Choice Architecture in Lifestyle Medicine in Viet Nam: Evidence on Nudges, Ethical Safeguards, and Real-World Evaluation

RESEARCH & IMPLEMENTATIONEVIDENCE & CLINICAL PRACTICE

11/22/202629 min read

Choice Architecture in Lifestyle Medicine in Viet Nam: Evidence on Nudges, Ethical Safeguards, and Real-World Evaluation

Last reviewed: November 22, 2026

A hospital introduces a revised food-ordering menu. The meals available to patients have not changed, but options meeting selected nutritional criteria now appear more prominently. Other choices remain available, although the new arrangement may influence which meals patients notice and select.

Elsewhere, a mobile health application sends reminders when users may be deciding whether to engage in physical activity. Another service simplifies its registration process, making it easier for people to enroll in a health-support program they have already chosen.

These adjustments may seem minor, but they reflect an important insight from behavioral science: the environment in which people make decisions can influence their choices, even when the available options remain unchanged.

This is the central idea behind choice architecture.

Interest in choice architecture has grown across healthcare and public health because many health-related decisions occur under conditions of limited time, competing demands, and imperfect attention. People may understand the benefits of healthier behaviors but still find it difficult to act consistently on that knowledge.

Making a suitable option easier to notice or select may help in some circumstances. Yet influencing how choices are presented raises questions that extend beyond whether an intervention changes behavior.

Does the intervention produce a meaningful improvement, or does it merely influence an immediate decision? Will the effect persist over time? Who determines which option should be preferred? Can individuals choose differently without unnecessary difficulty? Could the design disadvantage certain groups or serve commercial interests rather than patient welfare?

These questions are especially relevant to Lifestyle Medicine, where evidence-based recommendations concerning nutrition, physical activity, sleep, and other health behaviors must be translated into decisions made repeatedly in everyday life.

Choice architecture may offer a useful additional approach, but it is not a substitute for appropriate clinical care, meaningful patient participation, or action on the social and environmental conditions that influence health.

For Viet Nam, responsible use requires distinguishing plausible behavioral mechanisms from demonstrated benefits, protecting individual autonomy, and evaluating interventions under the conditions in which they will actually operate.

Understanding Choice Architecture and Nudging

Choice architecture refers to how options are organized, presented, and made available within a decision-making environment.

People rarely encounter choices in a completely neutral setting. A menu determines which dishes appear first. A mobile application decides what information receives the greatest visual attention. A registration form may contain preselected options. A healthcare facility organizes services and communication processes in ways that make some actions easier to perform than others.

These arrangements can influence decisions even when individuals are not explicitly instructed what to choose.

The concept of nudging is closely associated with Richard Thaler and Cass Sunstein, whose 2008 book Nudge: Improving Decisions About Health, Wealth, and Happiness helped establish choice architecture as an influential topic in behavioral economics and public policy.[1]

In its conventional formulation, a nudge changes aspects of the decision environment in ways that predictably influence behavior without prohibiting options or substantially altering economic incentives.

For example, placing a healthier food option near the beginning of a menu may increase its visibility without removing other meals. Similarly, simplifying an administrative process may make it easier for someone to complete an action they already intended to take.

However, not every intervention that influences behavior qualifies as a nudge.

Prohibiting a product, imposing a substantial financial penalty, or making an alternative practically unavailable involves different mechanisms. Conventional health education, motivational interviewing, and structured behavior-change counseling should not automatically be classified as nudging simply because they influence decisions.

The distinction matters because these approaches have different purposes, evidence requirements, and ethical implications.

Choice architecture is the broader concept, while nudging represents one approach within it.

In clinical practice, nudges may complement established methods of supporting behavior change. A patient who has received individualized advice about reducing excessive dietary sodium may benefit from a food-ordering system that makes suitable options easier to identify.

The clinical recommendation and the choice architecture serve different functions. The recommendation addresses what is appropriate for the patient, while the design of the menu influences how available options are encountered.

Another related concept is boosting, which generally aims to strengthen people's decision-making capabilities rather than primarily changing the environment to steer their choices.

Helping patients understand nutritional labels, recognize misleading health claims, or compare relevant treatment information may strengthen their ability to make informed decisions.

Some interventions combine these approaches. A nutritional label can provide information for deliberate consideration while also drawing attention to particular characteristics of a product.

The boundaries are not always clear. Describing an intervention as a nudge, boost, or educational strategy should depend on what it actually does rather than how innovative the terminology sounds.

For Lifestyle Medicine, the important question is whether the intervention addresses a legitimate problem, is supported by relevant evidence, and respects the people whose decisions it may influence.

How Choice Architecture Can Influence Health-Related Decisions

Choice architecture can operate through several mechanisms, each with different practical and ethical implications.

One involves the position and order of options. An item placed at the beginning of a menu, near the top of a digital screen, or in another prominent location may be more likely to attract attention.

Another involves defaults. A default is an option that applies unless the individual actively chooses something else. Defaults can influence decisions because changing the preselected option requires additional attention or action.

Their appropriateness, however, depends heavily on the decision involved.

A default setting for displaying general educational information is different from a preselected medical intervention requiring informed consent. The fact that a default can be changed does not automatically make its use ethically or legally acceptable.

Information can also be made more noticeable through visual emphasis, simplified presentation, or framing. The wording and organization of information may influence how people understand the consequences of different choices.

Reducing friction is another approach. Removing unnecessary steps from a process may help people complete an action they have already chosen, particularly when confusing administrative procedures or digital interfaces create avoidable difficulties.

The opposite approach, deliberately making alternatives harder to select, requires greater caution. A design that creates unnecessary obstacles to declining a service may become manipulative, even when the provider describes the preferred option as beneficial.

