Principles of medical ethics and informed consent

Modern medical ethics emerged from historical atrocities that revealed the catastrophic consequences of unchecked medical power. The Nuremberg Code (1947) was developed in response to inhumane medical experimentations during the Nazi era, followed by the World Medical Association's Declarations of Geneva (1948) and Helsinki (1964) as ethical cornerstones of biomedical research. Subsequent scandals including the Tuskegee Syphilis study, the Stanford Prison experiment, and the Tearoom Trade study prompted the Belmont Report (1978), which outlined three fundamental principles — respect, justice, and beneficence — that continue to guide human subject research. ^[1]

Beauchamp and Childress further developed and popularised these concepts in Principles of Biomedical Ethics, establishing the four-principle framework that dominates contemporary medical ethics: autonomy, beneficence, non-maleficence, and justice. This framework derives from the Belmont Report's original conceptualisation of respect for autonomy under the notion of respect for persons. As the traditional paternalistic medical model worldwide is gradually reformed, the autonomy of patients is becoming increasingly valued; in traditional models, patients are cared for but are not decision-makers. The emphasis on autonomy results from social transformation and is also derived from these four principles. ^{[4] [5]}

Informed consent is the practical expression of respect for autonomy. Obtaining informed consent from patients is the basis for respecting patient autonomy. However, exceptions exist: according to the Chinese Mental Health Commission (2009), in an emergency when the patient poses a direct danger to others, it is legal to use physical restraints without the patient's consent. When patients lack capacity, the informed consent of close relatives is feasible in some jurisdictions because it aligns with traditional beliefs. Informed consent also serves to protect healthcare staff from disputes. ^[5]

Autonomy occupies a central position in contemporary bioethics and end-of-life ethics in particular. The principle of respect for autonomy has been challenged from different theoretical standpoints, with many authors arguing that an individualistic interpretation needs reconceptualisation from a relational viewpoint. Relational autonomy emerged as an oppositional response to individualistic interpretations, portraying individual autonomy as a basic principle in end-of-life care ethics before critiquing its origins. This concept encompasses numerous related notions: autonomy-in-relation, extended autonomy, assisted autonomy, delegated autonomy, preference autonomy, second-order autonomy, diminished autonomy, partial autonomy, autonomy in responsibility, and autonomy in solidarity. European bioethicists have used the term accompaniment to describe an association of autonomy and solidarity as social values that promote and limit each other. ^[4]

Analysis of the literature reveals great variety within relational autonomy interpretations, drawing from feminist ethics, care ethics, ethical multiculturalism, phenomenology, personalist ethics, relational ethics, virtue ethics, and political-philosophical approaches including communitarianism and liberalism. Most publications refer to a single source of inspiration, presenting a one-sided interpretation from a narrow range of ethical approaches. This diversity precludes reductionist or homogenising categorisations of this rich concept. ^[4]

In clinical practice, protecting autonomy respects the patient's personality. A qualitative study found that nurses tried to explain why they used physical restraints and what controls they used, but because most restrained patients have limited mental status, informed consent is usually not obtained. Nevertheless, most nurses participate in decision-making with patients' families. Restraint use, like any treatment, violates a person's autonomy unless patients or their agents have provided informed consent or the requirement has been excluded. Providing this information is considered a necessary condition when considering restraints and shows respect for patients as humans. ^[5]

Beneficence and non-maleficence operate as complementary obligations. Beneficence requires acting in the patient's best interest, while non-maleficence prohibits causing harm. Justice demands fair distribution of healthcare resources and equitable treatment. Moral considerations should be based on all four principles: autonomy, beneficence, non-maleficence, and justice. These principles are not hierarchical; they must be balanced in each clinical situation, with informed consent serving as the primary mechanism for operationalising respect for autonomy while acknowledging legitimate exceptions in emergencies and for patients lacking capacity.

Valid consent is not a single event but a dynamic process requiring three concurrent elements: adequate information, decision-making capacity, and voluntariness. Without any one of these, consent is legally and ethically invalid.

