Chapter 12. Evaluating Biotechnology, Healthcare and Human-Centred Technologies — Part 1 of 2
EVALUATING BIOTECHNOLOGY, HEALTHCARE AND HUMAN-CENTRED TECHNOLOGIES
12.1 Introduction
Biotechnology and healthcare technologies require particularly rigorous evaluation because technological errors, weaknesses or inappropriate applications may directly affect human life, health, dignity, autonomy and quality of life. Unlike many conventional technologies, healthcare technologies frequently operate in environments where the consequences of failure can be immediate, serious and sometimes irreversible.
The field encompasses a remarkably broad range of technologies, including gene editing, synthetic biology, regenerative medicine, vaccines, pharmaceuticals, medical imaging, robotic surgery, precision medicine, artificial intelligence, clinical decision-support systems, wearable devices, telemedicine, electronic health records, implantable devices, assistive technologies, digital therapeutics and neurotechnology.
These technologies create substantial opportunities for improving prevention, diagnosis, treatment, rehabilitation and health management. At the same time, they create complex questions concerning safety, informed consent, privacy, cybersecurity, equity, accessibility, professional accountability, identity and long-term consequences.
Healthcare technology evaluation therefore requires a multidisciplinary perspective. Engineering and scientific performance are important, but they must be considered alongside clinical outcomes, patient experiences, ethical principles, economic implications, organizational capability and social consequences.
Traditional health technology assessment has commonly considered dimensions such as clinical effectiveness, safety, cost-effectiveness and patient outcomes. Contemporary evaluation increasingly incorporates patient experience, health equity, ethical considerations, implementation feasibility, organizational consequences, sustainability and real-world evidence (Drummond et al., 2015; Topol, 2019).
A technology may therefore be technically impressive while providing limited practical value. For example, an artificial intelligence diagnostic system may demonstrate high accuracy in a controlled dataset but perform poorly when introduced into hospitals serving populations that differ from the training data. Similarly, a wearable health device may collect enormous quantities of information without producing meaningful improvements in health outcomes.
Evaluation must consequently move beyond the question of whether the technology works. It should also ask whether it works safely, for whom it works, under what conditions it works, at what cost it operates, what consequences it creates and how its benefits and risks are distributed among individuals and communities.
12.2 Why Healthcare Technologies Are Different
Healthcare technologies differ from many other technologies because they operate within highly sensitive biological, social and ethical environments. They frequently involve high consequences of failure, vulnerable populations, professional responsibility, informed consent, sensitive personal information, regulatory oversight, complex evidence requirements and uncertainty concerning long-term effects. They also involve interactions between humans and technological systems and may be affected by unequal access to healthcare resources.
A malfunctioning consumer device may cause inconvenience or financial loss, whereas a malfunctioning medical device may cause injury or death. An incorrect recommendation from a clinical decision-support system may influence diagnosis or treatment. Similarly, a cybersecurity breach involving a hospital information system may expose highly sensitive health information.
Healthcare technologies are also embedded within professional relationships. Doctors, nurses, pharmacists, therapists, laboratory professionals and other healthcare workers interact with technologies while simultaneously exercising professional judgment. Consequently, responsibility cannot automatically be transferred to the technology. Technology should support accountable healthcare practice rather than obscure or eliminate human accountability.
Evaluation should therefore consider the entire sociotechnical system rather than evaluating the technological artifact in isolation. A useful conceptual relationship can be expressed as Technology → Healthcare Professional → Clinical Decision → Patient → Health Outcome. Failure may occur at any point within this chain. A technically reliable system can still produce poor outcomes if users misunderstand its recommendations, if organizational procedures are inadequate, if data are incomplete or if patients cannot access the resulting service.
12.3 Clinical Effectiveness
Clinical effectiveness concerns whether a technology produces meaningful improvements in health outcomes. Evaluators should distinguish among technical efficacy, clinical effectiveness and patient benefit because these concepts are related but not identical.
