Dr. Osser, as an attending psychiatrist at a VA unit in which you evaluate many patients with PTSD, how do you typically approach establishing a valid and reliable diagnosis of PTSD in your patients? I don’t get too fancy – I get out my DSM-IV and I ask the criteria from there. I find that a very valuable way to assess symptoms in a fairly objective way.

Any tips on how to ask the questions? One of the key criteria is avoidance. To nail down the avoidance criterion, you ask them if it is uncomfortable to recall, think about or talk about the event; whether they avoid movies, television programs, places, commercials or people that remind them of the traumas that they went through.

How do you ask the patients about flashbacks? I ask them if they have times during the day when they suddenly seem to be recalling the trauma and it seems very vivid to them, as if they are whisked back in time to the trauma and are stuck in that memory.

How do you ask about nightmares? I try to establish that the details of the nightmares are connected somehow with the trauma, and that they began after the trauma. Many people have nightmares, but these may have nothing to do with trauma.

A fair amount to establish. What about the “numbing” criterion? Numbness is diminished interest in life. These patients are estranged from everybody. They don’t want to be stimulated by contact with people, and in an effort to keep themselves on an even keel, they have a restricted range of affect.

What about the actual traumatic experience? How do you ascertain that a given experience actually qualifies as a DSM-IV “traumatic event?” This is usually a judgment call. I ask myself whether the event is so severe that it really is beyond the range of typical human experiences. There are those who have had repeated experiences over multiple years, others who have had one very, very major event, and everything in between.

How do you deal with patients who appear to be feigning symptoms in order to qualify for disability payments? We do see some patients like this. The most common clue is that after we present them with the various treatment options, both psychotherapeutic and medication-related, the patient seems uninterested in trying any of them. They seem quite willing to live with their symptoms and, not coincidentally, they are applying for disability coverage. This puts the clinician in an awkward position.

How is this handled in the VA system? In the VA, the disability evaluations are done by third parties, not by the actual treaters, so that there is some relative objectivity on that issue. And we encourage the clinicians to push the patient toward trying, systematically, the various treatments. We tell the patients that if they refuse treatments this will be held against them when they apply for disability. On the other hand, some people have been going through all the treatments, clearly are not responding to them, and you feel very strongly that they are disabled.

This brings us to the issue of treatment. So many of these patients come to us with the primary complaint of anxiety and the first thing you think about is, “I do have a medication, benzodiazepines, that will relieve anxiety quickly.” What do the guidelines say about this? This is an interesting topic. I have read guidelines and have participated with some guideline committees that concluded that benzodiazepines have no efficacy for PTSD. But when you go to the literature to see just how convincing these studies are, you find that it is actually a tiny literature. There is only one paper from 1990 by Braun where they compared 10 patients on alprazolam and 10 patients on placebo on the primary symptoms of PTSD and it showed no difference between the alprazolam and the placebo (Braun P et al., J Clin Psychiatry1990;51:236-238). And that is it for controlled studies addressing this issue: a ten-patient drug vs. placebo study.

Is there any other type of evidence bearing on this issue? There was one epidemiological study of a big database of patients in an HMO, which found that 12% of PTSD patients were abusing prescribed benzo’s versus 1.6% of patients without PTSD. So it looks like PTSD patients are more likely to abuse, but this was a retrospective observational study and there may have been confounding factors.

But from your perspective there is not really compelling evidence to tell clinicians that we should not be using benzodiazepines in patients with PTSD? Right. I feel the evidence is not that compelling. On the other hand, particularly for substance abusers, I think that it is reasonable to try to treat these patients without benzo’s. In my VA program, where 95% of patients are substance abusers, we prohibit benzo’s, and yet we are able to effectively treat these patients.

What is your overall treatment approach to PTSD? First, I rule out comorbidities, especially substance abuse. Many patients in an outpatient setting are actively abusing substances at the time of their first visit. It may not be that obvious at first; it may come to light as you start to work with them, but you should have a very high index of suspicion, because that can affect your treatment. The evidence base is that if you try to treat actively-using people, no treatment works very well, regardless of the specific modality. So I strongly stress that my patients need to be substance-free for at least a week or two before I am ready to start them on any medication, and I put a strong emphasis on that throughout and try to stay on top of that. Another issue, often neglected, is that some patients with a history of PTSD are undergoing continuing trauma, for example, with an abusing significant other. And you can’t effectively medicate in the presence of ongoing trauma.

What are the most important PTSD symptoms that you medicate? Sleep is often a major issue in PTSD. So before I start any medication, I try to address that problem with something relatively benign, because if someone can start to sleep when they have been sleep-deprived for weeks, months, if not years, that can make a big difference in how they handle themselves during the day as well. The first thing I try is trazodone for sleep. Trazodone has some efficacy for the nightmares and sleep difficulties in PTSD patients, and it has a pretty benign side effect profile. If they don’t respond to, or can’t tolerate trazodone, then I will often try prazosin.

What is prazosin, and why do you like to use it? Prazosin is a blood pressure medication that the International Psychopharmacology Algorithm Project Group endorsed as a treatment for PTSD. I heard about it from enthusiastic colleagues, so I reviewed the literature. It turns out that Murray Raskin has published several studies on it, though only one was a randomized controlled study, involving 10 patients (Raskin MA et al., Am J Psychiatry2003:160:371-373). And it had an impressive effect: not only did sleep improve, but daytime PTSD symptoms substantially improved as well. Recently, there was a letter to the editor published in Clinical Psychiatry News (February 2007) from Lloyd Spencer, who reported having treated 163 consecutive outpatients with PTSD with prazosin, and only one patient didn’t respond. Obviously, there’s not much published evidence on prazosin yet, but I have been using it and I am very impressed with it; I also have all my residents using it. Aside from being effective, it is inexpensive, and some patients like it because of the lack of stigma associated with it.

How should we dose prazosin? I start with a low dose because some people get postural hypotension and even, rarely, syncope. So my starting dose is 1 mg QHS; after three days I'll increase to 2 mg QHS, and then to 4 mg QHS after 4 more days if needed. You can then make weekly increases, first to 6 mg, and then to 10 mg (4 mg at 3 p.m. and 6 mg at bedtime). I find that the typical effective dose is from 4 to 6 mg QD, and this has worked for patients with intractable sleep problems related to PTSD, people who have literally been on everything.

What are the side effects that you warn patients about? The number one is the postural hypotension, dizziness, and the rare cases of syncope: I haven’t seen it but it could happen. And I warn them to get up slowly from bed and report any dizziness that they might have. It is very well tolerated in my experience.

What if prazosin doesn’t work? Sometimes I will use a sedating tricyclic as an option, such as low-dose doxepin at 25, 50, or 75 mg QHS. But if one or two efforts to help with sleep haven’t worked, then I will focus on the symptoms of hyperarousal, and I will start the standard antidepressant treatment, which is either an SSRI or an SNRI. They probably all work equally but my first choice is usually citalopram (Celexa), because it’s cheap, it has no drug interactions, and it has a good side effect profile.