Social information can influence behavior as well. Messages describing what other people commonly choose may affect decisions, but such information must be accurate. Invented statistics or misleading comparisons are not defensible simply because they are intended to encourage healthier behavior.

Some interventions allow individuals to organize aspects of their own decision environments. This is sometimes described as self-nudging.

A person may choose when to receive reminders, arrange their home environment to make physical activity more convenient, or establish personal preferences that support an independently selected goal.

Self-nudging may offer greater personal control, although its effectiveness still depends on the intervention and the individual's circumstances.

These mechanisms should not be treated as interchangeable.

Rearranging a physical menu involves different resources and risks from deploying a personalized digital system that continuously analyzes user behavior.

As interventions become more complex, particularly when they involve sensitive information or automated recommendations, the need for appropriate evidence, transparency, and governance becomes more important.

What Does the Overall Scientific Evidence Show?

Research on nudging has expanded considerably, but estimates of its effectiveness remain contested.

A meta-analysis by Mertens and colleagues, published in the Proceedings of the National Academy of Sciences in 2022, examined the effects of choice architecture interventions across different behavioral domains.[2]

Following a correction published in May 2022, the final analytical sample comprised 447 effect sizes from 212 publications, involving approximately 2.15 million participants. The corrected average effect was Cohen's d = 0.43, with a 95% confidence interval from 0.38 to 0.48.

The correction addressed errors in the underlying data, including observations from a subsequently retracted study, coding errors, and other analytical issues. The authors stated that the corrections did not change their overall conclusions.

The corrected results suggest a small-to-moderate average effect across the interventions studied. However, the effects varied substantially according to intervention type, behavioral domain, and study context.

Interventions changing the structure of choices generally showed larger effects than some approaches focused primarily on information or decision assistance. Food-related choices were among the more responsive domains in the included evidence.

The review also identified publication bias, which subsequently became an important point of scientific debate.

Later in 2022, Maier and colleagues published a methodological critique arguing that the overall evidence for positive nudging effects became substantially less convincing after adjustment for publication bias.[3]

The disagreement matters because estimates derived from research syntheses depend partly on which studies are available, how their results are combined, and the assumptions underlying statistical corrections.

A second-order meta-analysis by Hu and colleagues, published in the Journal of Behavioral Decision Making in December 2025, provided further evidence relevant to this debate.[4]

The authors synthesized 13 publications containing 14 meta-analyses, collectively covering 1,638 primary studies and approximately 30 million participants.

The unadjusted aggregate effect was relatively small, with Cohen's d = 0.27 and a 95% confidence interval from 0.16 to 0.38.

After applying a publication-bias adjustment, the estimated effect decreased to approximately d = 0.004.

These estimates should not be interpreted as two equally certain measurements of a universal nudging effect.

The second-order analysis combined reviews examining different behaviors, interventions, populations, and methodologies. The authors also found that most included meta-analyses were rated as having low or critically low methodological quality.

Some primary studies may have appeared in more than one meta-analysis, so the aggregate participant count should not be understood as a single independent study population.

Publication-bias adjustments also have limitations. Different statistical approaches can yield different estimates, especially when the underlying research is heterogeneous.

The findings therefore do not establish that every nudge is ineffective.

They do, however, challenge broad claims that nudging consistently produces substantial benefits across populations and settings.

The available evidence supports a more careful interpretation: some choice architecture interventions can influence behavior, but their effects vary in magnitude, reliability, and practical significance.

For healthcare decision-makers, the relevant question is not whether nudging works on average across unrelated studies. It is whether a particular intervention produces a meaningful benefit for a defined population, in a specific setting, under conditions that can be reproduced responsibly.

Digital Nudging in Lifestyle Medicine: What the 2026 Evidence Adds

Digital technologies have created additional opportunities to influence how people encounter health-related choices.

Mobile applications, websites, patient portals, chatbots, and other systems can present information, reminders, and options in different ways. Some allow users to establish their own preferences, while others use personal information to determine which messages appear.

A scoping review by Taniar and colleagues, published in Frontiers in Digital Health in July 2026, examined digital nudging techniques across Lifestyle Medicine-related domains.[5]

The review included 52 studies involving technologies such as mobile applications, websites, augmented reality, chatbots, and voice-based systems.

The researchers grouped interventions into four broad categories: one-way nudging, nudging through interactions with technology, nudging through peer interaction, and self-nudging.

Nutrition and physical activity accounted for more than three-quarters of the included studies.

Other domains, including sleep, mind-body health, and harmful substance use, were less extensively represented.

The review identified a growing diversity of approaches involving prompts, interface modifications, feedback, and personalized interactions. However, the authors also highlighted important evidence gaps, including short follow-up periods, limited diversity among study populations, and uncertainty about long-term engagement and sustained behavior change.

Because this was a scoping review, its primary purpose was to map the available research rather than establish a pooled estimate of clinical effectiveness.

The inclusion of 52 studies should therefore not be interpreted as demonstrating that digital nudging reliably improves health outcomes across all Lifestyle Medicine domains.

Technological sophistication also does not guarantee clinical value.

An application incorporating artificial intelligence, adaptive interfaces, or advanced personalization may be more technically capable than a simple reminder system without necessarily producing better health outcomes.

Digital systems can also introduce additional complexity. Users may receive repeated messages, experience notification fatigue, or become less responsive to prompts over time.

Personalized interfaces may be difficult to evaluate when their content changes continuously or when the reasons behind particular recommendations are not sufficiently transparent.

These considerations are particularly important when digital nudging is integrated into healthcare services rather than used solely for general wellness purposes.