Technical efficacy concerns whether the technology performs its intended function under controlled or specified conditions. For example, an evaluator may ask whether an imaging algorithm can identify abnormalities in a controlled dataset, whether a diagnostic device can detect a particular biomarker or whether a robotic surgical system can perform a specified movement accurately.
Clinical effectiveness concerns whether the technology improves outcomes when used in actual healthcare environments. Real-world clinical environments contain diverse patients, different professional practices, resource constraints, incomplete information, time pressures, equipment variation, organizational differences and unexpected clinical situations. Consequently, performance demonstrated under controlled conditions may not automatically translate into equivalent performance in practice.
Patient benefit concerns whether the technology produces an outcome that is meaningful to the individual receiving care. A treatment may produce a statistically significant improvement in a laboratory measurement without producing a sufficiently meaningful improvement in quality of life. Evaluation should therefore consider both statistical significance and clinical significance.
Relevant patient outcomes may include survival, symptom reduction, functional improvement, mobility, independence, quality of life, reduced hospitalization, reduced complications and psychological well-being. Patient satisfaction and patient-reported outcomes may also provide important evidence concerning whether the technology produces benefits that patients themselves perceive as meaningful.
12.4 Safety
Safety is one of the central dimensions of healthcare technology evaluation. Safety evaluation should identify foreseeable and reasonably foreseeable harms, including adverse events, device failure, software errors, incorrect outputs, misuse, inappropriate use, human error, interaction failures, cybersecurity incidents, system outages, environmental conditions and maintenance failures.
Safety should not be interpreted simply as the absence of known failures. Evaluators should consider uncertainty and the possibility that previously unidentified risks may emerge after widespread deployment.
A useful conceptual approach is to examine the relationship between hazard, exposure, failure mechanism, harm, severity, probability and risk control. This encourages evaluators to distinguish between a hazard and the actual risk associated with that hazard.
ISO 14971 provides a risk-management framework for medical devices and emphasizes systematic identification, evaluation, control and monitoring of risks (ISO, 2019). Safety evaluation should also consider residual risk because eliminating every possible risk may be impossible. The objective is therefore to identify risks, reduce them where reasonably possible and determine whether remaining risks are acceptable in relation to the expected benefits.
12.5 Evidence Hierarchy
Healthcare technology evaluation should consider the quality, relevance and limitations of available evidence. Evidence may originate from laboratory research, computational modelling, simulation, preclinical studies, animal studies where appropriate, observational studies, cohort studies, case-control studies, randomized controlled trials, systematic reviews, meta-analyses, health-economic evaluations, real-world evidence and post-market surveillance.
Different technologies require different forms of evidence. An implantable medical device, for example, may require evidence concerning mechanical reliability, biocompatibility, clinical outcomes and long-term complications. An artificial intelligence diagnostic system may additionally require evaluation of algorithmic performance, calibration, subgroup performance, interpretability, human interaction and post-deployment monitoring.
Evidence should also be examined in terms of validity, reliability, relevance, generalizability, reproducibility and transparency. A study may be scientifically rigorous but have limited generalizability if its participants are substantially different from the population in which the technology will ultimately be used.
This issue is particularly important for artificial intelligence. A model developed using data from one country, hospital system or demographic population may not perform identically elsewhere. Consequently, evidence must be interpreted in relation to the population, environment and circumstances in which the technology is intended to operate.
12.6 Biotechnology Evaluation
Biotechnology involves the application of biological knowledge and processes to develop products, services and technological systems. It may involve genetic information, microorganisms, cells, tissues, proteins, biological molecules, biological pathways, engineered organisms and biological production systems.
Biotechnology evaluation should therefore combine scientific, technological, clinical, environmental, ethical and regulatory dimensions. Evaluators should consider whether the underlying biological mechanism is supported by credible evidence, whether the technology is safe, whether results can be consistently reproduced and whether biological changes could produce unexpected outcomes.
