Treatment-Resistant Anxiety: What It Means and Why Reassessment Matters
Updated: Oct 8
Author: Ukrainian Psychological Hub · Published: October 1, 2026 · Editorial Policy
Treatment-resistant anxiety is best understood as a treatment-course descriptor, not as a separate mental disorder. It refers to an anxiety disorder that has not improved enough after evidence-based treatment that was actually appropriate for that disorder and delivered adequately. The phrase becomes clinically meaningful only after a more basic set of questions has been answered: What diagnosis is being treated? Which intervention was used? Was it a first-line treatment for that diagnosis? Was the dose, intensity, duration, and treatment fidelity adequate? Was the treatment completed and adhered to? How was change measured? What symptoms and functional problems remain?
That distinction matters because severe or persistent anxiety does not automatically equal “treatment resistance.” Anxiety can be a normal response, a symptom within many conditions, a medication or substance effect, a manifestation of a medical problem, or part of a formally diagnosed anxiety disorder. The World Health Organization describes anxiety disorders as conditions involving excessive fear or worry with associated behavioral, cognitive, and physical symptoms that cause significant distress or impairment, while also emphasizing that anxiety itself can occur normally. WHO’s current anxiety-disorders overview therefore provides the right starting boundary: the treatment-resistant concept applies to a clinical disorder and a treatment history, not simply to the presence of distressing anxiety.
A major international 2024 Delphi consensus guideline created the most detailed modern operational framework for treatment-resistant anxiety disorders in adults. It distinguishes resistance to pharmacotherapy from resistance to psychotherapy, specifies what an adequate failed trial should look like, and proposes a staged model rather than treating “resistant” as a single all-or-nothing state. Those criteria were developed mainly to improve research and regulatory consistency. They are useful in clinical reasoning, but they are not a new DSM-5-TR or ICD-11 diagnosis and they are not a self-test.
For the broader clinical map, see Anxiety Disorders: Symptoms, Types, Diagnosis, and Treatment. For how a clinician establishes and revisits a diagnosis, see How Is Anxiety Diagnosed? Assessment, History, and Differential Diagnosis.
What Does “Treatment-Resistant Anxiety” Mean?
In everyday language, people may use “treatment-resistant anxiety,” “refractory anxiety,” or “anxiety that will not go away” after one therapy course, one medication, several medication changes, or years of fluctuating symptoms. Research terminology is narrower. Treatment resistance refers to persistent clinically significant illness despite one or more adequate evidence-based treatment trials. The word “adequate” carries most of the scientific weight. A treatment that was never appropriate for the diagnosis, stopped before a reasonable evaluation period, taken inconsistently, delivered without the essential therapeutic components, or abandoned because of intolerable adverse effects does not provide the same evidence as a fully delivered treatment that simply failed to produce sufficient benefit.
The older literature used strikingly inconsistent definitions. A 2019 systematic review of 62 studies found that 62.9% of definitions labeled treatment resistance after the first treatment failure, 93% required a pharmacological failure, only 29% required a psychological-treatment failure, and more than half did not clearly define what treatment failure meant. Minimum treatment duration ranged from four weeks to six months. This heterogeneity made studies difficult to compare and helped motivate the later consensus work.
The 2024 consensus is therefore best read as an effort to standardize a previously unstable construct. It does not mean every healthcare system, guideline, clinician, or research protocol now uses exactly the same threshold. It also does not mean the label predicts that future treatment will fail. The consensus itself emphasizes that resistance to pharmacotherapy does not preclude response to psychotherapy, and resistance to psychotherapy does not preclude response to medication or a combination of modalities.
Treatment Resistance Is Not a DSM-5-TR or ICD-11 Diagnosis
DSM-5-TR and ICD-11 classify anxiety disorders such as generalized anxiety disorder, panic disorder, agoraphobia, social anxiety disorder, specific phobia, separation anxiety disorder, and selective mutism. “Treatment-resistant anxiety disorder” is not a separate diagnostic entity in those classification systems. The American Psychiatric Association’s DSM-5-TR overview explains that DSM diagnoses are based on disorder-specific criteria, exclusion rules, differential diagnosis, and trained clinical judgment. The WHO ICD-11 Clinical Descriptions and Diagnostic Requirements likewise exists to support accurate identification of the underlying mental disorder.