Information The core of information is disclosure. Doctors must explain the nature of the intervention, its risks, benefits, and alternatives. However, disclosure is not absolute. Therapeutic privilege (TP) exists as a defence where a doctor may withhold information if they hold a reasonable belief that disclosure would cause the patient serious physical or mental harm. This is a narrow exception. In practice, the trend is toward shared decision-making. For instance, community-based informed consent documents consider the influence of relatives that is wanted and expected by some patients. This acknowledges that for many, information is processed socially, not just individually. ^{[4] [3]}

Capacity Capacity is decision-specific and time-specific. It is distinct from age or diagnosis. A patient may lack capacity for complex financial decisions but retain it for simple medical ones. In older adults, intrinsic capacity declines with age. Research shows that intrinsic capacity scores tend to decline in higher age-groups even in the absence of diagnosed chronic conditions. Dementia has the greatest impact on intrinsic capacity scores in older age groups. Crucially, lack of capacity does not mean lack of autonomy. We must emphasise shared and supported decision-making that fosters autonomy for patients with compromised mental capacity. For example, a patient with mild dementia may still consent to a blood test if supported, but not to major surgery. ^[2]

Voluntariness Consent must be free from coercion. This includes pressure from family or staff. In nursing homes, the right of informed consent is often disputed. A common reason for prosecution in medical accident disputes is that the right of informed consent of family members is not guaranteed. However, the patient’s right to reject restraints is clear. Article 483 of Chapter 42 of the Federal Code of the United States prohibits caregivers from using physical restraints for discipline or convenience, and the patient has the right to reject them. Voluntariness also applies to advance care planning. Familial advance directives are signed by the patient together with the family to communicate the wish of the family as a whole. This highlights that voluntariness can be collective in certain cultural contexts, provided the patient’s voice is central. ^[5]

Special Populations In children, parents provide consent, but the child’s assent is ethically required as they develop capacity. In pregnancy, a competent woman’s refusal of treatment must be respected, even if it harms the fetus, as her autonomy is paramount. In low-resource settings, standardised documents are adapted. Two examples are familial advance directives and community-based informed consent documents. These adaptations ensure that consent remains valid despite literacy barriers or cultural norms. ^[4]

Informed consent is a cornerstone of ethical medical practice, but it is not absolute. Three recognised exceptions allow clinicians to proceed without full informed consent: emergency situations, therapeutic privilege, and waiver. Each has distinct ethical justifications and legal boundaries.

Emergency exception applies when a patient is unconscious, confused, or otherwise unable to consent, and immediate intervention is necessary to prevent death or serious deterioration. The ethical basis is beneficence—acting in the patient's best interests—combined with the practical impossibility of obtaining consent. The clinician must judge that a reasonable person would want treatment, and the intervention should be limited to what is immediately necessary. Once the patient regains capacity, consent should be sought for ongoing care. This exception does not permit treatment that a patient has previously refused while competent; an advance directive refusing specific interventions must be respected.

Therapeutic privilege (TP) is a more controversial exception. It allows a clinician to withhold information from a patient when disclosure would cause serious psychological or physical harm. However, there are very few reported cases where TP has been accepted as a defence for failing to disclose relevant information to a patient when seeking informed consent for treatment. An illustrative example is Tai v Saxon, where the doctor exercised TP and did not disclose to the patient a remote risk of perforation and fistula from a hysterectomy because she was very anxious and to do so would worsen her anxiety. The case illustrates that TP is narrowly construed—it does not justify withholding information merely because the patient might become upset or refuse treatment. It may therefore be more accurately described as therapeutic exception, rather than TP, and this is indeed the terminology adopted in some jurisdictions. ^[3]

Waiver occurs when the patient voluntarily relinquishes their right to receive information. A patient may say, "Doctor, you decide—I trust you." For a waiver to be valid, the patient must be informed that information is available, must understand they are giving up the right to receive it, and must make the choice freely without coercion. The clinician should document the waiver and periodically check whether the patient still wishes to remain uninformed. Waiver respects autonomy by honouring the patient's own choice about how much information they wish to receive—autonomy includes the freedom to delegate decisions.