Environmental consequences should also be examined, particularly where biological materials or organisms may interact with ecosystems. Ethical implications may concern human dignity, consent, ownership, identity and the appropriate boundaries of technological intervention. Regulatory compliance must also be considered to determine whether development and deployment meet applicable legal and regulatory requirements.
Social acceptance represents another important dimension. Communities, patients and other stakeholders may perceive biotechnology differently depending upon cultural values, knowledge, previous experiences and perceptions of risk. Biotechnology evaluation should therefore recognize that scientific evidence is essential but may not by itself resolve all questions concerning social adoption and public trust.
Biotechnology also requires attention to the distinction between intended and unintended consequences. A biotechnology intervention may achieve its primary objective while producing secondary effects that were not anticipated during development.
12.7 Gene Editing
Gene editing involves technologies that enable specific changes to genetic material. CRISPR-based technologies have significantly expanded the possibilities for modifying DNA and have become important tools in biological research and therapeutic development (Doudna & Charpentier, 2014).
Potential applications include the treatment of genetic diseases, development of disease models, biomedical research, development of therapeutic interventions and agricultural biotechnology. However, gene-editing technologies require particularly careful evaluation because genetic changes may have biological, ethical and societal implications.
An important distinction exists between somatic and germline applications. Somatic gene editing involves modifying cells within an individual without intentionally creating heritable genetic changes. Germline editing involves changes that may be transmitted to subsequent generations. These applications therefore raise different questions concerning safety, unintended genetic changes, long-term consequences, informed consent, intergenerational effects, human dignity, equity and governance.
Potential technical concerns include off-target effects, unintended genomic changes and uncertainty regarding long-term biological consequences. The National Academies (2017) and WHO (2021) have emphasized the importance of responsible governance, scientific oversight and careful consideration of ethical and societal implications in human genome editing.
Evaluation should therefore extend beyond the question of whether a gene can be edited. It should also consider whether the proposed editing should be undertaken, under what conditions it should be permitted, for which purposes it may be appropriate, who should determine acceptable uses and who should bear the associated risks and receive the potential benefits.
These questions demonstrate why biotechnology evaluation cannot be reduced to technical performance alone.
12.8 Synthetic Biology
Synthetic biology seeks to design, construct, modify or redesign biological systems for particular purposes. Potential applications include pharmaceuticals, vaccines, industrial biotechnology, agricultural production, biomaterials, food production, environmental remediation and biological manufacturing.
Synthetic biology can potentially improve manufacturing efficiency and enable biological systems to perform functions that are difficult or expensive to achieve through conventional technologies. However, evaluation should address both intended functionality and unintended consequences.
Biological safety is particularly important because engineered organisms may behave differently from what was anticipated during development. Environmental safety must also be considered where there is any possibility that a biological system could be unintentionally released into an ecosystem. Appropriate physical, biological or procedural containment mechanisms may therefore be necessary.
Governance is another important consideration. Researchers, organizations and governments need mechanisms capable of addressing emerging risks and ensuring that applications are developed and used responsibly. Dual-use concerns should also be considered because knowledge or technology intended for beneficial purposes may potentially be misused.
Ecological consequences require particular attention because engineered organisms may interact with existing species and ecological systems. Public acceptance is also relevant because communities may have legitimate questions concerning the purposes, safety and consequences of synthetic biology.
Synthetic biology therefore demonstrates the importance of evaluating technology throughout its lifecycle rather than only during development. A simplified evaluation pathway can be represented as Research → Design → Development → Testing → Validation → Controlled Deployment → Monitoring → Review → Modification or Retirement.
Continuous monitoring is particularly important because biological systems can behave differently when transferred from controlled environments into more complex real-world contexts.
12.9 Precision Medicine and Personalized Healthcare
Precision medicine seeks to use information about an individual's biological characteristics, environment and lifestyle to improve prevention, diagnosis and treatment. Relevant information may include genetic information, biomarkers, medical history, lifestyle information, environmental exposure, clinical records and imaging.