This has an important practical consequence. Before asking whether an anxiety disorder is resistant, the clinician needs to be reasonably confident about which disorder is present. Persistent worry in generalized anxiety disorder is not assessed and treated exactly like recurrent unexpected panic attacks in panic disorder, fear of social scrutiny in social anxiety disorder, situation-bound fear in specific phobia, or separation-related fear in separation anxiety disorder. OCD and PTSD can involve intense anxiety but occupy separate diagnostic territories; illness anxiety, depressive disorders, bipolar disorders, substance-related conditions, sleep disorders, and medical conditions can also change the formulation.
A resistance label that is attached too early can hide a more useful question: is the current treatment targeting the right clinical problem? That is why a careful anxiety differential diagnosis becomes central when expected improvement has not occurred.
The 2024 Adult Consensus: What Counts as an Adequate Failed Treatment?
The 2024 international consensus separated treatment failure, pharmacological treatment resistance, psychotherapeutic treatment resistance, and multimodal resistance. Its purpose was to provide reproducible operational criteria, particularly for research. The details are useful because they show why “I tried something and I am still anxious” is not enough to establish treatment resistance.
How treatment failure is measured
For general operational purposes, the panel defined treatment failure as failure to achieve a clinically significant reduction in symptom severity. It proposed less than a 50% reduction on the Hamilton Anxiety Rating Scale or Beck Anxiety Inventory, or a Clinical Global Impression–Improvement score above 2, as broad options. It also proposed optional disorder-specific response criteria for social anxiety disorder, generalized anxiety disorder, and panic disorder/agoraphobia.
These are research and outcome-monitoring rules, not diagnostic cutoffs. A questionnaire score cannot establish an anxiety disorder, prove that a treatment was adequate, rule out another diagnosis, or determine that a person is “resistant.” The American Psychiatric Association’s guidance on DSM-5-TR assessment measures explicitly states that symptom measures should enhance clinical decision-making rather than serve as the sole basis for diagnosis.
Pharmacotherapy resistance
For pharmacological treatment resistance, the 2024 panel recommended at least two separate failed full trials of first-line pharmacological monotherapy approved for the anxiety disorder and recommended by guidelines, using two different medication classes, at at least the minimum approved dose for at least six to eight weeks each, ideally with documented adherence. The consensus gives examples for generalized anxiety disorder, but those examples are not universal prescriptions for every anxiety disorder or every country. Approval status, guideline recommendations, contraindications, adverse effects, prior response, age, pregnancy considerations, other medications, and patient preference all matter.
The consensus also says apparent pharmacological resistance should be checked for “pseudo-resistance,” including whether treatment was taken as intended and whether factors such as age and renal or hepatic function could affect exposure. Drug-level testing may sometimes be informative but was considered impractical for routine use in many settings. A medication stopped early because of adverse effects is clinically important, but “intolerable” and “ineffective despite an adequate trial” are different treatment outcomes.
Psychotherapy resistance
For psychotherapy resistance, the consensus recommended at least one failed trial of an evidence-based first-line standardized, ideally manualized psychotherapy such as CBT, delivered by a qualified clinician with adequate intensity and duration. Adequacy includes whether core procedures were actually implemented—for example, sufficient exposure or behavioral experiments when those are central to the disorder-specific treatment—and whether between-session practice and treatment engagement were adequate.
Duration cannot be reduced to one universal number. The consensus gives a range from one-session approaches for some specific phobias to substantially longer courses for generalized anxiety disorder, panic disorder/agoraphobia, and social anxiety disorder. This is exactly why the phrase “I did therapy for anxiety” is too broad for a resistance assessment. A supportive or insight-oriented therapy may have value for many purposes while still not constituting a full trial of the disorder-specific first-line intervention being used as the benchmark for treatment resistance.
A staged model rather than a single label
The consensus proposed three escalating stages. Stage 1 reflects failure of either two adequate pharmacotherapy courses or at least one adequate psychotherapy trial. Stage 2 reflects failure of both those pharmacological and psychotherapeutic benchmarks. Stage 3 reflects failure of multiple adequate pharmacological and multiple adequate psychotherapeutic trials and was termed multimodal treatment resistance. The value of this model is conceptual: two people described as having “treatment-resistant anxiety” may have very different treatment histories and very different remaining options.
No Response, Partial Response, Residual Symptoms, Relapse, and Treatment Resistance Are Different
Treatment outcomes are not binary. A person can have no meaningful response, a partial response, substantial symptom reduction without remission, remission with residual avoidance, improved symptoms but persistent functional impairment, or an initial recovery followed by recurrence. Those patterns can look similar from a distance but lead to different clinical questions.
No response
No response generally means symptoms and functioning show little clinically meaningful improvement after an adequate intervention. Before interpreting this as biological or psychological “resistance,” clinicians need to confirm treatment adequacy and the validity of the target diagnosis.
