This site is intended for health professionals only

What not to do – mental health

NICE has published an online database of outdated or non-evidence-based practices that should no longer be used. This guidance represents the view of NICE, which was arrived at after careful consideration. However, clinicians and healthcare practitioners should be advised that there is a lack of robust, high-quality evidence in mental health as defined by review groups from NICE and the Cochrane Collaboration, which largely focus on randomised controlled trials and quantitative evidence.

Patient assessment

Do not diagnose generalised anxiety disorder (GAD) solely on number, severity or duration of symptoms.1

For people who may have GAD, conduct a comprehensive assessment that does not rely solely on the number, severity and duration of symptoms, but also considers the degree of distress and functional impairment.

Do not use risk assessment tools and scales to predict future suicide or repetition of self-harm.2

There is a dearth of research in this area. A systematic review of the use of risk assessment tools on Iraq war veterans to predict suicide or suicidal self-directed violence was inconclusive. In other populations, suicide risk assessment tools should never be used diagnostically as no tool can predict this. They should be used as an adjunct to clinical experience and expertise.3,4


Do not offer a benzodiazepine for the treatment of GAD in primary or secondary care except as a short-term measure during crises

Benzodiazepines are highly effective in the short-term management of GAD.5 But the original clinical trials for this class of medication date from the early 1960s, before the separation of GAD from other anxiety disorders.6,7 There is no evidence from clinical trials to support usage greater than four to six weeks. However, a Cochrane review is under way to explore this further and has found evidence that benzodiazepines may be more effective than antidepressants in older people.8

Do not offer an antipsychotic for the treatment of GAD in primary care.

Consider referral if there is an inadequate response to step three interventions (individual high-intensity psychological intervention or drug treatment), a risk of self-harm or suicide or significant comorbidity, such as substance misuse, personality disorder or complex physical health problems or self-neglect. While monotherapy with quetiapine is effective in reducing symptoms of GAD, consider tolerability and side-effects such as weight gain, sedation and effect on an individual’s cardiometabolic profile.9

Do not offer computerised CBT (CCBT) as a treatment for GAD, specific phobias or panic disorder.

CCBT is a promising low-intensity intervention for GAD but does not yet have a substantial evidence base, although studies provide encouraging preliminary results for effectiveness and acceptability in children, young people and adults.10

Do not make a diagnosis of panic disorder after a single panic attack.11

At least two unexpected panic attacks are necessary for diagnosis and the attacks should not be accounted for by the use of a substance, a general medical condition or another psychological problem.

Do not prescribe benzodiazepines, sedating antihistamines or antipsychotics for the treatment of panic disorder.

The evidence for benzodiazepines has already been discussed. There is insufficient evidence to recommend hydroxyzine as a first-line therapy.12 The recommended treatment options have an evidence base – psychological therapy (CBT), medication (SSRI or tricyclic antidepressant) and self-help have all been shown to be effective.13

Do not use antidepressants routinely to treat persistent subthreshold depressive symptoms or mild depression.14

The risk-benefit ratio for these patients is poor, but consider them for people with a history of moderate or severe depression, with an initial presentation of subthreshold depressive symptoms that have been present for a long period (typically at least two years) or subthreshold depressive symptoms or mild depression that persists after other interventions. Studies have shown that drugs are effective in the treatment of dysthymia with no differences between and within drug classes, although tricyclic antidepressants are more likely to cause side-effects.

Do not prescribe or recommend St John’s wort for depression.

There is evidence that St John’s wort may be of benefit in mild to severe depression, but practitioners should be cautious when recommending or prescribing because of uncertainty about appropriate doses, persistence of effect, variation in the nature of preparations and potential serious interactions with other drugs (including oral contraceptives, anticoagulants and anticonvulsants). However, side-effects of St John’s wort extracts are usually minor and uncommon, and advice should be tailored to the individual.15

In unipolar major depression, do not switch to, or start, dosulepin.

Evidence supporting its tolerability relative to other antidepressants is outweighed by increased cardiac risk and toxicity in overdose. Fluoxetine was less effective than dothiepin or dosulepin, sertraline, mirtazapine and venlafaxine, which suggests that treatment decisions should also be based on patient acceptability, previous medications prescribed and cost.16

Do not use drug treatment specifically for borderline personality disorder.17

Nor should it be used for the individual symptoms or behaviour associated with the disorder – for example, repeated self-harm, marked emotional instability, risk-taking behaviour and transient psychotic symptoms. Antidepressants are not widely supported for borderline personality disorder treatment, but may be helpful in the presence of comorbid conditions. Total borderline personality disorder severity was not significantly influenced by any drug.18

Do not offer clomethiazole for community-based assisted withdrawal because of the risk of overdose and misuse.19

Clomethiazole should be used in hospital under close supervision or, in exceptional circumstances, on an outpatient basis by specialist units when daily dosage must be monitored closely. Caution is advised when prescribing clomethiazole for individuals known to be addiction prone and to outpatient alcoholics.

Do not use benzodiazepines as ongoing treatment for alcohol dependence.

Benzodiazepines showed a protective benefit against alcohol withdrawal symptoms, in particular seizures, when compared to placebo and a potentially protective benefit for many outcomes when compared with other drugs. Nevertheless, no definite conclusions about the effectiveness and safety of benzodiazepines were possible because of the heterogeneity of the trials, both in interventions and the assessment of outcomes.20

Do not offer drug treatment as an intervention to reduce self-harm.