Potential benefits include more targeted treatment, improved diagnosis and reduced unnecessary interventions. However, precision medicine also creates challenges involving privacy, consent, data ownership, data sharing, discrimination, insurance, cybersecurity and interpretation of genetic risk.
The interpretation of genetic information requires particular care. A genetic association is not necessarily equivalent to a statistical risk, a clinical prediction or a clinical decision. These represent different stages of interpretation and should not be treated as interchangeable.
Precision medicine therefore requires evaluation not only of biological and clinical accuracy but also of the systems through which biological information is collected, interpreted, communicated and incorporated into healthcare decisions.
12.10 Artificial Intelligence in Healthcare
Artificial intelligence is increasingly used in medical imaging, diagnosis, clinical decision support, drug discovery, patient monitoring, hospital administration, predictive analytics, personalized treatment and medical research.
AI evaluation requires more than measuring accuracy. A healthcare AI system may need to be assessed according to accuracy, sensitivity, specificity, calibration, robustness, fairness, explainability, usability, safety and clinical utility.
For example, an AI model may achieve high overall accuracy while performing differently across demographic groups. Evaluation should therefore consider subgroup performance and determine whether differences could create clinically important consequences.
AI systems may also be affected by biased training data, missing data, changes in clinical practice, data drift, distribution shifts, automation bias and inappropriate user interpretation. Cybersecurity threats may introduce additional risks.
A model that performs well when initially deployed may deteriorate if the underlying population, disease patterns, clinical procedures or data-generating processes change. Consequently, AI evaluation should continue after deployment rather than ending when the system is first introduced into practice.
12.11 Medical Data and Privacy
Healthcare increasingly depends on large quantities of digital information, including electronic health records, genomic data, medical images, laboratory results, wearable-device data, patient-generated health data, hospital operational data and remote-monitoring information.
These datasets can support research and improved healthcare delivery but create substantial privacy and security responsibilities. Evaluation should therefore consider data minimization, access control, authentication, encryption, secure storage, data sharing, retention, anonymization or pseudonymization, patient consent, secondary use, cybersecurity and breach response.
A useful principle is that the more sensitive the information, the greater the responsibility for protecting it. Data governance should establish who can access information, why access is required, how access is controlled and how long information should be retained.
Privacy evaluation should also consider secondary uses of health information. Data initially collected for clinical care may later be used for research, artificial intelligence development, public-health analysis or commercial purposes. Such secondary uses create additional questions concerning consent, governance, transparency and accountability.
12.12 Digital Health and Telemedicine
Telemedicine allows healthcare services to be delivered remotely through digital communication technologies. Potential benefits include improved geographic access, reduced travel, faster consultation, support for rural communities, continuity of care and remote access to specialist expertise.
However, telemedicine may introduce limitations involving internet connectivity, digital literacy, language, accessibility, inability to conduct physical examinations, privacy, cybersecurity, technology failure and unequal access to devices.
Evaluation should therefore consider whether telemedicine improves access without creating a new digital divide. The assessment should examine access, usability, clinical quality, patient experience, safety, equity, cost and continuity of care as interconnected dimensions rather than isolated measures.
The effectiveness of telemedicine may also depend upon the type of healthcare service being delivered. Some consultations may be highly suitable for remote delivery, whereas other clinical situations may require physical examination, diagnostic testing or immediate intervention.
12.13 Wearable and Implantable Technologies
Wearable technologies can continuously or periodically collect physiological and behavioural information. Smart watches, fitness trackers, continuous glucose monitors, wearable ECG systems and sleep-monitoring devices are examples of technologies that may generate information concerning individual health and behaviour.
Implantable technologies may include cardiac devices, neural implants, drug-delivery systems, cochlear implants and implanted sensors.