Partial response and residual symptoms
A partial response means treatment helped but not enough. Residual symptoms can include worry, anticipatory anxiety, avoidance, safety behaviors, panic-related fear, sleep disruption, or functional restrictions. Someone can improve on a symptom scale while still avoiding work, travel, social interaction, school, or other important activities. Conversely, function may improve before symptoms fully recede. A reassessment should therefore examine both symptom change and real-world functioning.
Relapse or recurrence
Relapse describes the return or substantial worsening of symptoms after meaningful improvement or remission. That is conceptually different from never responding to an adequate treatment. A person who improved and later worsened may need a relapse-focused formulation: what changed, whether treatment was discontinued, whether discontinuation symptoms are being confused with recurrence, whether stressors or health conditions changed, and whether previously helpful strategies are still in place. For the dedicated guide, see Anxiety Relapse: Why Symptoms Return and How Relapse Prevention Works.
Treatment intolerance
A treatment can also fail because it cannot be tolerated. Clinically, intolerance matters enormously, but it is not identical to evidence that the treatment was ineffective when delivered at an adequate therapeutic dose or intensity. Conflating intolerance, nonadherence, early discontinuation, and true nonresponse can make the treatment history look more resistant than it actually is.
Why Reassessment Matters Before Escalating Treatment
Reassessment is not a ceremonial repeat of the original diagnosis. It is the process of rebuilding the treatment map from the evidence already generated by the person’s course. The question changes from “What treatment works for anxiety?” to “What exactly has been treated, how well was it treated, what changed, what remains, and what could explain the mismatch between expected and observed response?”
This principle is embedded in major clinical guidance. For complex treatment-refractory generalized anxiety disorder, NICE CG113 recommends specialist assessment of symptom duration and severity, functional impairment, comorbidities, risk, current and past treatments, medication adherence, fidelity of prior psychological interventions, and the person’s home, support, and relationship context. For panic disorder with significant symptoms after two interventions, NICE recommends specialist referral and a thorough holistic reassessment that includes previous treatment effectiveness and concordance, substance use, comorbidities, functioning, social networks, chronic stressors, avoidance, and risk.
Recheck the diagnosis and the exact treatment target
The first question is whether the clinical target is sufficiently specific. “Anxiety” can refer to generalized worry, panic attacks, fear of scrutiny, phobic avoidance, separation fears, trauma-related hyperarousal, obsessive-compulsive distress, illness fears, depressive agitation, medication effects, substance withdrawal, or physical symptoms from a medical condition. Treating all of these as one interchangeable anxiety syndrome creates predictable nonresponse.
A diagnostic reassessment can identify a different primary disorder, a comorbid disorder that changes treatment planning, or a mixed picture in which several conditions require attention. For example, depression frequently co-occurs with anxiety disorders and can affect functioning, motivation, sleep, cognition, and perceived recovery. See Anxiety and Depression: Why They Commonly Co-Occur for that boundary. The same logic applies to OCD and PTSD: anxiety may be prominent, while the disorder requires its own treatment framework.
Confirm that the treatment matched the disorder
A treatment can be evidence-based for one anxiety disorder and still be a poor test of treatment resistance in another. Disorder-specific CBT is a good example. Panic-focused CBT often includes interoceptive exposure and work on catastrophic interpretations of bodily sensations. Social-anxiety CBT targets self-focused attention, safety behaviors, negative predictions, behavioral experiments, and exposure to feared social situations. Specific-phobia treatment is strongly exposure-centered. Selective mutism typically requires a developmentally and contextually adapted behavioral plan involving communication environments such as school and family.
The same boundary applies to medication evidence. A drug recommended for generalized anxiety disorder is not automatically an evidence-based first-line treatment for specific phobia or selective mutism. The 2026 systematic review of treatment-resistant anxiety interventions primarily found direct evidence in treatment-resistant GAD, panic disorder/agoraphobia, and social anxiety disorder—not across every diagnosis classified as an anxiety disorder.
Confirm dose, duration, intensity, fidelity, and adherence
An apparently failed medication trial may have been too short, inconsistently taken, limited by adverse effects, or never brought to a clinically appropriate dose. An apparently failed psychotherapy may have lacked the central therapeutic procedures, involved too few sessions, stopped before sufficient exposure or behavioral practice occurred, or been interrupted by practical barriers. None of these observations assigns blame. They identify where the treatment process and the evidence benchmark diverged.