These exceptions are not unlimited. They are subject to legal scrutiny, and institutional review boards may question their use, particularly in research contexts. The ethical integrity of a project depends on protecting both participants and the consent process itself. In practice, the exceptions should be invoked rarely, transparently, and always with the patient's welfare as the guiding principle.

Informed consent is the practical expression of the ethical principle of respect for autonomy. It is not a signature on a form, but a process of communication in which a patient or research participant, having been given the relevant information, voluntarily agrees to a proposed intervention or study. The legal and ethical foundations of consent rest on the idea that every adult with capacity has the right to decide what happens to their own body. For consent to be valid, three elements must be present: it must be given voluntarily, by a person with capacity, and on the basis of adequate information.

The assessment of capacity is central to valid consent. Two major legal frameworks exist for determining mental capacity, both of which produce a binary outcome: an individual either has capacity or lacks capacity to make a specific decision. The first is the functional test, where the individual's ability to make the specific decision is assessed directly. The second is a two-stage test, where an individual lacks mental capacity if their inability to decide (the functional stage) is caused by an impairment or disorder of the mind or brain (the clinical stage). In practice, capacity is decision-specific: a patient may have capacity to consent to a simple blood test but lack capacity to weigh up the risks of a complex neurosurgical procedure. The threshold for capacity should be set no higher than the complexity of the decision demands. ^[3]

For procedures, the information that must be disclosed includes the nature and purpose of the procedure, its expected benefits, the material risks, and the reasonable alternatives, including doing nothing. The standard for what counts as "material" risk is shifting from what a reasonable doctor would disclose to what a reasonable patient would want to know. When a patient lacks capacity, the law requires that all practicable steps to help them make the decision must be taken without success before treating them as unable to decide. This is a statutory principle in the Mental Capacity Act of Singapore, similar to that of England and Wales. It means that doctors must consider whether information can be presented differently, whether a family member or advocate can help, or whether more time would allow the patient to decide. ^[3]

In research, consent takes on additional weight because the primary purpose is to generate generalisable knowledge, not to benefit the individual participant. Research participants must understand that they are volunteering for a study, that they may not benefit personally, and that they can withdraw at any time without penalty. The ethical challenges in research are amplified in low- and middle-income countries (LMICs), where research staff face a range of ethical difficulties that existing guidelines and institutional review boards (IRBs) largely neglect. Current ethics guidelines and regulations often do not take into account specific conditions found in local research settings, which makes them less suited for supporting and protecting research staff in the field. This gap matters because research staff play a pivotal role in guaranteeing the success of any research endeavour, and the guiding ethical principle of "do no harm" should equally apply to any research staff involved. ^[1]

A common pitfall is treating consent as a one-off event rather than an ongoing process. Patients may agree to a procedure on Tuesday, but by Thursday, when the operation is scheduled, they may have forgotten key details or changed their minds. Consent should be revisited whenever new information emerges about risks, benefits, or alternatives, or when the patient's condition changes. A signature on a form is evidence of the process, not a substitute for it.

Another pitfall is the assumption that providing more information always improves decision-making. Doctors are expected to weigh up the probability of informational and other decisional assistance failing before deciding whether to withhold it from a patient. This suggests that doctors need not even try to provide such assistance if they predict it is unlikely to work. However, this approach is problematic because the outcome is entirely dependent on how the particular doctor interpreted the situation in determining whether to withhold the information or not. The danger in providing the patient with information is that the patient might misapprehend it and refuse a treatment that the doctors perceive is low risk and beneficial. But the patient might not consider foregoing the treatment as harmful. If the refusal is due to a lack of understanding, suitable assistance could potentially support and enhance the patient's ability to decide rather than nullifying it entirely. The lesson is that doctors should not withhold information merely because they predict the patient will make a choice they disagree with. ^[3]

