Last month a series of letters appeared in Science. They were written in response to a study by trauma researchers at Columbia University who examined the extent of long-term stress in Vietnam veterans.
When it was published last August, the Columbia article made headlines because it concluded that the psychological fallout from the Vietnam war was considerably less than previously thought. In short, it claimed that a “landmark” 1988 report–the National Vietnam Veterans Readjustment Study–had overstated the prevalence of residual post-traumatic stress disorder.
As is usually the case with letters to the editor, some of the correspondence was critical. But it was not the work of the Columbia team that the Science letters contested. Instead, they were aimed at Richard McNally, a professor of psychology at Harvard, who wrote the accompanying editorial.
His fault, in the critics’ eyes, was to dismiss the possibility that non-infantry soldiers could have been exposed to wartime trauma. In fact, he said no such thing, as his published reply made clear. And elsewhere he had plainly acknowledged that certain support roles could put soldiers in harm’s way. But McNally struck a nerve because he was perceived as underplaying the traumatogenic nature of war and, by extension, the vast potential for psycho pathology–and victim status–among returning troops.
The epistolary tension between McNally and his critics may seem like routine academic back-and-forth. But it goes deeper. The letters represent a slice of a larger conflict within traumatology. It is a discipline in turmoil, torn apart by the passions of political advocacy on one side and the principles of scientific method on the other.
This dynamic was on more colorful display just last November at the annual meeting of the International Society for Traumatic Stress Studies, the leading association of trauma specialists. At a symposium called “Controversies Surrounding the Psychological Risks of Vietnam for U.S. Veterans: Multiple Perspectives on New Evidence,” McNally walked the audience through his own analysis of the proportion of Vietnam veterans afflicted with post-traumatic stress disorder. (Because he was out of the country giving a long-planned lecture, McNally’s talk was “presented” as a DVD of him giving a lecture on his reanalysis. I watched the DVD and listened to a recording of the symposium.)
Before getting to McNally’s scientific presentation, a little background is in order. The Columbia study that appeared in Science in August 2006 was a reanalysis of data from a 1988 study mandated by Congress to determine the degree of psychological stress among Vietnam veterans. Known as the National Vietnam Veterans Readjustment Study (NVVRS), it put the prevalence of post-traumatic stress disorder at 15.2 percent of all men who had served in Vietnam. This figure was for cases at the time the study was conducted, between 1986 and 1988–well more than a decade after the subjects had come home from the war.
The Columbia reanalysis found a more plausible estimate of post-traumatic stress disorder to be 9.1 percent. McNally applauded the rigor of the Columbia reanalysis but went a step further. He believed the true estimate was even lower, more like 5 percent.
McNally’s main contention was that the Columbia team used too lenient a definition of “impairment” in assigning a diagnosis of post-traumatic stress disorder to veterans. (Impairment must be present before a diagnosis of post-traumatic stress disorder can be made.) By recalibrating the definition of impairment, McNally found the prevalence of post-traumatic stress disorder to be 5.4 percent among men who served in Vietnam. At last November’s meeting he took the audience through the numbers.
After hearing McNally’s presentation, three leading traumatologists, one of whom was a coauthor of the 1988 study, began their invited commentaries. All politely noted ambiguities in the data that may have led the Columbia team to reduce the post-traumatic stress disorder prevalence more than they believed was warranted, but none pressed the point.
Then they turned their guns on McNally. “I want to thank Rich for his anecdotes,” said the NVVRS coauthor in his introductory remarks, implying, of course, that while he himself did science, McNally merely told stories. The three commentators–one was an adviser to the 1988 study–spoke of the “spin” McNally put on his “misleading” and “immoderate” presentation. They issued impassioned pleas for “accurate” and “responsible” research, clearly implying that McNally’s was neither.
The president of the International Society for Traumatic Stress Studies was one of the three commentators. He virtually accused McNally of lying. “What I would like to do,” he told the audience, “is to swear Rich and other critics in under oath to tell the truth, the whole truth, and nothing but the truth. If that were done you would have seen an entirely different presentation, I think.”
During the commentaries there were ripples of approving laughter from the audience. At least one person, however, found the atmosphere so unsettling that he asked aloud, “Is Rich McNally the Anti-Christ?”
Ad hominem remarks aside, none of the panelists made a single mention of McNally’s methodology. This was remarkable because McNally’s reworking of the data was the centerpiece of his presentation. The proportion of veterans afflicted was the sole policy-relevant aspect of the National Vietnam Veterans Readjustment Study; it was the very reason Congress mandated the study.
Curiously, too, none of the panelists directed any criticism toward the Columbia team. In fact, they praised its work lavishly. Why assail only McNally when the Columbia analysis also resulted in a significant drop in estimated post-traumatic stress disorder?
One reason was that McNally was already in their cross-hairs for being a member of a vocal cadre of psychologists and historians of military psychiatry who insisted that the original estimate made by the NVVRS–namely, that 15 percent of troops had chronic post-traumatic stress disorder–was too high. How, they asked, could the prevalence be equal to the percentage of men assigned to combat, also 15 percent? This suggested, implausibly, that most infantrymen had developed cases of post-traumatic stress disorder that had lasted more than a decade.
The Columbia team, by comparison, expressed no particular doubts. It undertook the reanalysis because trying to resolve the controversy impressed the researchers as a scientifically important matter, especially so, they believed, in light of a new generation of soldiers returning from combat in Iraq.
The McNally affair is a set piece in the sociology of science, a backdrop against which heated reaction to unpopular inquiry exposes the troubled state of an academic enterprise. The hostility towards a colleague and the complete failure to engage the novel and data-driven assertion he has made–indeed, the only truly new finding presented during the entire panel–reveal traumatology to be a field in crisis.
This is no secret of course. The raging controversy over repressed memories of child abuse–which reached a fever pitch in the mid-1980s and ’90s–gave the field a self-inflicted black eye. That scandal ruined the lives of many patients and their families. The current tension over the National Vietnam Veterans Readjustment Study numbers, it is true, is far less sensational than “recovered memory therapy.” But the tempest swirling around the NVVRS is a sorry episode in its own right, destructive to the conduct and culture of inquiry. It shows, as McNally has put it, how vigorously “the advocacy tail can wag the scientific dog” in the world of trauma research.
Sally Satel is a psychiatrist and coauthor, with Christina Hoff Sommers, of One Nation Under Therapy. She is a resident scholar at the American Enterprise Institute.