A 2012 review of nonadherence and nonresponse found that roughly one fifth of participants in anxiety-treatment research dropped out early and about one third of treatment completers were classified as nonresponders across the meta-analyses reviewed. Predictors were inconsistent, but practical barriers, side effects, motivation, severity, and comorbidity were among the factors discussed. The useful clinical lesson is that treatment completion, treatment quality, and treatment response are separate variables.
Review avoidance and safety behavior
Persistent avoidance can make improvement look smaller than symptom reduction alone would suggest, and safety behaviors can limit opportunities for corrective learning in exposure-based treatments. The goal is not to force exposure indiscriminately; it is to determine whether the therapy actually addressed the avoidance processes that maintain the diagnosed fear problem. For the mechanism, see Anxiety and Avoidance: How Avoidance Maintains Fear.
Review comorbidity without treating it as an automatic explanation
The 2024 consensus recommends recording and considering comorbid depression, substance-use disorders, and personality disorders but does not make their presence part of the operational definition of treatment resistance. This is a useful distinction. Comorbidity can complicate presentation and treatment planning without proving why a particular treatment failed. The reassessment should ask what each condition contributes, which problems are most impairing, and whether the treatment plan is addressing the relevant combination.
Review substances, medications, sleep, and physical health
Caffeine, alcohol, cannabis and other substances, withdrawal states, prescribed or over-the-counter medications, sleep loss, endocrine or cardiopulmonary problems, and other medical conditions can produce or intensify symptoms that feel like anxiety. An older but still influential clinical review of treatment-refractory anxiety used the term pseudo-resistance for apparent nonresponse caused by factors such as inadequate treatment delivery, nonadherence, unrecognized substance effects, or an incorrect diagnosis. The modern consensus retains the need to exclude pseudo-resistance while framing it more operationally.
This is also where medical safety matters. New or changing chest pain, marked shortness of breath, fainting, persistent palpitations, neurologic symptoms, or other significant physical changes should not be attributed to anxiety simply because anxiety is already part of the history. See Anxiety Symptoms and Medical Differential Diagnosis for the symptom-first medical boundary.
Measure function as well as symptom severity
A scale can show meaningful symptom reduction while a person remains unable to work, attend school, travel independently, maintain relationships, sleep normally, or enter feared situations. The reverse can also occur: function may improve even while symptoms remain noticeable. Reassessment should therefore track the person’s own goals and functional outcomes alongside symptom measures. This prevents a purely numerical definition of success or failure.
Include preference, feasibility, and access
A theoretically optimal treatment is not an adequate real-world trial if the person cannot access it, cannot participate in its essential procedures, or chooses not to continue after an informed discussion of benefits and harms. The 2024 consensus explicitly considered patient preferences and attitudes, and major guidelines use shared decision-making when selecting among evidence-based options. Reassessment should distinguish “treatment did not work” from “treatment was not realistically deliverable in this context.”
An “Adequate Trial” Is Disorder-Specific
One of the most common errors in treatment-resistance discussions is to treat all anxiety disorders as if they share one standard trial. They do not. Anxiety disorders overlap in fear, worry, arousal, avoidance, and functional impairment, but their treatment evidence and essential therapeutic components differ.
Generalized anxiety disorder
For GAD, treatment may involve disorder-appropriate CBT, applied relaxation in some guideline frameworks, and pharmacological options supported for GAD. NICE’s GAD guideline uses a stepped-care model and reserves complex treatment-refractory GAD for specialist care after inadequate response at earlier steps. At that point, NICE recommends a formal review of previous treatment adherence and psychotherapy fidelity before considering more complex regimens, and it cautions that evidence for combination pharmacotherapy is limited while adverse effects and interactions may increase.
The dedicated Generalized Anxiety Disorder Treatment: CBT, Medication, and Self-Help page owns the detailed GAD treatment evidence. The present article does not turn a GAD-specific sequence into a universal anxiety algorithm.
Panic disorder and agoraphobia
Panic disorder treatment has its own mechanisms and evidence, including panic-focused CBT and interoceptive exposure when appropriate. Agoraphobic avoidance may require exposure to feared situations in addition to work on panic-related interpretations. NICE CG113 recommends specialist mental-health referral in most instances when two interventions have been provided and significant panic symptoms persist, followed by a holistic reassessment rather than automatic medication escalation.
For detailed disorder-specific options, see Panic Disorder Treatment: CBT, Exposure, Medication, and Self-Help.
Social anxiety disorder
Social anxiety disorder provides a clear example of why treatment sequencing is diagnosis-specific. NICE CG159 recommends individual social-anxiety-specific CBT as the preferred initial psychological treatment for adults, and then gives distinct options for partial response to CBT, partial response to an SSRI, nonresponse or intolerance to an SSRI, and later treatment steps. Its recommended CBT protocols include behavioral experiments, exposure, reduction of safety-seeking behavior, work on self-focused attention, and relapse prevention.