For healthcare organizations, the decision to introduce digital nudging should depend on the problem being addressed, the intended users, the available evidence, and the potential risks.

Novelty alone is not sufficient justification.

Food Choice: From Simulated Decisions to Real-World Behavior

Food choice is among the most extensively studied applications of choice architecture.

Visibility, convenience, price, available options, and the presentation of nutritional information can influence decisions about meals and groceries.

These characteristics make food-related environments useful settings for evaluating nudges.

A systematic review by Ge and colleagues, published in BMC Public Health in August 2025, examined digital nudging interventions intended to improve dietary choices.[6]

The review included 22 randomized controlled trials reported across 17 articles, involving 33,352 participants.

The interventions included nutritional labels, changes in product accessibility or prominence, framing, salience, and priming.

Some studies found that particular labeling or presentation strategies improved the nutritional characteristics of selected foods.

However, an important methodological limitation affects interpretation.

The review excluded interventions conducted in real-world purchasing settings. The included trials used simulated online restaurant, supermarket, or grocery-shopping environments.

Many measured selected products or calculated the nutritional characteristics of hypothetical orders rather than documenting actual purchases or consumption.

The authors also identified risks of bias, particularly concerns related to participant attrition. Because the intervention types and outcomes varied substantially, the review did not produce a single pooled effectiveness estimate.

These limitations matter because selecting a healthier item in a simulated environment does not necessarily mean that the person will purchase, prepare, or consume that item in everyday life.

Even an actual purchase does not independently establish improved overall dietary quality, nutritional status, or clinical outcomes.

The findings provide useful evidence about how interface design can influence decisions under experimental conditions. They do not demonstrate durable improvements in health.

Research conducted in actual food environments offers another perspective.

A quasi-experimental study by Hünninghaus and colleagues, published in the International Journal of Behavioral Nutrition and Physical Activity in July 2025, examined menu-ordering interventions at University Hospital Essen in Germany.[7]

The researchers analyzed 26,949 meal choices made by 6,575 adult inpatients across three consecutive four-week periods.

During the initial period, patients used the standard menu. In the second period, a plant-based meal was displayed first. In the third period, the same ordering change was combined with a verbal recommendation.

After adjustment for age and sex, the odds of selecting the plant-based meal were approximately 1.95 times those observed during the baseline period for both intervention conditions.

In absolute terms, the proportion of plant-based meal selections increased by approximately 4.7 percentage points during the ordering intervention and 4.4 percentage points during the combined intervention.

Adding the verbal recommendation did not produce a statistically significant improvement over the ordering intervention alone.

These findings suggest that the position of an option can influence actual meal selection in a hospital environment.

However, an odds ratio of 1.95 does not mean that the probability of selecting the meal doubled for every patient.

The study was also quasi-experimental rather than randomized. Differences between study periods, changes in patient characteristics, and other unmeasured influences cannot be fully excluded.

Patients unable to choose from the standard main-menu options, including some with special dietary requirements, were excluded from the analysis.

The findings therefore should not be generalized automatically to hospitalized patients requiring individualized therapeutic diets.

Most importantly, the study examined meal selection. It did not establish improvements in long-term dietary intake, nutritional status, blood pressure, or other clinical outcomes.

A second example comes from a pooled analysis of four controlled trials conducted in real-world grocery stores, published by Stuber and colleagues in December 2024.[8]

The longitudinal analysis included 638 participants with the necessary purchasing and demographic information.

The pooled results did not demonstrate a statistically significant overall intervention effect on the proportion of fruit and vegetable purchases.

The analysis also did not provide convincing evidence that the interventions were more effective among participants experiencing socioeconomic disadvantage.

These findings do not invalidate every grocery-store nudge. They demonstrate that favorable results observed in some experimental settings may not translate consistently into everyday purchasing behavior.

Taken together, the simulated, hospital, and grocery-store studies illustrate why the actual setting matters.

The effect of choice architecture depends on more than the design itself. Available options, prices, patient needs, service arrangements, and the surrounding environment may all influence the outcome.

Choice Architecture Beyond Nutrition

Choice architecture may also be relevant to physical activity, sleep, and other Lifestyle Medicine behaviors, although the evidence is less extensive for several domains.

A healthcare organization might make it easier for patients to identify available physical activity programs, locate accessible facilities, or complete enrollment in a service they have already chosen.

A digital system could simplify an administrative process that previously discouraged participation.

However, improving visibility or reducing inconvenience is not the same as demonstrating sustained behavior change.

A reminder may increase application engagement without increasing physical activity. A simplified registration process may improve enrollment without improving attendance. A message about sleep may attract attention without changing the circumstances that prevent adequate rest.

For someone working rotating night shifts, reminders about maintaining a consistent bedtime may have limited value if they fail to account for working arrangements.

Similarly, encouraging physical activity through digital prompts is unlikely to resolve the problem when a person has no safe or accessible place to exercise.

Choice architecture may be most useful when it addresses a barrier that can genuinely be influenced by the decision environment.

It cannot replace changes to employment conditions, food affordability, transportation, accessibility, or healthcare service availability when those are the principal obstacles.

For Lifestyle Medicine, nudging should therefore be considered a potential complement to clinical counseling, behavioral support, and broader health-promoting environments.

It should not be promoted as a substitute for them.

Ethical Choice Architecture Must Protect Autonomy

An intervention may successfully influence behavior without being ethically justified.

Before considering how to design a nudge, a healthcare organization should be able to explain why it wants to influence a particular decision and whose interests the intervention is intended to serve.

Making evidence-supported options easier to identify may be appropriate. Designing an interface primarily to increase sales of a commercially favored product while presenting the arrangement as personalized healthcare guidance raises different concerns.