A third pitfall is conflating the ethical principle of autonomy with a purely individualistic model of decision-making. A "relational turn" in current thinking on autonomy is taking place, particularly in end-of-life care. Relational autonomy recognises that patients make decisions within a web of relationships, cultural values, and social circumstances. Much misunderstanding exists about relational autonomy, since it is used as an umbrella term covering a range of diverse perspectives. In practice, this means that a patient's decision may be shaped by family members, religious beliefs, or financial constraints. Ignoring these relational factors can lead to consent that is technically valid but ethically hollow, because it fails to respect the patient as a situated person rather than an isolated rational agent. ^[4]

A fourth pitfall is the misuse of restraints in clinical care, which raises consent issues even when the patient is not explicitly asked. The use of restraints protects patients and reduces self-harm, but it may cause psychological damage to the patient. Directly or indirectly restricting or isolating patients is considered a violation of the primary goals and ethical traditions of the nursing industry. Physical restraints use conflicts with the nurse's moral responsibilities of beneficence and non-maleficence to patients. Restraint should never be a default response; it requires a documented, time-limited, and reviewed decision, ideally with the patient's consent where capacity exists. ^[5]

A fifth pitfall is failing to document the consent conversation adequately. If a dispute arises later, the medical record is the primary evidence that consent was obtained properly. Documentation should include the key information given, the patient's questions, the patient's understanding, and the fact that voluntariness was confirmed. Vague entries such as "risks discussed" are unhelpful; the material risks should be named.

Consent is used in virtually every clinical interaction, from taking a blood pressure reading to performing major surgery. In everyday practice, implied consent covers routine examinations: a patient who extends their arm for a blood pressure cuff has impliedly consented to that intervention. Expressed consent, whether verbal or written, is required for procedures that carry material risk, involve sedation or anaesthesia, or are invasive. Written consent is conventionally obtained for surgical procedures, endoscopy, chemotherapy, and participation in research studies, although the law does not require writing for most procedures; writing is an evidential safeguard.

In research, consent must be obtained from participants before any study-specific procedure is performed. This includes screening tests, questionnaires, and biological sample collection. Research ethics committees and IRBs review the consent process as part of their approval. However, the protection of research staff themselves is often neglected by existing ethics guidelines and IRBs, particularly regarding ethical standards and needs in host countries. Research staff face ethical challenges that vary by their hierarchical position—principal investigators, PhD students, research assistants, and data collectors—and by whether they are local or non-local to the study setting. In practice, this means that a research team should consider not only the consent of participants but also the ethical integrity of the working conditions for those who obtain consent. The UN Sustainable Development Goal 8.8 aims to "protect labour rights and promote safe and secure working environments for all", and research projects should not violate this goal. ^[1]

In end-of-life care, consent and capacity take on particular urgency. Decisions about resuscitation, ventilation, and palliative sedation require careful attention to the patient's current capacity, which may fluctuate. The relational dimension of healthcare is especially pronounced in end-of-life care. A patient may wish to involve family members in decision-making, or may delegate decisions entirely to a relative. Clinicians must navigate these relational dynamics while ensuring that the patient's own voice, where possible, remains central. Advance care planning allows a patient with capacity to document preferences for future care, but these preferences must be revisited if the patient retains capacity at the time the decision arises.

In emergency situations, consent may be impossible to obtain. The law permits treatment without consent when it is necessary to save life or prevent serious deterioration, provided the treatment is in the patient's best interests. This is not an exception to consent but a recognition that consent is presumed where the patient would reasonably agree if they could. Once the emergency has passed, ongoing treatment requires consent.

Finally, consent is used in medical education and training. Students must obtain consent from patients before examining them, taking histories, or performing procedures, even when the supervising clinician has already obtained consent for the overall procedure. Patients must be told that a student will be involved and that they may decline without affecting their care. This is a matter of respect for autonomy and transparency, and it is a skill that students must practise from their earliest clinical attachments.