The disorder-specific owner is Social Anxiety Disorder Treatment: CBT, Exposure, Medication, and Self-Help.
Specific phobia
Specific phobia is a particularly important boundary case. Exposure-based psychological treatment is central, and adequate treatment can be much shorter than treatment for GAD, panic disorder, or social anxiety disorder. The 2024 consensus explicitly notes that the adequate psychotherapy time frame can range down to one-session approaches for specific phobias. Therefore, a generic rule such as “therapy must always last three months before it counts” would be wrong.
For this diagnosis, see Specific Phobia Treatment: Exposure Therapy and What Evidence Shows.
Separation anxiety disorder and selective mutism
The 2024 treatment-resistance consensus states that its proposed criteria are limited to adults and that treatment-resistance criteria for children, adolescents, and older adults remain to be established. It also specifically notes that treatment resistance in adult separation anxiety disorder and selective mutism warrants further investigation. That makes extrapolation especially risky because both diagnoses have substantial developmental considerations and much of their clinical literature concerns younger populations.
Use the diagnosis-specific evidence instead: Separation Anxiety Disorder Treatment: CBT, Family Support, and Medication and Selective Mutism Treatment: Behavioral Therapy, School Support, and Family Involvement. Adult treatment-resistance thresholds should not simply be copied onto a child or adolescent.
What Does the Evidence Say About Treatment After Nonresponse?
There is no single evidence-based “treatment-resistant anxiety protocol” that applies across diagnoses. The most recent broad synthesis is the 2026 integrative systematic review by Schiele and colleagues, which examined pharmacological, psychotherapeutic, and neurostimulatory treatments in adults with treatment-resistant generalized anxiety disorder, panic disorder/agoraphobia, and social anxiety disorder. It identified 26 randomized controlled trials and 36 open-label studies. Most studies were small and had important methodological limitations.
The review found some disorder- and sequence-specific signals. In panic disorder, randomized evidence supported certain antidepressant options after CBT nonresponse. In pharmacologically treatment-resistant anxiety disorders, several randomized studies supported CBT after medication nonresponse. For GAD and social anxiety disorder, various medication-switching or augmentation strategies have been studied, but the evidence varies substantially in rigor, sample size, and directness. Neurostimulation evidence remained inconclusive.
That pattern supports a restrained conclusion: nonresponse to one evidence-based treatment does not establish that all treatments will fail, and switching modality or combining modalities can be reasonable in selected cases. It does not support a universal list of “stronger” medications for anyone who feels anxious despite treatment. The relevant question is which diagnosis, which prior treatment, which response pattern, which adverse effects, which risks, and which next option has direct evidence for that clinical situation.
A separate 2020 meta-analysis of psychotherapy trials in treatment nonresponders across mood and anxiety disorders found that introducing psychotherapy to treatment nonresponders produced a medium-to-large benefit overall, although the evidence base was small and heterogeneous. This again argues against treating prior nonresponse as a prognosis of permanent untreatability.
Medication Changes and Augmentation Require More Caution, Not Less
Persistent symptoms can create pressure to escalate treatment quickly. Yet treatment resistance is exactly the context in which the evidence base becomes thinner and the risk-benefit calculation becomes more individualized. The 2024 consensus notes that internationally endorsed evidence-based guidelines specifically for treatment-resistant anxiety disorders are still lacking. The 2026 systematic review shows why: the literature includes many small trials, open-label studies, disorder-specific strategies, and interventions with materially different adverse-effect profiles.
NICE therefore places complex treatment-refractory GAD in specialist care and explicitly cautions that evidence for combination drug treatments is limited while adverse effects and interactions become more likely. The German anxiety-disorders guideline revision similarly recommends considering a switch to the other major modality or a combination when medication or psychotherapy has not been effective, with an individual plan based on efficacy, adverse effects, costs, and patient preference.
The practical implication is not that medication escalation is inappropriate. It is that an educational article cannot safely translate heterogeneous second-step evidence into an individualized medication recipe. A prescriber needs to review the exact diagnosis, prior drugs and doses, duration and adherence, adverse effects, other medications, substance use, medical history, pregnancy considerations when relevant, withdrawal risk, and the evidence for the specific anxiety disorder before choosing a next step.