For example, a wellness platform may prominently display an expensive supplement package while making less costly or clinically appropriate alternatives difficult to locate.

Such an interface may influence purchasing behavior, but its commercial success does not establish that it benefits health.

The Organisation for Economic Co-operation and Development published Good Practice Principles for Ethical Behavioural Science in Public Policy in October 2022.[9]

The guidance emphasizes the importance of ethical consideration throughout the development and use of behavioral interventions, including problem definition, design, research, evaluation, and implementation.

Although developed for public policy, these principles provide useful considerations for healthcare organizations.

Ethical assessment should begin with the intended behavior, the reasons for influencing it, the alternatives available, and the possible consequences for different people.

A technically effective intervention may still be inappropriate if its objective is poorly justified or its design undermines the interests of those affected.

A scoping review by Shang and colleagues, published in Frontiers in Public Health in November 2025, examined ethical discussions concerning healthcare nudges.[10]

The review included 23 studies and identified recurring concerns involving autonomy, informed consent, transparency, equity, and unintended harm.

The authors proposed a framework emphasizing transparency, proportionality, meaningful alternatives, and fairness.

Much of the literature reviewed was conceptual rather than experimental. The findings therefore should not be interpreted as proof that any particular safeguard eliminates the possibility of harm.

Nevertheless, they provide a useful basis for responsible design.

An alternative that exists only in theory may not represent meaningful choice if it is hidden, difficult to access, or accompanied by unnecessary obstacles.

A default that can be changed through a clear process differs from one that requires confusing navigation, repeated attempts, or disclosure of unnecessary personal information.

The latter may create inappropriate pressure even when the preferred option is described as beneficial.

Transparency does not require a lengthy explanation beside every minor design element, but the intervention should not depend on deception or deliberately misleading presentation.

Organizations should be able to explain their purpose, the reasons for favoring an option, and how individuals can select an alternative.

People should also have appropriate opportunities to raise concerns or complaints.

The ethical question is not simply whether another option remains technically available. It is whether individuals can understand their choices and exercise meaningful control over them.

Clinical Decisions, Informed Consent, and Manipulative Design

Not every healthcare decision is appropriate for nudging.

Some choices involve significant uncertainty, potential harms, or outcomes that different patients may reasonably value differently.

An article by Ainar Miyata, published in the Journal of Medical Ethics in 2025, examined when physicians may ethically nudge patients toward particular treatments.[11]

The author argued that nudging may sometimes be permissible when it does not impair patient understanding or make refusal difficult.

However, when several clinically reasonable options exist and the patient's preferences should determine the choice, nudging toward one option may be unjustified.

This distinction is particularly important in preference-sensitive care.

A healthcare organization should not use a preselected option to steer patients toward a significant medical intervention merely because it is administratively convenient or commercially favorable.

Defaults should not be confused with informed consent.

A prechecked box or prominently displayed acceptance option is not a substitute for the information, voluntariness, and other requirements applicable to clinical consent.

Choice architecture may help organize information or make agreed care easier to access, but it must not replace clinical explanation, shared decision-making, or professional responsibility.

For Lifestyle Medicine, an application may help someone identify suitable behavior-support options or manage reminders they have chosen to receive.

It should not independently alter prescribed treatment, initiate a regulated clinical service, or make medication decisions without appropriate professional authority and clinical assessment.

The more consequential the decision, the more carefully the appropriateness of nudging should be examined.

In some situations, the responsible choice may be not to use a nudge at all.

The boundary between supportive design and manipulation is particularly important in commercial digital environments.

The term dark pattern is commonly used to describe interface designs that mislead users, obscure important information, or make certain decisions unnecessarily difficult.

Examples include confusing cancellation procedures, hidden charges, misleading visual emphasis, and repeated prompts designed to pressure users into accepting unwanted options.

In health-related services, these practices can have serious consequences. A patient may be encouraged to purchase an unnecessary service, disclose sensitive information, or believe that declining an optional product will compromise their clinical care.

Such approaches are inconsistent with patient-centered Lifestyle Medicine.

Not every default or visually prominent option is manipulative. The assessment depends on its purpose, the information provided, the practical availability of alternatives, and the consequences for the individual.

Commercial involvement is not automatically inappropriate either. However, financial relationships should be transparent, conflicts of interest appropriately managed, and health-related claims supported by relevant evidence.

An intervention should not be considered successful merely because more users select the organization's preferred option.

For healthcare, the more important questions are whether the option is appropriate, whether the decision remains informed and voluntary, and whether the intervention provides meaningful value without avoidable harm.

Equity and Cultural Relevance in Viet Nam

Nudges are sometimes presented as relatively inexpensive interventions capable of reaching large populations.

Their effects, however, may not be distributed equally.

A visually prominent message may be accessible to someone with good vision but difficult for a person with visual impairment. Digital reminders may benefit individuals with reliable connectivity while excluding others. A healthier food option may become more noticeable without becoming more affordable.

An intervention can therefore improve average outcomes while leaving existing inequalities unchanged or potentially widening them.

The pooled grocery-store analysis published in 2024 did not establish that the interventions examined were more effective among socioeconomically disadvantaged participants.[8]

This reinforces the importance of testing assumptions about equity rather than treating equitable benefit as an automatic consequence of nudging.

For Vietnamese healthcare organizations, relevant considerations may include income, age, disability, education, language, digital literacy, working arrangements, and access to services.

These characteristics should not be used to make unsupported assumptions about individual abilities or preferences. Their relevance lies in understanding whether the intervention is accessible, feasible, and respectful of different circumstances.

