How the PTSD Treatment Guideline Ranks the Therapies

When the 2023 U.S. Department of Veterans Affairs and Department of Defense guideline for posttraumatic stress disorder placed only three psychotherapies at its highest tier, the reason was not new doubt about trauma care. It was a change in how the evidence was weighed. The panel used GRADE, a system that deliberately keeps two questions apart: how confident are we in what the studies show, and how firmly should that translate into advice. When each therapy was scored on its own trial record instead of borrowing credit from a category, the rankings shifted, and the guideline (summarized in Annals of Internal Medicine) is frank that a top rating has to be earned by a body of consistent, low-bias trials rather than one promising result.

Key points#

• GRADE rates the certainty of the evidence separately from the strength of the recommendation.
• The 2023 update graded each trauma-focused therapy on its own record, the way medications are judged.
• Three psychotherapies kept a strong recommendation: cognitive processing therapy (CPT), eye movement desensitization and reprocessing (EMDR), and the therapy abbreviated PE that has patients gradually confront trauma memories and reminders.
• A move to "suggest" is not a warning; it usually reflects a thinner or less consistent trial base, not evidence of failure.
• Trauma-focused psychotherapy is favored over medication for people able to take it up, with sertraline, paroxetine, and venlafaxine as the best-supported drugs.

Two questions GRADE never blends#

GRADE (Grading of Recommendations Assessment, Development and Evaluation) starts by asking how close the estimated effect is likely to be to the truth. For randomized trials that certainty begins high and is marked down for problems such as risk of bias, results that disagree across studies, indirect comparisons, imprecise numbers, and hints of publication bias. Only after that does the panel ask the second question: weighing that certainty against the balance of benefit and harm, patient values, and practicality, how strong should the recommendation be. The answers fall into tiers running from strong for, through weak or "suggest" for, to neither for nor against, and on to weak and strong against.

Keeping these questions separate has a useful consequence. A treatment can genuinely help and still receive only a weak recommendation when the evidence behind it is limited. Read correctly, "suggest" is an invitation, not a caution. It tells you the reviewers expect benefit while remaining honest about how large or dependable that benefit is.

The procedural change that moved the list#

The most important shift from the 2017 edition was methodological rather than clinical. Earlier guidelines had assessed the trauma-focused psychotherapies partly as a bloc, which let an individual therapy inherit the standing of the whole group. The 2023 panel instead reviewed each therapy against its own studies, applied GRADE more strictly, and folded in newer trials. That tighter accounting, not any change in how trauma works, is what rearranged the page.

Which psychotherapies held the top tier#

Of the seven trauma-focused psychotherapies that carried a strong recommendation in 2017, three retained it: cognitive processing therapy, eye movement desensitization and reprocessing, and the approach known as PE, in which a person repeatedly and deliberately faces trauma memories and avoided situations until their grip loosens. These three sit on the largest randomized-trial base and the steadiest pattern of results, which is why the National Center for PTSD still describes them as the most strongly backed choices.

Why other therapies slipped to "suggest"#

Several well-regarded options moved down to "suggest" or to "neither for nor against," among them brief eclectic psychotherapy, narrative therapy for trauma (NET), a specific cognitive therapy for PTSD, and a short writing-based protocol (WET). None of this means they do not work. Once each therapy stood alone, its evidence tended to rest on fewer independent trials or to yield less precise estimates. Under GRADE, that lower certainty caps how forcefully a panel can speak, even about a therapy clinicians respect and use.

The writing-based protocol is a clean example. It is brief, practical, and drawing growing interest, yet by 2023 the pooled trials were not deep enough to clear the bar for a strong recommendation. A guideline is a snapshot of the evidence on its publication date, and a modest rank now can firm up as more trials report.

Where the medications landed#

For people able to take it up, the guideline favors trauma-focused psychotherapy over medication, pointing to more durable improvement. Among drugs it recommends sertraline, paroxetine, and venlafaxine, the antidepressants with the most consistent randomized support for easing PTSD symptoms. Fluoxetine, chemically similar to two of those, fell into "neither for nor against" because its direct evidence was judged insufficient, a reminder that GRADE rates specific treatments rather than whole drug classes.

At the far end, the panel recommended against benzodiazepines and against cannabis or cannabis-derived products, reflecting weak evidence of benefit alongside real potential for harm. Those "against" ratings run on the same logic as the endorsements; the method cuts in both directions.

What "strong" actually costs#

The thread running through all of this is that the strongest label is earned rather than assumed. It generally takes several well-run randomized trials, a consistent direction of effect, low risk of bias, and estimates precise enough that reviewers trust the benefit is both real and meaningful. A guideline that saves its firmest language for that standard is being candid about uncertainty instead of pretending every listed therapy is interchangeable.

The useful lesson for you is about literacy rather than ranking anxiety. A therapy at the top of the list carries the sturdiest evidence, while one a tier below can still be a sound, evidence-based option, particularly when access or your own preference points that way. Choices like these belong with a clinician who knows the individual.