Evaluation should consider measurement accuracy, reliability, battery life, comfort, biocompatibility, cybersecurity, data quality, false alarms, user adherence, maintenance and long-term effects.
A device that produces large quantities of data does not necessarily produce better healthcare. The central question is whether additional information leads to better decisions and better outcomes.
Wearable technology also raises questions concerning continuous monitoring. Users may become increasingly dependent upon measurements that are imperfect or may interpret normal variations in physiological data as signs of disease. Consequently, evaluation should consider not only technical measurement accuracy but also how users interpret and act upon the information generated.
12.14 Assistive Technologies
Assistive technologies are designed to support people with disabilities, limitations or age-related functional challenges. They include mobility devices, hearing technologies, visual-assistance systems, communication technologies, prosthetic devices, adaptive interfaces, smart home systems and robotic assistance.
Evaluation should place considerable emphasis on user experience. The evaluator should determine whether the technology improves independence, whether it is affordable and accessible, whether it is comfortable, whether users can operate it safely, whether maintenance services are available, whether it respects user dignity and whether it can be adapted to individual needs.
A technology designed without meaningful involvement from its intended users may satisfy technical specifications while failing to meet practical needs. Human-centred evaluation therefore requires users to participate meaningfully in technology development, testing and improvement.
12.15 Neurotechnology
Neurotechnology involves technologies that interact with the nervous system or measure neurological activity. Potential applications include neurological diagnosis, rehabilitation, prosthetic control, brain-computer interfaces and treatment of neurological conditions.
Because these technologies interact with aspects of human cognition, movement and potentially communication, they raise particularly complex ethical questions.
Evaluation may therefore need to consider safety, neurological effects, autonomy, privacy, mental integrity, informed consent, data ownership, long-term effects, identity and psychological consequences.
Neurotechnology demonstrates that some technologies may affect not only the physical body but also how individuals experience themselves and interact with the world. This makes conventional technical evaluation insufficient by itself. Human, ethical and social dimensions become central components of the evaluation process.
12.16 Informed Consent and Human Autonomy
Informed consent is a central principle in healthcare. Patients should receive information that enables them to make informed decisions concerning relevant interventions.
Technology can complicate informed consent because patients may not understand how an algorithm operates, how their data will be used, what risks are involved, what uncertainties remain, who may access their information or whether automated recommendations can be challenged.
Consent should therefore not become a purely administrative process. Meaningful consent involves information, understanding, voluntariness, decision-making and ongoing respect for autonomy.
For technologies that continuously collect information, consent may also need to be considered as an ongoing process rather than a single event. Patients may need opportunities to understand changes in data use, technological capabilities and potential risks over time.
12.17 Equity and Accessibility
A healthcare technology may be clinically effective while remaining inaccessible to substantial sections of the population. Evaluation should therefore consider affordability, geographic availability, infrastructure, digital literacy, disability accessibility, language, cultural appropriateness, socioeconomic differences and other barriers that may influence access.
A technology available only to wealthy populations may widen existing inequalities. This creates an important distinction between equality and equity. Equality generally concerns providing the same resources or opportunities to everyone, whereas equity recognizes that people may have different circumstances and barriers and may therefore require different forms of support to achieve meaningful opportunities to benefit.
Healthcare technology evaluation should consequently examine who benefits, who does not benefit and whether barriers to access are systematically distributed.
Accessibility should also be considered during the design stage rather than only after deployment. Designing accessible systems from the beginning can reduce the need for costly modifications and can improve usability for a wider population.
12.18 Economic Evaluation
Healthcare technologies can be expensive to develop, implement and maintain. Economic evaluation may therefore consider acquisition costs, implementation costs, training costs, maintenance, infrastructure, staffing, consumables, cybersecurity, upgrades, replacement, patient costs and wider societal costs.
Cost analysis examines what the technology costs. Cost-effectiveness analysis examines the relationship between cost and health outcomes. Cost-utility analysis examines the relationship between costs and outcomes that incorporate considerations of quality of life. Cost-benefit analysis examines whether the monetary value of benefits justifies the monetary costs.