Why “Pseudo-Resistance” Is a Useful Concept—and a Risky Label
Pseudo-resistance refers to apparent treatment resistance when the treatment history does not yet demonstrate true nonresponse to adequately delivered evidence-based care. Examples include the wrong treatment for the diagnosis, an insufficient dose or duration, inconsistent medication use, psychotherapy that omitted essential components, major barriers to participation, or an incorrect or incomplete diagnostic formulation.
The concept is useful because it prevents premature therapeutic pessimism. It is risky if it is used to blame a person for not improving. Adherence itself can be affected by adverse effects, financial constraints, transportation, work schedules, caregiving, fear of exposure exercises, ambivalence, poor therapeutic alliance, cognitive or communication needs, stigma, or misunderstanding of the treatment plan. The clinical task is to identify what happened in the treatment process and correct what can be corrected.
The most precise framing is therefore “apparent resistance pending reassessment” when treatment adequacy is uncertain. Once adequate trials are documented and the underlying diagnosis has been carefully reviewed, the treatment-resistance construct becomes more defensible.
Can the Diagnosis Be Wrong?
Yes, but persistent symptoms do not automatically prove misdiagnosis. A diagnosis can be incomplete, another condition can be comorbid, the dominant problem can change over time, or the original diagnosis can remain correct while the person genuinely does not respond to a well-delivered treatment. Reassessment is about updating probabilities, not searching for a single hidden explanation.
A thorough diagnostic review asks whether the current symptoms still meet the expected pattern for the diagnosed anxiety disorder; whether panic attacks, avoidance, intrusive thoughts, trauma symptoms, depressive symptoms, elevated or unusually activated mood, substance use, sleep disturbance, attentional problems, or physical symptoms point toward additional assessment; and whether medication or medical conditions could better account for some of the presentation. For a structured overview of mimics and overlaps, see Anxiety Differential Diagnosis: What Can Look Like an Anxiety Disorder?.
The DSM-5-TR framework itself emphasizes differential diagnosis and conditions that must be ruled out, while WHO’s ICD-11 CDDR describes accurate diagnosis as a first step toward appropriate care. Treatment nonresponse raises the value of revisiting that diagnostic work because the treatment history provides new information.
There Is No Single Test for Treatment-Resistant Anxiety
No blood test, brain scan, genetic test, questionnaire, or digital marker can currently determine that an anxiety disorder is treatment resistant in routine practice. Symptom scales can document severity and change, and the 2024 consensus proposes specific outcome thresholds for operational research definitions. Their role is measurement, not standalone diagnosis.
Likewise, a high GAD-7, Beck Anxiety Inventory, Hamilton Anxiety Rating Scale, Panic Disorder Severity Scale, Liebowitz Social Anxiety Scale, or another instrument does not prove that prior treatment was appropriate or adequate. A person can score highly because the treatment failed, because the target diagnosis was incomplete, because the treatment was never completed, because another condition is contributing, or because the scale is capturing a symptom domain that requires additional clinical interpretation.
A resistance assessment is therefore longitudinal. It combines diagnostic history, treatment history, adherence and fidelity, symptom trajectories, function, adverse effects, comorbidity, medical and substance review, and the person’s goals and preferences.
Does “Treatment-Resistant” Mean Permanent?
No. “Treatment-resistant” describes what has happened so far under defined treatment conditions; it is not a declaration that recovery is impossible. The 2024 consensus explicitly treats resistance as dimensional and modality-specific, and notes that failure of pharmacotherapy does not rule out response to psychotherapy or vice versa. The 2026 systematic review found evidence for benefit from switching modalities or using selected second-step strategies in some adults with GAD, panic disorder/agoraphobia, or social anxiety disorder.
The label can nevertheless feel definitive. That is one reason some experts have also discussed “difficult-to-treat” anxiety as a broader clinical concept. “Difficult to treat” can include complexity that does not fit a clean failed-trial algorithm: multiple comorbidities, adverse effects, fluctuating symptoms, access problems, chronic stressors, or treatments that are technically available but not feasible for a particular person. The 2024 panel considered this broader language while retaining treatment resistance as a useful operational research construct.
The strongest realistic message is therefore neither “just keep trying” nor “nothing works.” It is that persistent symptoms justify a more exact formulation, better documentation of what has already been tried, and a diagnosis-specific review of remaining evidence-based options.
Treatment-Resistant Anxiety in Children and Adolescents
The adult treatment-resistance criteria should not be copied directly to children or adolescents. The 2024 consensus explicitly states that its proposed criteria are limited to adults and that criteria for childhood and adolescence remain to be established. Development changes how fear is expressed, how impairment is judged, which treatments are evidence-based, how parents or caregivers participate, how school context affects symptoms, and how medication decisions are made.
