierenberg asserts: “People’s lives are at stake when they become needlessly frightened of antidepressants and stop taking them.” Yet there are substantive reasons to be cautious about antidepressants. Some patients who discontinue antidepressants relapse into depression, whereas others do not. While this fact might partly reflect differential responsiveness to drugs, a recent meta-analysis of discontinuation studies shows that the risk of relapse after discontinuation is positively related to the degree to which the antidepressant increases prefrontal serotonin and norepinephrine. This suggests that antidepressant use itself increases the relapse risk.

Paul W. Andrews // Department of Psychology, Neuroscience & Behaviour, McMaster University, Hamilton, Ontario
J. Anderson Thomson Jr. // Counseling and Psychological Services, Institute of Law, Psychiatry and Public Policy, University of Virginia, Charlottesville


“Enter Hospitalists” (Spring 2012) raises concerns that are the focus of many a hospital committee, including effective strategies for transitioning patients out of the hospital. My organization has addressed this with a program in which a hospitalist serves as hospital liaison for our physician group’s 10 internists and their collective panel of patients, and makes rounds daily with a care coordinator. The care coordinator not only makes needed arrangements and follow-up appointments at the time of discharge but also handles post-discharge medication reconciliation and follow-up phone calls within 48 hours.

It can certainly be argued that readmission is inevitable for the sickest patients. For a large number, however, a focused care-coordination plan can make a great difference—not only to the care patients actually receive but to the care they perceive, which often is just as important to one’s sense of well-being.

Adam Smith // Director of Inpatient Care, Assistant Chief of Acute Care Services, Yale Health Center, New Haven

What’s Your Take?
Write to the editor to comment on a story—or offer suggestions for future topics.

Protomag on Facebook Protomag on Twitter