Economic evaluation should not be interpreted as meaning that financial considerations are more important than patient welfare. Rather, economic evaluation helps decision-makers understand resource implications and opportunity costs.
Healthcare resources are often limited, and the adoption of one technology may require resources that could otherwise be used for another intervention. Understanding these trade-offs is therefore an important part of responsible healthcare planning.
12.19 Human Factors and Usability
Many healthcare technology failures are not caused by technological incapability alone. They may arise from interactions between people and technology.
Human factors evaluation examines usability, interface design, cognitive workload, alarm design, workflow, training, error prevention, accessibility and communication. It also considers how people interpret information and make decisions when interacting with technological systems.
A technically sophisticated system can fail if healthcare professionals find it difficult to use. For example, a complex interface may increase cognitive workload, which may contribute to user error and ultimately result in an incorrect action or patient risk.
Human-centred design therefore attempts to design technologies around actual users and working environments rather than expecting users to adapt completely to technological systems.
12.20 Organizational and Implementation Evaluation
A healthcare technology does not operate independently of the organization. Successful implementation may depend on leadership, organizational culture, staff skills, infrastructure, financial resources, training, workflow integration, technical support and institutional policies.
A technology may therefore be technically effective but organizationally unsuitable.
Evaluation should determine whether the organization can implement the technology, whether staff can use it effectively, whether it can be maintained, whether it can be integrated into existing workflows, whether the organization can monitor its performance and whether it can respond effectively when something goes wrong.
Implementation evaluation should also recognize that introducing technology may change professional roles and organizational processes. New technologies can alter responsibilities, communication patterns, decision-making procedures and relationships between professional groups.
12.21 Regulation and Governance
Healthcare and biotechnology technologies require appropriate governance because their consequences can extend beyond individual users. Governance may involve governments, regulatory agencies, healthcare organizations, professional bodies, research ethics committees, technology developers, patients and public-interest organizations.
Regulatory evaluation may consider safety, efficacy, manufacturing quality, data protection, clinical evidence, ethical compliance and post-market surveillance.
Governance should also evolve as technologies evolve. A regulatory system designed primarily for conventional medical devices may require adaptation when technologies become continuously learning, software-driven and connected to large data ecosystems.
Effective governance should establish responsibilities, accountability mechanisms, reporting procedures, oversight structures and processes for responding to emerging risks. Governance should also provide mechanisms through which patients and the public can understand how important healthcare technologies are being developed and used.
12.22 Post-Market and Real-World Monitoring
Evaluation should not necessarily end when a technology receives approval or enters clinical practice. Real-world monitoring can identify unexpected adverse events, new patterns of misuse, performance deterioration, population-specific problems and cybersecurity vulnerabilities. It can also reveal changes in user behaviour and long-term effects that may not have been apparent during controlled testing.
The evaluation cycle can therefore be understood as Development → Testing → Approval → Deployment → Monitoring → Evidence Collection → Reassessment → Modification.
This approach recognizes that technologies exist within changing environments. New patient populations, clinical practices, software updates, cybersecurity threats, organizational changes and external conditions may alter technological performance.
Continuous evaluation is therefore particularly important for digital and AI-enabled healthcare technologies because these systems may be updated or may interact with changing data environments after initial deployment.
12.23 Environmental Sustainability
Healthcare technologies also have environmental consequences. Evaluation may consider energy consumption, electronic waste, pharmaceutical waste, manufacturing emissions, packaging, transportation, water consumption, raw materials, disposal and recycling.
For example, widespread deployment of connected medical devices may generate substantial quantities of electronic waste. Digital healthcare systems may also require significant computing infrastructure and energy consumption.
Sustainability should therefore be incorporated into technology evaluation rather than treated as a completely separate issue.