This is especially important for separation anxiety disorder and selective mutism, both of which have major pediatric evidence bases. A child who remains highly anxious after a brief or nonspecific intervention needs a developmentally appropriate reassessment rather than an adult “resistant anxiety” label. The dedicated treatment pages linked above preserve those diagnosis- and age-specific evidence boundaries.
When Persistent Anxiety Needs Broader or More Urgent Reassessment
Treatment resistance is not an explanation for every worsening symptom. A person whose anxiety pattern changes substantially, who develops new physical symptoms, who has major medication changes, or whose substance use changes may need a fresh medical and psychiatric assessment rather than simply another anxiety intervention.
Urgency rises when there is severe functional deterioration, self-neglect, significant risk of self-harm or suicide, suspected mania or psychosis, intoxication or dangerous withdrawal, or physical symptoms that could represent an acute medical condition. NICE explicitly includes risk and self-neglect in specialist assessment for complex treatment-refractory GAD. Online symptom matching cannot safely distinguish anxiety from every medical or psychiatric emergency.
What a Good Reassessment Should Clarify
A strong reassessment should leave the person and clinician with a clearer map than they had before. It should name the primary diagnosis or diagnoses being treated, identify the symptoms and functional problems that still matter most, summarize every major treatment trial in a comparable way, distinguish nonresponse from intolerance or incomplete treatment, and specify which evidence-based options remain genuinely untried.
For medication, that means knowing the drug, indication, approximate dose range, duration, adherence, adverse effects, reason for stopping, and objective or functional response. For psychotherapy, it means knowing the treatment model, whether it was disorder-specific, who delivered it, approximate number and spacing of sessions, whether core procedures such as exposure or behavioral experiments were used when indicated, the extent of between-session practice, barriers to engagement, and what changed.
It should also review physical health, sleep, other medications and substances, comorbid mental disorders, chronic stressors, family or social context, avoidance and safety behaviors, the person’s goals, previous helpful experiences, and preferences about next steps. The result is a treatment plan that is more specific because the failures have been analyzed, not merely counted.
Questions to Discuss With a Clinician
Useful questions are concrete rather than label-centered. You can ask: What exact diagnosis are we treating now? What evidence makes that diagnosis the best fit? Did my previous treatment count as an adequate trial for this disorder? If not, what was missing—duration, dose, treatment component, adherence, or something else? Did I have no response, a partial response, residual symptoms, intolerance, or relapse?
It is also reasonable to ask: Are there comorbid conditions or medical contributors that could change the plan? Are any substances, medications, sleep problems, or withdrawal effects complicating the picture? Is the proposed next treatment supported for my specific anxiety disorder, or is the evidence being extrapolated from another condition? What benefits, adverse effects, monitoring needs, and alternatives should I understand? Would specialist assessment add anything at this stage?
Those questions support shared decision-making without reducing treatment to a sequence of trial-and-error prescriptions.
Frequently Asked Questions
Is treatment-resistant anxiety a diagnosis?
No. It is a clinical and research descriptor for an anxiety disorder that has not responded sufficiently to adequately delivered evidence-based treatment. DSM-5-TR and ICD-11 diagnose the underlying anxiety disorder, not a separate disorder called “treatment-resistant anxiety.”
How many treatments have to fail before anxiety is called treatment resistant?
There is no single universal clinical rule. The 2024 adult consensus defines pharmacotherapy resistance as failure of at least two adequate first-line medication trials from different classes, psychotherapy resistance as failure of at least one adequate first-line evidence-based psychotherapy trial, and then stages resistance according to whether one or both modalities and later multiple modalities have failed. These are operational criteria, especially useful for research and standardized clinical description.
Does failure of two medications mean all anxiety treatment has failed?
No. It may satisfy an operational definition of pharmacotherapy resistance if the trials were appropriate and adequate, but the consensus explicitly states that pharmacotherapy resistance does not preclude response to psychotherapy or combined treatment. The next step depends on the diagnosis, prior therapy, medical history, adverse effects, preferences, and available evidence.
Can therapy “fail” because the wrong therapy was used?
Yes. For a treatment-resistance assessment, psychotherapy should be evidence-based and appropriate for the specific disorder, with adequate intensity, duration, and fidelity. A therapy can be helpful in other ways without constituting an adequate first-line trial for the diagnosis being evaluated.
What is pseudo-resistance?
Pseudo-resistance means apparent treatment resistance when inadequate delivery, insufficient duration or dose, nonadherence, treatment mismatch, an incomplete diagnosis, or another modifiable factor may explain the lack of improvement. The term should be used descriptively, not as blame.
