13.1.7 Supporting an informed patient decision

Making the choice explicit

Shared decision-making combines evidence and clinical judgement with the person's goals, concerns, previous experiences and circumstances.

The three talk approach has three steps:

1. Introduce the choice.

2. Discuss the options, including no treatment or no change, and any legal or urgent clinical constraints.

3. Support a decision.

• For less severe depression, NICE (National Institute for Health and Care Excellence) advises against routinely offering antidepressants first line unless they are the person's informed preference.

• Active monitoring, for someone who declines treatment or is improving, means information, access to help, reassessment normally within 2 to 4 weeks and follow-up after non-attendance.

Absolute numbers over a stated time

• Give absolute frequencies with the same denominator and time period for every option.

• Halving a risk from 60 in 100 to 30 in 100 matters far more than halving it from 2 in 1,000 to 1 in 1,000.

• Present both outcomes, relapse and no relapse, rather than framing only the one that favours your preferred option.

Tap to enlarge

Worked example: antipsychotic maintenance

A 2012 review pooled 65 trials with 6,493 participants. The outcomes come from different trial subsets and follow-up periods, so they illustrate the method rather than forecast one person's year.

• Relapse falls by 37 percentage points, so the number needed to treat (NNT) is about 3.

• The pooled relative risk (RR) was 0.40, slightly different from 27 / 64 because pooling uses weighted study estimates.

• Do not compare benefits and harms as though their follow-up and definitions matched when they did not.

Baseline risk

• Absolute benefit depends on baseline risk: if the RR of 0.40 held constant, higher baseline risk would give greater absolute benefit, but constancy is an assumption.

• Stabilised participants in withdrawal trials, and the withdrawal methods used, may differ from planned gradual reduction in routine care.

Communicating uncertainty

• The pooled RR had a 95% confidence interval of 0.33 to 0.49, and the NNT was 3 with a confidence interval of 2 to 3.

• When an interval includes no effect, describe the range of benefits and harms compatible with it.

• A confidence interval does not capture bias, long-term uncertainty or individual variability.

• Say plainly when the evidence covers weeks or months but the person is asking about years.

Making the options usable

• Discuss each option to comparable depth: benefits, harms, practical demands, time to benefit and review.

• Quality-assured decision aids support the discussion but do not replace it.

• Give information in manageable chunks and use teach-back to check understanding.

• Explain terms such as "response", which may mean a fall in a score rather than recovery.

• Provide suitable interpreters and communication support.

• Equality Act reasonable adjustment duties include accessible information where relevant.

• England's Accessible Information Standard (DAPB1605) requires services to identify, record, flag, share, meet and review disability-related communication needs; it does not cover every language need.

Eliciting what the person is weighing

• Ask about desired outcomes, adverse effects to avoid, tablets versus injections, monitoring, work, family, travel, costs and tolerance of uncertainty.

• "What would a good six months look like?" can reveal priorities a diagnosis does not.

• Order of presentation, detail and enthusiasm all steer choices, so explain the basis of any recommendation and do not hide alternatives.

Legal standard for disclosure

A risk is material if a reasonable person in the patient's position would likely attach significance to it, or the clinician should recognise that this particular person would (tremor or sexual effects matter differently to different people).

Capacity and supported decision-making

• Support includes hearing aids, interpreters, clearer information, a better time of day, or delaying a non-urgent decision until acute disturbance settles.

• Disagreeing with advice does not by itself show incapacity.

• Family and supporters are involved according to the person's wishes and confidentiality; they gain no decision-making authority automatically.

Detention in England and Wales

• Detention does not remove capacity or the duty to involve the person.

• Section 63 authorises specified treatment for mental disorder, subject to Part IV safeguards; unrelated physical treatment follows ordinary consent law.

• Under section 58, medication beyond three months needs consent certification or a second opinion appointed doctor (SOAD) certificate; the three months count medication given under the Act, not time since admission.

Recording the decision

• Record the options, material information, what mattered to the person, the decision and review, and offer an accessible copy.

• A person with capacity may choose differently from the clinician, subject to statutory powers.

• If they stop medication, plan the withdrawal rather than implying abrupt cessation is harmless.

• Record agreed warning signs and follow-up rather than relying on "against medical advice".