A choice architecture intervention developed for a European online supermarket may not perform similarly in a Vietnamese hospital canteen or community healthcare program.

Available foods, prices, dietary practices, household routines, and reasons for making particular choices may differ.

Translation alone does not establish cultural appropriateness.

A locally relevant design process should begin by understanding how people currently make decisions, what prevents them from acting on their preferences, and which aspects of the environment can realistically be changed.

Healthcare professionals, service designers, patients, and other intended users may need to examine existing workflows and discuss how proposed changes would be experienced.

This process can help identify unclear information, accessibility problems, or design elements that may be uncomfortable or inappropriate.

Co-design does not establish clinical effectiveness. An intervention may be understandable and acceptable yet have little measurable influence on behavior.

Its value lies in improving relevance and identifying problems before wider implementation.

For Viet Nam, careful local adaptation and evaluation are more defensible than assuming that an intervention shown to influence behavior elsewhere will produce equivalent effects.

From a Behavioral Idea to a Credible Evaluation

A responsible choice architecture project should begin with a clearly defined problem rather than a desire to introduce nudging as an innovative technique.

A healthcare organization may observe that patients struggle to identify appropriate information, complete registration, or recognize available health-supporting options.

Before changing the design, the team should investigate why the difficulty occurs.

The problem may involve unclear information, inconvenient workflows, limited availability, cost, accessibility, or insufficient professional support. Not every barrier can be addressed through choice architecture.

Once the problem is understood, the organization can define the intended behavior and explain why influencing it is appropriate.

The proposed mechanism should also be clear. Does the intervention make information easier to notice, remove unnecessary steps, change the order of options, or help users establish preferences they have independently selected?

The evaluation method should match the question.

Randomized controlled trials can help estimate causal effects when randomization is feasible and ethical. In a digital environment, eligible users may sometimes be randomly assigned to different interface versions. In physical settings, comparison or randomization at the level of departments, facilities, or service periods may be more appropriate.

Such designs require attention to allocation, contamination, clustering, and other methodological considerations.

Observational and quasi-experimental evaluations may be useful when randomization is impractical.

A hospital, for example, may examine whether meal choices change after a menu redesign. However, differences between periods may also reflect changes in prices, product availability, patient characteristics, staffing, or seasonal patterns.

A simple before-and-after comparison should not automatically be presented as proof that the nudge caused the observed change.

Evaluation should distinguish between exposure to an intervention, immediate decisions, subsequent behavior, and clinically relevant outcomes.

A person may notice a promoted option without selecting it. Someone may select an item in an application without purchasing it. A purchased meal may not be fully consumed.

Even sustained changes in food choice do not automatically establish improvements in clinical outcomes.

A higher application engagement rate does not prove increased physical activity. A change in menu selection does not demonstrate improved nutritional status. A short-term purchasing effect does not necessarily indicate sustained dietary improvement.

The follow-up period should therefore reflect the intended outcome.

Short experiments may be suitable for testing immediate decisions, while claims about behavior maintenance or health benefits generally require appropriately designed longer-term evaluation.

Researchers should also consider whether effects diminish as users become familiar with the intervention.

Repeated notifications may lose their influence or become irritating. A design that initially attracts attention may eventually be ignored.

Effectiveness should therefore be evaluated alongside acceptability, accessibility, user burden, and sustainability.

Behavioral interventions also require attention to safety.

A digital system may encourage excessive attention to health metrics or create pressure to achieve inappropriate targets. A food-related nudge may affect patients whose nutritional requirements differ from those of the general population. A social comparison message may discourage individuals with disabilities or other barriers.

Repeated prompts may create unnecessary interruptions or contribute to disengagement.

These possibilities depend on the intervention and population, but they should be considered even when the objective is to improve health.

An intervention that increases one desirable behavior while creating avoidable harm may not represent an overall improvement.

Ongoing monitoring should inform whether the design should be maintained, modified, evaluated further, suspended, or discontinued.

A Hypothetical Hospital Food-Choice Project in Viet Nam

Consider a hypothetical Vietnamese hospital interested in improving how meal options are presented to patients and visitors.

The hospital already offers several meals that meet relevant nutritional and food-safety requirements, but some appropriate options are difficult to identify in the ordering process.

The organization considers revising the menu so that these options are easier to notice and compare.

Before making the change, the hospital distinguishes between meals intended for the general public and those provided as part of clinical nutrition care.

Patients requiring therapeutic diets must continue receiving food appropriate to their clinical condition and prescribed nutritional requirements. A general menu preference must not override individualized dietary management.

The hospital then examines how patients and visitors currently use the ordering system. Staff may identify difficulties involving layout, language, accessibility, or inconsistent information.

Patients and other intended users may be asked whether the proposed presentation is understandable and whether appropriate alternatives remain easy to select.

Different menu layouts could then be tested through a suitable evaluation design.

The hospital might measure changes in actual meal orders and, where feasible, subsequent consumption. It should not claim improvements in blood pressure, body weight, nutritional status, or other clinical outcomes unless those outcomes have been appropriately studied.

The evaluation should also consider accessibility for older adults, people with disabilities, and individuals who may have difficulty using digital interfaces.

If the revised menu increases selection of the promoted meals, the finding would support a conclusion about meal-choice behavior under the conditions studied. It would not independently demonstrate better long-term health.

If the intervention produces no meaningful benefit, the organization should be prepared to reconsider or discontinue it.

This example is hypothetical. It does not describe an existing Vietnamese program, a validated Lifestyle Medicine intervention, or an activity implemented or endorsed by VLMA.

Any actual project would require assessment of its clinical, operational, ethical, and regulatory implications.

