The Department of Veterans Affairs announced on September 28 that, starting in 2027, its clinicians will obtain patient education and a signed consent form before prescribing an antidepressant to a veteran younger than 30. The same announcement adds signature consent before any concurrent prescription of an opioid and a benzodiazepine, standardized patient education for psychotropic medications, and documented discussions of risks, benefits, alternatives, and how to stop a medication safely. None of it applies yet. VA says it will use the rest of 2026 to train staff and update its policies.

The press release presents the change as a transparency measure. Read against VA’s own policy documents and its congressional testimony, it is also three other things: a much narrower version of a signature mandate VA argued against before the Senate in December; a requirement VA’s current consent directive does not accommodate as written, since that directive bars signature consent for treatments that fail a high-risk test; and an age threshold that does not match the one in the FDA’s antidepressant labeling. VA has not publicly explained either the shift from its December position or the choice of age 30.

What a signature changes inside VA

Every treatment at VA already requires informed consent. Under 38 C.F.R. § 17.32 and VHA Directive 1004.01(3), the governing informed-consent policy, a practitioner must discuss the indication, risks, benefits, and alternatives before any treatment, including a prescription. For most treatments, a progress note recording that discussion is the documentation.

Signature consent is a separate, heavier tier: the patient’s and practitioner’s signatures on a nationally standardized form, completed in VA’s electronic consent software and filed in the health record. The regulation requires it for treatments that meet at least one of five criteria: sedation; anesthesia or narcotic analgesia; significant discomfort; significant risk of complication or morbidity; or injection into a joint space or body cavity. Long-term opioid therapy for pain already requires it, a requirement VA folded into Directive 1004.01 when it rescinded the standalone opioid-consent directive, Directive 1005, in December 2023.

The directive then goes a step past the regulation. For treatments that meet none of the five criteria, it states that signature consent “is not permitted,” because the extra documentation burden on the patient is “ethically unjustifiable.” Practitioners must obtain a signature for an individual patient when a treatment meets the criteria for that patient, even if it does not for patients generally.

That clause is why, on the directive’s own terms, the announcement is a policy change and not a clarification. Reading the directive as written, a signature requirement for every antidepressant prescription under age 30 would need one of two things: a determination that the prescription meets the high-risk test, most plausibly the criterion for significant risk of complication or morbidity, or an amendment to the prohibition itself. VA has not said which route it will take. As of September 29, the version on VHA’s publications index was still 1004.01(3), last amended May 1, 2024, and VA had not posted an implementing directive or notice.

What VA told the Senate in December

The announcement follows a bill VA opposed. The Written Informed Consent Act, introduced as H.R. 4837 in August 2025 and as S. 3314 in December 2025, directs VA to extend Directive 1005’s signature-consent model to antipsychotics, stimulants, antidepressants, anxiolytics, and narcotics. Neither version contains an age limit. Both name a directive VA had already rescinded.

At a December 3, 2025 Senate Veterans’ Affairs Committee hearing, VA’s written testimony said the bill appeared to mandate signature consent for approximately 100 FDA-approved medications, “many of which are routinely prescribed and do not meet the high-risk threshold.” It said requiring signatures for them “would not enhance patient outcomes and could, in fact, have adverse effects,” and that the added steps could increase the risk of negative outcomes, “including suicide,” by delaying access to psychiatric and pain medications. A coalition led by the American Psychiatric Association sent the committees a letter the day before opposing the bill on similar grounds.

The September announcement lists nearly the same five drug classes as the bill, describing psychotropic medications as antipsychotics, stimulants, antidepressants, anxiolytics, and opioids. For most of that list, what it adds is standardized education and a documented discussion. The new signature requirements cover one drug class in one age band, antidepressants under 30, and one combination, an opioid with a benzodiazepine. Neither bill has moved past committee referral; the House bill’s last recorded action was referral to the Health Subcommittee on December 19, 2025. VA’s announcement does not mention the legislation or explain what changed between its December testimony and September.

The age line

The FDA’s boxed warning, carried as class labeling on antidepressants, says they increased the risk of suicidal thoughts and behaviors “in pediatric and young adult patients in short-term studies.” The labeling quantifies that risk by age in a table drawn from pooled placebo-controlled trials covering approximately 77,000 adults and more than 4,400 pediatric patients. The drug-placebo difference per 1,000 patients treated was 14 additional patients with suicidal thoughts or behaviors under age 18 and 5 additional patients at ages 18 to 24. At ages 25 to 64 it was 1 fewer patient, and at 65 and older 6 fewer. The labeling says it is unknown whether the risk in younger patients extends beyond about four months of use. The warning and the table remain in current antidepressant labeling.

