01Principles and purposeThe professional or clinical skill and the decisions it supports.
Ethical research is scientifically credible, socially valuable and proportionate to burden. A weak design can expose participants without generating useful knowledge, so ethics begins with a clear question, feasible methods, appropriate inclusion and honest uncertainty. Selection should distribute burdens and benefits fairly: exclusion requires justification, but recruitment pressure, digital-only processes or inaccessible information can also create inequity. Public involvement can improve relevance and materials but does not replace independent review.
Governance starts with classification and accountability. Research is commonly designed to generate generalisable or transferable knowledge using a defined protocol, whereas audit asks whether care meets a standard, service evaluation judges an existing service and QI tests local change. Borderline projects should use the HRA decision route and organisational research or clinical-governance office. For research, identify a sponsor that accepts overall responsibility, a qualified chief investigator, sites and delegated duties. Depending on design, approval may involve an NHS Research Ethics Committee, HRA and HCRW Approval or devolved administration, the MHRA for a clinical trial of an investigational medicinal product, confidentiality support, radiation or other specialist review. REC opinion alone does not authorise recruitment.
Consent should be adapted to risk and participant needs, not reduced to a signature. Explain purpose, procedures, alternatives, randomisation or placebo where relevant, burdens, foreseeable risk, confidentiality, data and sample use, commercial interests, expenses, compensation arrangements, withdrawal and contact routes. Check comprehension and voluntariness, especially where the treating clinician recruits a dependent patient. Capacity law is nation-specific: Mental Capacity Act 2005 section 30 and related provisions govern intrusive research involving adults lacking capacity in England and Wales; Adults with Incapacity (Scotland) Act 2000 section 51 governs Scotland; Mental Capacity Act (Northern Ireland) 2016 section 132 and associated framework apply in Northern Ireland. Children, emergency research and tissue or confidential information introduce additional rules requiring specific advice.
Delivery requires fidelity and learning. Train and delegate appropriately, use the approved version, maintain a participant identification and delegation trail, protect source data and document deviations. Safety assessment separates clinical care from research attribution: treat first, then determine seriousness, expectedness and relationship using protocol and sponsor processes. Report urgent safety measures, serious breaches and relevant adverse reactions to the correct bodies within applicable timelines. Since 28 April 2026, the Medicines for Human Use (Clinical Trials) (Amendment) Regulations 2025 are fully in effect; the current regime emphasises proportionate requirements and clearer sponsor-investigator responsibilities. Exact reporting obligations depend on study type and current guidance, so do not invent universal deadlines.
Key points
- Classify the activity before recruitment: research, audit, service evaluation and quality improvement have different purposes and governance routes. Publication intent alone does not decide classification.
- A favourable research ethics opinion is not the same as all approvals. Confirm sponsor, regulatory, HRA or devolved-nation permissions, local capacity and capability, contracts and data arrangements before starting.
- Valid consent is an ongoing process: give understandable relevant information, support questions and voluntariness, assess capacity for the decision, document consent and respect withdrawal within explained data limits.
- Adults lacking capacity are governed by different legislation across the UK. England and Wales, Scotland and Northern Ireland have distinct research provisions; never cite the Mental Capacity Act 2005 as UK-wide.
- Use only the approved protocol and current documents. Deviations, amendments, safety events, breaches and urgent safety measures must follow sponsor and regulator procedures with timely reporting.
- The amended UK clinical-trials regulations took full effect on 28 April 2026. Use current sponsor and investigator duties rather than relying on teaching based solely on the earlier regime.
02Situations and prioritiesThe context, relevant information and actions that matter most.
A participant may believe every procedure is chosen for personal benefit. Explain the research question, allocation, uncertainty and alternatives while maintaining ordinary clinical care independently.
Money, authority, detention, educational hierarchy or reliance on the treating team can compromise voluntariness. Adjust recruiter, timing, payment and advocacy rather than assuming a signed form proves free choice.
Using an outdated information sheet, adding unapproved procedures or repeatedly departing from eligibility criteria threatens rights and data integrity. Stop, protect participants and seek sponsor guidance.
An unexpected event, external signal or changed risk-benefit balance may require treatment, re-consent, amendment, urgent safety measures or suspension. Participant protection comes before administrative convenience.
Randomisation, experimental intervention or an aim to generate new knowledge may indicate research despite a “service evaluation” label. Conversely, publication does not automatically convert local evaluation into research.
03Assessment and interpretationHow to gather information, assess the situation and recognise uncertainty.
Consider the information, its meaning and its limitations before deciding what follows.
- 01
HRA project-classification route - Why
- Determine whether the activity is research and which governance office should lead.
- Interpretation and limitations
- Apply purpose and design, document the decision and seek formal advice when borderline. Labels chosen for speed do not remove research obligations.
- 02
Approval and responsibility map - Why
- Confirm sponsor, REC, regulator, national and site permissions before activity begins.
- Interpretation and limitations
- List each required approval, conditions, version and responsible person. A favourable REC opinion is one component rather than universal permission.
- 03
Consent-process assessment - Why
- Establish information, comprehension, voluntariness, capacity and documentation for this decision.