Digital Nudging, Personal Data, and AI Governance in Viet Nam

Digital choice architecture creates additional responsibilities when interventions collect or use personal information.

A simple interface may present the same options to every user. A personalized system may instead analyze previous decisions, activity patterns, health conditions, or other characteristics to determine which messages appear.

These approaches have different implications for privacy, transparency, safety, and governance.

In Viet Nam, the Law on Personal Data Protection No. 91/2025/QH15 entered into effect on January 1, 2026.[12]

Government Decree No. 356/2025/NĐ-CP, providing implementing requirements, also became effective on that date.[13]

Decree No. 102/2025/NĐ-CP, effective July 1, 2025, regulates digital health data management within its applicable scope.[14]

Decree No. 330/2026/NĐ-CP, effective August 19, 2026, establishes administrative sanctions for specified violations involving cybersecurity and personal-data protection.[19]

These instruments are relevant when digital interventions collect, analyze, retain, disclose, or otherwise process identifiable personal information.

Organizations must determine the lawful basis and purpose of processing, identify which information is necessary, and implement applicable safeguards concerning access, storage, security, retention, and disclosure.

Consent, where required, must satisfy the applicable legal requirements.

Participation in a health program should not be treated as unrestricted authorization to monitor personal behavior or use information for unrelated commercial purposes.

A patient who agrees to receive health-related reminders has not necessarily agreed to the disclosure of sensitive information to advertising partners or its use for additional profiling.

Digital interventions should also consider how notifications appear.

A message displayed on a shared device may unintentionally reveal information about a person's health condition or treatment. Such risks should be considered during design, not only after deployment.

When the legitimate objective can be achieved without extensive profiling, a simpler approach using less personal information may be preferable.

Some digital nudging systems may incorporate artificial intelligence to personalize messages, predict behavior, or dynamically modify interfaces.

Where an intervention involves an AI system falling within the applicable legal scope, Viet Nam's Law on Artificial Intelligence No. 134/2025/QH15, effective March 1, 2026, must also be considered.[20]

The law establishes a legal framework for AI-related activities, including requirements that may concern safety, transparency, human oversight, accountability, and risk management according to the nature and classification of the system.

Its relevance to healthcare is particularly important where AI functions may affect patient safety, health-related decisions, or access to services.

This does not mean that every reminder application or basic interface change is automatically subject to the same obligations as an AI system used for clinical decision support.

The relevant requirements depend on the system's actual functions, intended use, risk characteristics, and applicable legal provisions.

Healthcare organizations should distinguish ordinary digital communication from systems performing clinical risk assessment, decision support, or other functions that may introduce additional safety and governance responsibilities.

The use of AI does not independently authorize medical advice, diagnosis, or treatment.

Existing professional, institutional, and regulatory responsibilities remain applicable.

Patient Rights, Consumer Protection, and Commercial Boundaries

Choice architecture may be used in hospitals, healthcare applications, commercial wellness services, and other digital environments.

The applicable Vietnamese legal requirements depend on the nature of the activity.

The Law on Medical Examination and Treatment No. 15/2023/QH15, as amended, governs regulated clinical activities and relevant patient rights.[15]

Its requirements remain applicable when digital interfaces or behavioral design elements are integrated into medical examination and treatment.

A nudge does not authorize an organization or individual to diagnose disease, prescribe treatment, or make medication decisions outside their lawful professional scope.

Nor should choice architecture be used to bypass informed consent, appropriate clinical assessment, or other professional responsibilities.

Commercial activities may involve additional requirements.

The Law on Protection of Consumer Rights No. 19/2023/QH15, effective July 1, 2024, provides a legal framework relevant to consumer transactions, including those conducted in digital environments.[16]

Law No. 75/2025/QH15, amending and supplementing the Law on Advertising, became effective on January 1, 2026.[17]

These laws may be relevant when organizations promote health-related products, design purchasing interfaces, or communicate claims about the expected benefits of services.

A commercial interface should not mislead users about the nature, price, necessity, or expected benefits of an offering.

At the same time, not every default or persuasive design element constitutes a legal violation. The assessment depends on the actual conduct, claims, contractual arrangements, data processing, and other relevant circumstances.

Scientific effectiveness and legal compliance should be considered separately.

An intervention may be legally permissible without having convincing evidence of health benefit. Conversely, a behaviorally effective intervention may still require modification before it can be used lawfully.

Organizations should avoid presenting commercial influence as clinical guidance when the primary objective is to increase revenue.

Relevant financial relationships should be disclosed and conflicts of interest appropriately managed.

For VLMA, discussing choice architecture should not be confused with endorsing particular vendors, applications, supplements, or commercial behavioral marketing strategies.

The scientific question is whether an intervention provides meaningful and appropriate benefits. Its popularity or commercial performance does not establish clinical value.

Research Ethics and Institutional Responsibilities

Choice architecture projects may range from routine service improvement to formal research involving human participants.

Their classification depends on their actual purpose, design, activities, and applicable institutional and regulatory requirements.

A hospital changing the presentation of general information as part of ordinary service operations is not necessarily conducting biomedical research.

However, a project involving experimental assignment, identifiable health information, additional research procedures, or systematic investigation of clinical outcomes may create further ethical and research governance obligations.

In Viet Nam, Ministry of Health Circular No. 43/2024/TT-BYT, effective February 1, 2025, regulates the establishment, organization, and operation of ethics committees in biomedical research within its scope.[18]

The need for ethics review and other authorization should be determined according to the specific activity.

It would be incorrect to assume that every routine quality-improvement project automatically requires the same review process as an interventional clinical trial.

It would also be incorrect to assume that describing a project as service improvement removes ethical obligations that otherwise apply.