The FDA also looked past 24. In its adult meta-analysis, published in the BMJ in 2009 by staff of its Center for Drug Evaluation and Research and covering 372 placebo-controlled trials and 99,231 adults, the authors pre-specified that if they confirmed increased risk under age 25 they would test the next 10-year band, 25 to 34. They did. Among adults treated for psychiatric disorders, the population in which the age analysis was run, the odds ratio for suicidal ideation or worse at ages 25 to 34 was 0.76 (95% CI 0.53 to 1.08), and for suicidal behavior 0.81 (0.43 to 1.52). Neither estimate shows an increase, and both intervals are wide enough that a small increase cannot be excluded. Modeled continuously in the same population, the odds ratio for ideation or worse fell 2.6 percent per year of age. The analysis describes a gradient, not a cliff at 25.

Veterans aged 25 through 29 fall inside VA’s new rule, outside the FDA’s labeled elevated-risk band, and inside the band FDA tested next without finding an increase. That does not make the VA threshold wrong. It means the threshold needs a reason the FDA record does not supply.

VA has not given one. The age appears in the press release without explanation. It does not appear in the December testimony or in either bill, and VA has not yet published an implementing document that could explain it. Whether it rests on veteran-specific data, an internal safety review, a clinical judgment to extend the FDA band, or something else is not established.

One documented case sits in the background. VA’s Office of Inspector General told the same December hearing about a veteran in their twenties who died by suicide in 2022, about 50 days after starting mirtazapine at the Tuscaloosa VA, and found the nurse practitioner had not directly informed the patient of the boxed-warning risk. The OIG said its recommendation that the facility director create “processes to ensure that providers provide patient education about applicable boxed warnings when prescribing psychiatric medication” remained open. VA’s announcement does not cite the case, and it does not establish the case as the basis for the rule. What it documents is a breakdown in the ordinary consent discussion, the step VA’s new education and documentation requirements govern.

The opioid and benzodiazepine requirement

The second signature requirement covers a combination the FDA already warns against. Opioid labeling carries a boxed warning that concomitant use with benzodiazepines may result in profound sedation, respiratory depression, coma, and death, and directs prescribers to reserve that combination for patients for whom alternative treatment options are inadequate. VA told the Senate that its Opioid Safety Initiative, launched in 2013, had reduced concurrent opioid and benzodiazepine prescriptions by 90 percent. The new requirement adds a signature to a prescribing pattern VA reports it has already sharply reduced.

What the announcement leaves open

The press release does not define the rule’s scope, and none of the following can be inferred from it:

  • which drugs count as antidepressants for the rule, and whether it attaches to the drug class or to the indication for which the drug is prescribed
  • whether it applies only to new starts or also to veterans under 30 already taking an antidepressant in 2027, and whether a renewal, dose change, or switch to another antidepressant triggers a new signature
  • whether it covers Community Care and other non-VA clinicians who prescribe to veterans; the current directive’s definition of a VA practitioner does not settle how a 2027 rule will treat them
  • how it works in telehealth and inpatient settings. The current directive says consent requirements are the same whether care is delivered in person or by telehealth, and allows an alternative signature workflow when an in-person signature is not possible, but VA has not said how the new requirement will use those provisions
  • how it interacts with the current rule that a signed form is valid for 60 days, with a course of treatment allowed to continue beyond that without a new signature unless the plan or the patient’s condition changes significantly

What to watch

The implementing document is the first thing that will answer these questions. It will show whether VA amends Directive 1004.01’s prohibition on signature consent for treatments outside the high-risk criteria, or classifies antidepressant prescriptions under 30 as meeting them, and what the national consent form tells a 26-year-old veteran about a risk the FDA’s labeling places at 24 and younger. Whether VA explains the age threshold, and on what evidence, is the second.

On the legislative side, the question is whether the bills’ sponsors treat a narrower administrative rule as settling their proposal or keep pressing for all five drug classes at every age. VA has not said whether it will report how the requirement affects antidepressant starts or time to treatment among younger veterans, the outcome its own December testimony warned about.

VA has drawn its line five years past the FDA’s, and has not yet said why.