Can a questionnaire diagnose treatment-resistant anxiety?
No. Questionnaires can measure symptoms and change over time. They cannot by themselves establish the underlying diagnosis, determine whether past treatment was adequate, rule out medical or psychiatric alternatives, or diagnose treatment resistance.
Can treatment-resistant anxiety still improve?
Yes. Treatment resistance describes prior response, not a fixed prognosis. Evidence supports benefit from other modalities or selected second-step treatments in some people, although direct evidence becomes thinner and more disorder-specific after first-line treatment failure.
Why might medication seem to stop working?
Several possibilities can produce that experience: relapse after earlier improvement, a change in stressors or health, inconsistent use, interactions or substance effects, discontinuation or withdrawal phenomena, an evolving comorbid condition, or true loss of benefit. Sorting these apart requires a treatment and clinical reassessment rather than assuming one cause.
When should specialist reassessment be considered?
Persistent significant symptoms after adequate first-line interventions, substantial functional impairment, diagnostic complexity, important comorbidity, high-risk medication decisions, or safety concerns can justify specialist review. NICE specifically recommends specialist assessment for complex treatment-refractory GAD and specialist referral for panic disorder when significant symptoms persist after two interventions in most instances.
Do the adult treatment-resistance criteria apply to children?
No. The 2024 consensus states that its criteria are limited to adults and that childhood and adolescent criteria remain to be established. Pediatric treatment decisions should follow age- and disorder-specific evidence and assessment.
Related Articles
References
American Psychiatric Association. About DSM-5-TR. https://www.psychiatry.org/psychiatrists/practice/dsm/about-dsm
American Psychiatric Association. DSM-5-TR Online Assessment Measures. https://www.psychiatry.org/psychiatrists/practice/dsm/educational-resources/assessment-measures
Bandelow, B., Werner, A. M., Kopp, I., Rudolf, S., Wiltink, J., & Beutel, M. E. (2022). The German Guidelines for the treatment of anxiety disorders: first revision. European Archives of Psychiatry and Clinical Neuroscience, 272(4), 571–582. https://doi.org/10.1007/s00406-021-01324-1
Bokma, W. A., Wetzer, G. A. A. M., Gehrels, J. B., Penninx, B. W. J. H., Batelaan, N. M., & van Balkom, A. L. J. M. (2019). Aligning the many definitions of treatment resistance in anxiety disorders: A systematic review. Depression and Anxiety, 36(9), 801–812. https://doi.org/10.1002/da.22895
Domschke, K., et al. (2024). The definition of treatment resistance in anxiety disorders: A Delphi method-based consensus guideline. World Psychiatry, 23(1), 113–123. https://doi.org/10.1002/wps.21177
Gloster, A. T., Rinner, M. T. B., Ioannou, M., et al. (2020). Treating treatment non-responders: A meta-analysis of randomized controlled psychotherapy trials. Clinical Psychology Review, 75, 101810. https://doi.org/10.1016/j.cpr.2019.101810
National Institute for Health and Care Excellence. Generalised anxiety disorder and panic disorder in adults: management (CG113). Recommendations. https://www.nice.org.uk/guidance/cg113/chapter/Recommendations
National Institute for Health and Care Excellence. Social anxiety disorder: recognition, assessment and treatment (CG159). Recommendations. https://www.nice.org.uk/guidance/cg159/chapter/recommendations
Roy-Byrne, P. (2015). Treatment-refractory anxiety; definition, risk factors, and treatment challenges. Dialogues in Clinical Neuroscience, 17(2), 191–206. https://doi.org/10.31887/DCNS.2015.17.2/proybyrne
Schiele, M. A., Fagan, H. A., Baldwin, D. S., & Domschke, K. (2026). Integrative systematic review on pharmacological, psychotherapeutic, and neurostimulatory treatment options in treatment-resistant anxiety disorders. Psychotherapy and Psychosomatics, 95(3), 214–247. https://doi.org/10.1159/000547926
Taylor, S., Abramowitz, J. S., & McKay, D. (2012). Non-adherence and non-response in the treatment of anxiety disorders. Journal of Anxiety Disorders, 26(5), 583–589. https://doi.org/10.1016/j.janxdis.2012.02.010
World Health Organization. (2026). Anxiety disorders. https://www.who.int/news-room/fact-sheets/detail/anxiety-disorders
World Health Organization. (2024). Clinical descriptions and diagnostic requirements for ICD-11 mental, behavioural and neurodevelopmental disorders. https://www.who.int/publications/i/item/9789240077263