Responsible institutions should determine the applicable requirements before beginning research or implementing interventions that may create significant risks.

Participant information, consent, confidentiality, foreseeable harms, and appropriate professional oversight should be addressed according to the nature of the project.

International behavioral science frameworks may help organizations examine ethical issues, but they do not replace Vietnamese legal requirements or institutional responsibilities.

From Behavioral Ideas to Responsible Healthcare Decisions

Choice architecture offers a useful perspective on health behavior because it recognizes that decisions are influenced not only by knowledge and motivation but also by the environments in which choices occur.

This perspective is relevant to Lifestyle Medicine, where many health-supporting behaviors must be repeated during everyday life.

The way food is displayed, information is presented, or a service is organized may influence those behaviors.

However, the international evidence does not support a simple conclusion that nudges reliably improve health.

Some studies report favorable changes in immediate choices, while others identify modest, inconsistent, or absent effects. Research syntheses have also highlighted methodological limitations, publication bias, and uncertainty about how well findings translate into real-world settings.

The 2026 review of digital nudging in Lifestyle Medicine demonstrates the growing diversity of approaches under investigation, but also identifies gaps involving several behavioral domains, long-term outcomes, and population diversity.

Experimental studies of food-related choices suggest that particular design changes may influence decisions, while research in actual hospitals and grocery stores shows that effects can depend heavily on context.

These differences are not reasons to reject behavioral science. They are reasons to apply it more carefully.

For Vietnamese healthcare organizations, responsible use begins with identifying a meaningful problem, understanding why it occurs, and determining whether changing the decision environment is an appropriate response.

The intervention should have a legitimate purpose, preserve meaningful alternatives, and remain proportionate to the importance of the decision.

Patient and community perspectives can help improve its relevance and acceptability. Healthcare professionals should assess potential implications for clinical safety, while organizations need to consider data protection, consumer rights, and other applicable legal requirements.

Evaluation should measure meaningful outcomes rather than rely exclusively on engagement statistics, immediate selections, or promotional claims.

Where feasible, appropriate comparative designs can help determine whether the intervention produces an effect and whether that effect is sufficiently important to justify continued use.

Unintended consequences, unequal effects across population groups, and the resources required to maintain the intervention should also be examined.

Null or unfavorable findings are valuable when reported transparently. An intervention that produces no meaningful benefit may reveal that the original problem was misunderstood or that a different approach is needed.

A nudge that works initially but loses effectiveness over time should not automatically be expanded.

Responsible implementation includes the willingness to modify or discontinue an intervention when evidence does not support its continued use.

For the Vietnam Lifestyle Medicine Alliance (VLMA), examining choice architecture is consistent with advancing evidence-based Lifestyle Medicine through professional education, scientific knowledge exchange, and responsible discussion of implementation methods.

The objective is not to promote nudging as a universal solution or establish a separate clinical standard for behavioral design.

It is to help healthcare professionals understand how decision environments can influence behavior, recognize the limitations of current research, and appreciate the ethical and professional responsibilities involved in shaping choices.

The most important question is not simply whether a nudge changes what people select.

It is whether the intervention helps people make meaningful, informed, and beneficial decisions without creating unreasonable burdens, misleading them, or compromising their autonomy.

That is the standard against which choice architecture should be considered in Lifestyle Medicine in Viet Nam.

References and Further Reading
  1. Thaler RH, Sunstein CR. Nudge: Improving Decisions About Health, Wealth, and Happiness. New Haven: Yale University Press; 2008.

  2. Mertens S, Herberz M, Hahnel UJJ, Brosch T. The effectiveness of nudging: A meta-analysis of choice architecture interventions across behavioral domains. Proceedings of the National Academy of Sciences. 2022;119(1):e2107346118. doi:10.1073/pnas.2107346118. See also: Correction for Mertens et al. Proceedings of the National Academy of Sciences. 2022;119(19):e2204059119. doi:10.1073/pnas.2204059119.

  3. Maier M, Bartoš F, Stanley TD, Shanks DR, Harris AJL, Wagenmakers EJ. No evidence for nudging after adjusting for publication bias. Proceedings of the National Academy of Sciences. 2022;119(31):e2200300119. doi:10.1073/pnas.2200300119.

  4. Hu B, Xia Z, Guo Q, Lu C, Constantino S, Ju X. Assessing Nudge Impact: A Comprehensive Second-Order Meta-Analysis. Journal of Behavioral Decision Making. 2025;38(5):e70053. Published online December 9, 2025. doi:10.1002/bdm.70053.

  5. Taniar D, King E, Manger S, Carlisle K. Digital nudging techniques for behaviour change in lifestyle medicine: a scoping review. Frontiers in Digital Health. 2026;8:1799205. Published July 30, 2026. doi:10.3389/fdgth.2026.1799205.

  6. Ge L, Guan X, Tang L, Zhu Q, Zhong Y, Lan M. The effectiveness of digital nudging in improving dietary quality: A systematic review of randomized controlled trials. BMC Public Health. 2025;25:2804. doi:10.1186/s12889-025-24069-1.

  7. Hünninghaus K, Schäfer HC, Plonka M, et al. Exploring nudging strategies for plant-based dietary choices in hospital patients: a quasi-experimental study. International Journal of Behavioral Nutrition and Physical Activity. 2025;22:86. doi:10.1186/s12966-025-01793-w.

  8. Stuber JM, Beulens JWJ, Ayala GX, et al. Can nudge interventions targeting healthy food purchases in real-world grocery stores reduce diet-related health disparities? A pooled analysis of four controlled trials. International Journal of Behavioral Nutrition and Physical Activity. 2024;21:137. doi:10.1186/s12966-024-01687-3.

