Nearly 20% of Veterans who served in Iraq are said to have to have post traumatic stress or major depression according to a study by the Rand Corporation. Everyone seems to be racing ahead of everyone else to embrace these findings lest they be thought negligent or insensitive. A word of caution, though the full report is available from Rand, no one commenting about it (including me) seems to have read it. And none of the accounts of it I have seen mentions controls.

For this study to be meaningful there would have to be at least two control groups similarly studied. One would contain age and sex-matched civilians. A second would contain age and sex-matched veterans who had not served in a war zone. Absent these controls the finding cannot be adequately interpreted.

Doubtless, this study will fill agendas that go beyond health care for veterans. But we can hope that this report was based on adequate information. If I had to bet I would place my wager on the absence of adequate controls.

But even if there were adequate controls the issue of observer bias would still be unresolved. The person doing the psychiatric evaluation will have his own feelings and prejudices about war and military service. These cannot be avoided or suppressed. There’s no way that the evaluator can evaluate without knowing the military history of the subject. Bias no matter how surreptitious will creep into a study like this one. If the study was based on a questionnaire its results are even more tenuous.

Studies like this one often excite but far less often inform.

Here’s how the Rand Corp’s website describes the study:

Data collection for this study began in April 2007 and concluded in January 2008. Specific activities included a critical review of the extant literature on the prevalence of post-traumatic stress disorder, major depression, and traumatic brain injury and their short- and long-term consequences; a population-based survey of servicemembers and veterans who served in Afghanistan or Iraq to assess health status and symptoms, as well as utilization of and barriers to care; a review of existing programs to treat servicemembers and veterans with the three conditions; focus groups with military servicemembers and their spouses; and the development of a microsimulation model to forecast the economic costs of these conditions over time.

It doesn’t sound very reassuring. It seems to be a questionnaire with all the attendant errors such surveys intrinsically contain. It also doesn’t seem to have any controls. You can buy the Rand monograph for $55.50.