In particular, do not offer tricyclic antidepressants as there is a high risk of toxicity with overdose. There remains uncertainty about which forms of psychosocial and physical treatments of self-harm patients are most effective.21

Do not routinely offer anticonvulsants, tricyclic antidepressants, benzodiazepines or antipsychotic medication to treat social anxiety disorder in adults.22

SSRI medication appears effective in treating social phobia over the short term (particularly amongst the SSRIs), and the long term. Nevertheless, the possibility of publication bias has to be acknowledged.23

Do not routinely offer mindfulness-based interventions or supportive therapy to treat social anxiety disorder.

The small number of studies available does not permit any conclusions to be drawn on the effectiveness of meditation therapy (transcendental meditation, relaxation therapy and Kundalini Yoga).24 Mindfulness-based cognitive therapy (MBCT) is a meditation program based on an integration of CBT and mindfulness-based stress reduction. However, the evidence base does not routinely support using MBCT for anxiety disorders as a first line treatment at present.

Do not offer St John’s wort or other over-the-counter medications and preparations for anxiety to treat social anxiety disorder.

Explain the potential interactions with other prescribed and over-the-counter medications and the lack of evidence to support their safe use.

Do not offer botulinum toxin to treat hyperhidrosis (excessive sweating) in people with social anxiety disorder.

This is because there is no good-quality evidence showing benefit from botulinum toxin in the treatment of social anxiety disorder and it may be harmful.

Dr Elizabeth England is a GPSI in mental health in Birmingham and RCGP mental health clinical commissioning champion


  1. NICE. CG113: Generalised anxiety disorder and panic disorder (with or without agoraphobia) in adults. Management in primary, secondary and community care. London: NICE; 2011
  2. NICE. CG113: Self harm (longer term management) London: NICE; 2011
  3. Suicide risk factors and risk assessment tools: a systematic review. Evidence-based synthesis program. Investigators: Elizabeth M Haney et al. March 2012
  4. Chehil S, Kutcher SP. Suicide Risk Management: A Manual for Health Professionals. 2nd Ed. Publisher 2012.
  5. Mitte K, Noack P, Steil R et al. A meta-analytic review of the efficacy of drug treatment in generalized anxiety disorder. J Clin Pharmacol 2005;25:141-50.
  6. Grosser HH, Ryan E. Drug treatment of anxiety: a controlled study of opipramol and chlordiazepoxide. Brit J Psychiat 1965;111:134-41.
  7. Nesselhof WJ, Gallant DM, Bishop MP. A double-blind comparison of WY-3498, diazepam and placebo in psychiatric outpatients. Am J Psychiat 1965;121:809-11.
  8. Goncalves DC, Byrne GJ. Interventions for generalized anxiety disorder in older adults: systematic review and meta-analysis. J Anxiety Disord 2012;26:1-11
  9. Depping AM, Komossa K, Kissling W et al. Second-generation antipsychotics for anxiety disorders. Cochrane Database of Systematic Reviews 2010, Issue 12. Art No CD008120. DOI: 10.1002/14651858.CD008120.pub2
  10. Stallard P , Richardson T, Velleman S et al. Behav Cogn Psychoth 2011;39:273
  11. NICE. CG113: Generalised anxiety disorder and panic disorder (with or without agoraphobia) in adults. Management in primary, secondary and community care. London: NICE; 2011
  12. Guaiana G, Barbui C, Cipriani A. Hydroxyzine for generalised anxiety disorder. Cochrane Database of Systematic Reviews 2010, Issue 12. Art No CD006815. DOI: 10.1002/14651858.CD006815.pub2.
  13. Schneider AJ , Mataix-Cols D , Marks IM et al. Internet-guided self-help with or without exposure therapy for phobic and panic disorders: a randomised controlled trial. Psychother Psychosom 2005;74:154
  14. NICE CG90: Depression in adults. The treatment and management of depression in adults. London: NICE; 2009
  15. Linde K, Berner MM, Kriston L. St John’s wort for major depression. Cochrane Database of Systematic Reviews 2008, Issue 4. Art No: CD000448. DOI: 10.1002/14651858.CD000448.pub3.
  16. Magni LR, Purgato M, Gastaldon Cet al. Fluoxetine versus other types of pharmacotherapy for depression. Cochrane Database of Systematic Reviews 2013, Issue 7. Art No: CD004185. DOI: 10.1002/14651858.CD004185.pub3.
  17. NICE. CG78: Borderline Personality Disorder. London: NICE; 2009
  18. Stoffers J, Völlm BA, Rücker G et al. Pharmacological interventions for borderline personality disorder. Cochrane Database of Systematic Reviews 2010, Issue 6. Art No: CD005653. DOI: 10.1002/14651858.CD005653.pub2.
  19. NICE. CG115: Alcohol dependance and harmful alcohol abuse. London: NICE; 2011
  20. Amato L, Minozzi S, Vecchi S et al. Benzodiazepines for alcohol withdrawal. Cochrane Database of Systematic Reviews 2010, Issue 3. Art No: CD005063.
  21. Hawton KKE, Townsend E, Arensman E et al. Psychosocial and pharmacological treatments for deliberate self-harm. Cochrane Database of Systematic Reviews 1999, Issue 4. Art No: CD001764. 
  22. NICE. CG159: Social Anxiety Disorder. London: NICE; 2013
  23. Stein DJ, Ipser JC, van Balkom AJ. Pharmacotherapy for social anxiety disorder. Cochrane Database of Systematic Reviews 2000, Issue 4. Art No: CD001206. 
  24. Krisanaprakornkit T, Sriraj W, Piyavhatkul N et al. Meditation therapy for anxiety disorders. Cochrane Database Syst Rev 2006;1:CD004998