  9. Organisation for Economic Co-operation and Development. Good Practice Principles for Ethical Behavioural Science in Public Policy. OECD Public Governance Policy Papers. Paris: OECD Publishing; 2022. doi:10.1787/e19a9be9-en.

  10. Shang H, Xiong L, Chen K, Tian R, Tu J, Shang X. Ethical dimensions of healthcare nudges: a PRISMA-ScR-guided scoping review and framework for responsible behavioral governance. Frontiers in Public Health. 2025;13:1716466. Published November 26, 2025. doi:10.3389/fpubh.2025.1716466.

  11. Miyata A. When should doctors nudge? Nudging and preference-sensitive care. Journal of Medical Ethics. Published online January 2, 2025. doi:10.1136/jme-2024-110268.

  12. National Assembly of Viet Nam. Law on Personal Data Protection No. 91/2025/QH15, dated June 26, 2025. Effective January 1, 2026.

  13. Government of Viet Nam. Decree No. 356/2025/NĐ-CP, dated December 31, 2025, detailing selected provisions and implementation measures of the Law on Personal Data Protection. Effective January 1, 2026.

  14. Government of Viet Nam. Decree No. 102/2025/NĐ-CP, dated May 13, 2025, on Health Data Management. Effective July 1, 2025.

  15. National Assembly of Viet Nam. Law on Medical Examination and Treatment No. 15/2023/QH15, dated January 9, 2023, as amended, together with applicable implementing regulations.

  16. National Assembly of Viet Nam. Law on Protection of Consumer Rights No. 19/2023/QH15, dated June 20, 2023, as amended. Effective July 1, 2024.

  17. National Assembly of Viet Nam. Law No. 75/2025/QH15, dated June 16, 2025, amending and supplementing the Law on Advertising. Effective January 1, 2026.

  18. Ministry of Health of Viet Nam. Circular No. 43/2024/TT-BYT, dated December 12, 2024, regulating the establishment, organization, and operation of ethics committees in biomedical research. Effective February 1, 2025.

  19. Government of Viet Nam. Decree No. 330/2026/NĐ-CP, dated August 19, 2026, on Administrative Sanctions in Cybersecurity and Personal Data Protection. Effective August 19, 2026.

  20. National Assembly of Viet Nam. Law on Artificial Intelligence No. 134/2025/QH15, dated December 10, 2025. Effective March 1, 2026.

Editorial, Clinical, Ethical and Regulatory Note

This article is published by the Vietnam Lifestyle Medicine Alliance (VLMA) for professional education, scientific knowledge exchange, and discussion of responsible approaches to choice architecture, behavioral nudging, and evidence-based Lifestyle Medicine in Viet Nam.

It reviews selected international research, behavioral science concepts, ethical frameworks, and relevant Vietnamese legal and regulatory considerations. It is not a Vietnamese national clinical practice guideline, an approved behavioral intervention protocol, a professional licensing standard, a certification framework, or an authoritative legal opinion.

The choice architecture approaches and hypothetical implementation examples discussed in this article are illustrative. They are not presented as universally effective interventions, validated Vietnamese Lifestyle Medicine protocols, or programs already implemented or formally endorsed by VLMA.

Evidence concerning immediate choices, digital engagement, food purchasing, or other behavioral outcomes must not automatically be interpreted as evidence of improved clinical health, sustained behavior change, or reduced healthcare costs.

International research findings should not be presumed directly applicable to Vietnamese populations or healthcare settings without appropriate consideration of context, feasibility, equity, and clinical relevance.

Behavioral design must not be used to mislead individuals, obstruct meaningful alternatives, bypass informed consent, or place commercial interests above patient welfare.

Clinical assessments, treatment decisions, and other regulated healthcare activities must remain with appropriately authorized healthcare professionals and facilities operating within their lawful scope.

Digital interventions collecting or processing personal information must comply with applicable Vietnamese personal-data protection, health-data management, confidentiality, and security requirements. Where AI systems are involved, relevant AI-specific legal obligations must also be assessed according to the system's functions and intended use.

Commercial interfaces, health-related advertising, and consumer transactions must comply with the laws applicable to their actual nature and purpose.

Any research involving human participants or other activities requiring ethical review must comply with the relevant institutional and regulatory requirements.

International behavioral science frameworks and ethical guidance are presented as professional reference materials. They do not independently establish legally binding Vietnamese standards.

VLMA does not independently authorize clinical services, grant healthcare practice licenses, certify nudging interventions for regulatory purposes, approve treatment protocols, or endorse particular commercial behavioral design products through this article.

The scientific and regulatory materials referenced were reviewed through November 22, 2026. Organizations considering implementation should verify current legal requirements, professional responsibilities, and the suitability of proposed interventions for their intended population and setting.

Vietnam Lifestyle Medicine Alliance

Evidence-Based Lifestyle Medicine for Viet Nam.

Contact

contact@vlma.org

Saigon Paragon Building, No. 3 Nguyen Luong Bang Street, Tan My Ward, HCMC

+84 909 228 476

+84 84 9999 911

© 2025 - 2026. Vietnam Lifestyle Medicine Alliance. All rights reserved.

A physician-led professional platform advancing Lifestyle Medicine through education, collaboration, research, and responsible implementation.

VLMA is operated by Vietnam Lifestyle Medicine Alliance Company Limited, Enterprise Registration Number 0318902666.

VLMA is a Sister Organization of the Lifestyle Medicine Global Alliance and a Member Organization of the Planetary Health Alliance.

VLMA is not a medical service provider and does not issue medical licenses or IBLM certification.