Lifecycle thinking is especially useful because environmental consequences may occur at multiple stages. A technology may have low operational environmental impact while requiring environmentally intensive materials during manufacturing or creating substantial waste at the end of its useful life.
12.24 A Human-Centred Evaluation Framework
The evaluation of biotechnology, healthcare and human-centred technologies can be organized into an integrated framework beginning with the technology itself and progressing through scientific validity, technical performance, safety and risk, clinical effectiveness, patient benefit, human factors and usability, ethics and autonomy, privacy and security, equity and accessibility, economic sustainability, organizational implementation, environmental sustainability and real-world monitoring.
Figure X.X A Human-Centred Evaluation Framework — Conceptual Framework for the Multidimensional Evaluation of Biotechnology, Healthcare and Human-Centred Technologies
- TECHNOLOGY
- Scientific Validity
- Technical Performance
- Safety and Risk
- Clinical Effectiveness
- Patient Benefit
- Human Factors and Usability
- Ethics and Autonomy
- Privacy and Security
- Equity and Accessibility
- Economic Sustainability
- Organizational Implementation
- Environmental Sustainability
- Real-World Monitoring
- Continuous Evaluation and Improvement
Framework interpretation: The framework emphasizes that technology evaluation is multidimensional. A technology should not be considered successful simply because it achieves technical performance. Its broader effects on individuals, organizations, communities and the environment must also be examined.
Source: Developed by the researcher.
The framework emphasizes that technology evaluation is multidimensional. A technology should not be considered successful simply because it achieves technical performance. Its broader effects on individuals, organizations, communities and the environment must also be examined.
The framework can also be understood as a continuous rather than purely linear process. Evidence obtained during real-world monitoring may identify new safety concerns, usability problems, equity issues or environmental consequences. These findings may then require modification of the technology, its implementation procedures or the conditions under which it is used.
12.25 Integrated Evaluation Questions
A comprehensive evaluation should begin by determining whether the underlying science is credible and whether the technology performs its intended technical function. The evaluation should then establish whether the technology produces meaningful health benefits and identify what could go wrong and how serious the consequences could be.
The human dimension requires consideration of whether the technology respects dignity, autonomy and individual needs. Ethical evaluation should consider whether the purposes and consequences of the technology are consistent with relevant ethical principles. Privacy evaluation should determine whether personal and health data are appropriately protected and whether data are being collected, processed and shared responsibly.
Equity evaluation should examine who can access the technology and who may be excluded. Economic evaluation should consider whether the costs are proportionate to the benefits and whether the necessary resources are available. Organizational evaluation should determine whether healthcare organizations can implement, operate and maintain the technology effectively.
Environmental evaluation should examine the technology's lifecycle consequences, including production, operation, maintenance and disposal. Regulatory evaluation should establish whether the technology satisfies applicable legal, professional and regulatory requirements.
Finally, long-term evaluation should consider what may happen after widespread adoption. Technologies can produce consequences that are difficult to identify during initial development or controlled testing. Continuous monitoring and reassessment are therefore essential components of responsible technological governance.
The central principle of this chapter is that the evaluation of healthcare and biotechnology must extend beyond whether a technology works to whether it works safely, effectively, ethically, equitably and sustainably within the human systems in which it operates.
Healthcare technologies do not simply interact with machines. They interact with human bodies, human decisions, human relationships, institutions and societies. Consequently, responsible technology evaluation must recognize the interconnected nature of technological performance, human welfare, organizational capability, social equity, ethical responsibility and environmental sustainability.
imaging;
12.26 Case Study: Wearable Health Monitoring
Consider a wearable health-monitoring device designed to detect abnormal heart rhythms. Such a device may continuously or periodically monitor physiological signals and provide an alert when an abnormal pattern is detected. At first sight, the evaluation of the device might appear to be primarily a technical exercise involving the accuracy of the sensor and the reliability of the algorithm. A comprehensive evaluation, however, requires substantially broader consideration.