- Interpretation and limitations
- Use accessible materials and teach-back where helpful. Revisit consent when information, procedures or capacity changes; a signature cannot repair coercion or material omission.
- 04
Jurisdiction-specific capacity check - Why
- Apply the correct framework if an adult cannot consent.
- Interpretation and limitations
- Identify where recruitment occurs and consult the applicable England and Wales, Scottish or Northern Irish provisions and approved protocol; a relative is not automatically a substitute decision-maker for research.
- 05
Safety and deviation assessment - Why
- Protect the participant and determine reporting and study implications.
- Interpretation and limitations
- Give clinical care immediately, document source facts, assess seriousness and relationship with appropriate investigators and notify sponsor and authorities through current study-specific procedures.
04Worked approachesCases with ordered reasoning, an action and a check of the outcome.
01Worked caseCapacity fluctuates during an approved studyAn adult participant in England develops delirium before a research-only follow-up procedure. The original consent form is signed and a relative says to continue.+
- 1Pause the research-only procedure and treat or assess the delirium clinically. Determine whether delay changes safety and whether the participant currently has capacity for this specific procedure; do not treat the old signature as irrevocable authority.
- 2Check the approved protocol, participant information and England-and-Wales provisions for adults lacking capacity. Contact the principal investigator or sponsor and research team; clarify whether the study lawfully includes this situation and whether a consultee process is approved.
- 3Seek the participant’s present wishes and signs of objection, involve the relative in the proper advisory role rather than assuming proxy consent, and use the least burdensome lawful option. Clinical care continues regardless of research participation.
- 4Final action: proceed only if the approved study and applicable legal framework authorise it and participation remains consistent with the person’s interests or presumed wishes; otherwise defer or withdraw from the procedure and notify the sponsor as required.
- 5Verification: document capacity, consultation, version and decision, confirm any deviation or safety report, reassess if capacity returns and provide the participant with updated information.
02Practical approachA ward project is labelled quality improvementA team proposes randomising patients between an experimental device and usual care to produce generalisable evidence, without sponsor or ethics review.+
- 1Do not recruit or randomise while classification is unresolved.
- 2Use the HRA decision tools and organisational research office, documenting purpose, design, allocation and added procedures.
- 3If research, identify sponsor and obtain every required ethical, regulatory and site approval before starting.
- 4If redesigned as genuine QI, follow clinical-governance approval and protect data and patients; the label must follow the activity, not convenience.
03Escalation approachUnapproved protocol change after a safety signalAn investigator quietly changes monitoring after serious unexpected events and asks staff to use a new spreadsheet without sponsor approval.+
- 1Protect current participants, arrange clinical review and preserve the approved documents and factual safety information.
- 2Notify the principal investigator and sponsor immediately and follow urgent-safety-measure and safety-reporting routes appropriate to the trial.
- 3Stop affected research activity if continuing under the unapproved process is unsafe or unauthorised, while preserving ordinary care.
- 4Escalate suspected serious breach to governance and regulatory bodies through authorised channels and verify corrective action before resumption.
05Feedback, follow-up and evidenceReview outcomes, seek feedback and identify what to improve.
- Track recruitment, inclusion, refusals and withdrawals to identify selection pressure, inaccessible processes or unjustified exclusion, not merely numerical targets.
- Maintain current approvals, versions, delegation, training and consent evidence; reconcile site practice against the protocol and document corrective action.
- Review adverse events, deviations, data quality and emerging external evidence with the sponsor, applying current study-specific reporting requirements.
- Confirm that participants receive relevant new information, clinical follow-up and routes for questions or complaint, including after withdrawal where agreed obligations continue.
- For WPBA or research competency evidence, show classification, consent reasoning and governance closure using anonymised information; formal sponsor and regulator approval remains separate.
06Special situationsVariants, exceptions and circumstances that change the usual approach.
Ethics opinion is not operational permission
REC review addresses ethical acceptability, while sponsor, regulator, national and site bodies hold different responsibilities. Recruitment starts only when the complete route is satisfied.
Consent is revisable
A participant may change their mind, lose or regain capacity, or receive new risk information. Explain prospectively what withdrawal can and cannot mean for data already collected or safety follow-up.
Relative does not equal research proxy
The role of a consultee, guardian or representative depends on the applicable statute and approved protocol. Family knowledge informs wishes but does not create automatic authority.
Science and protection are linked
Poorly justified endpoints, underpowered design or unreliable data can make burden unethical. Methodological review is therefore part of participant protection.
Current trial law matters
The 2025 amending regulations became fully effective in April 2026. Legacy training may describe superseded processes, so check current MHRA and HRA instructions for active studies.
07Common pitfallsFrequent interpretation and management errors.
- 01
Starting recruitment after REC opinion but before sponsor, regulator, national or local permissions are complete.
- 02
Treating the consent form as permanent proof despite new information, coercion, declining capacity or participant objection.
- 03
Calling experimental randomisation quality improvement to avoid review, or calling audit research only because publication is planned.
- 04
Applying the Mental Capacity Act 2005 to Scotland or Northern Ireland without using their distinct research provisions.
- 05
Inventing a universal adverse-event deadline rather than following current study type, sponsor and regulatory requirements.