The first question is technical: can the sensor accurately detect the relevant physiological signal under the conditions in which the device is intended to operate? Evaluation should examine signal quality, sensor accuracy, reliability, battery performance, connectivity, environmental interference and the ability of the system to distinguish genuine physiological abnormalities from artefacts generated by movement or other sources.
The second question is clinical: does detecting an abnormal heart rhythm actually improve patient outcomes? Detection by itself does not necessarily constitute a health benefit. The notification must lead to an appropriate clinical response, and that response must have the potential to improve diagnosis, treatment, prevention or patient safety. Evaluation should therefore consider whether the device contributes to earlier intervention, improved monitoring, reduced complications or other meaningful clinical outcomes.
The third question is human-centred: will users understand what the notification means and what they should do after receiving it? A patient may receive an alert without understanding whether the situation is urgent, whether medical attention is required or whether the notification could represent a false alarm. Poor communication may therefore create confusion or unnecessary anxiety. The design of the notification system, the language used, the accessibility of the information and the availability of appropriate guidance are important components of evaluation.
The fourth question is ethical: does continuous monitoring create anxiety, excessive surveillance or inappropriate medicalization of ordinary physiological variation? Continuous access to health information may provide reassurance for some users while creating uncertainty or worry for others. Users may repeatedly check their physiological measurements and interpret minor variations as evidence of serious illness. The evaluation should therefore consider not only the potential benefits of continuous monitoring but also its psychological and behavioural consequences.
The fifth question is economic: is the technology affordable for the intended population? The total cost may include the device itself, software subscriptions, replacement components, connectivity, technical support and clinical follow-up. A technology that is technically effective but available only to a small proportion of the population may have limited potential to improve population health.
The sixth question concerns governance and information management. The evaluator must determine who receives the health information generated by the device, how the information is stored, who can access it, whether it is shared with healthcare providers, whether third parties can use it and what procedures exist in the event of unauthorized access or data loss.
The seventh question concerns environmental sustainability. The evaluation should examine the materials used to manufacture the device, energy consumption, packaging, transportation, battery replacement, electronic waste, recycling and final disposal. A technology that is beneficial from a clinical perspective may nevertheless create environmental burdens that should be recognized within a complete lifecycle assessment.
This example demonstrates why multidimensional evaluation is necessary. The success of wearable health monitoring cannot be established merely by demonstrating that the sensor detects a physiological signal. The evaluation must determine whether the technology produces meaningful health benefits while maintaining safety, privacy, affordability, usability, autonomy, equity and environmental responsibility.
12.27 Human-Centred Technology
Human-centred technology begins with human needs rather than technological capability. Instead of asking what a technology can do and subsequently searching for a problem to which it can be applied, human-centred evaluation begins by identifying a genuine human, organizational or societal need and then determining whether technology provides an appropriate means of addressing that need.
This principle is closely related to usability, accessibility, inclusion, safety and responsible innovation. A technology may possess sophisticated technical capabilities while providing little practical value if users cannot understand it, cannot access it, cannot afford it or do not perceive it as useful.
Human-centred design therefore requires attention to the context in which technology is actually used. ISO 9241-11 provides a framework for understanding usability in relation to effectiveness, efficiency and satisfaction within a specified context of use (ISO, 2018b). The concept of usability is consequently broader than simply asking whether an interface is easy to operate. It concerns whether users can achieve intended objectives effectively, efficiently and satisfactorily within the environment in which the technology is used.
Human-centred evaluation should therefore examine actual users in realistic environments. Laboratory testing can provide valuable evidence, but real-world conditions may introduce distractions, resource limitations, time pressures, accessibility barriers, organizational constraints and unexpected patterns of use. Evaluation should consequently include representative users and realistic contexts wherever possible.
Human-centred technology should also recognize diversity among users. Individuals differ in age, education, language, digital literacy, physical ability, cognitive capacity, cultural background and technological experience. A design that works well for one group may create barriers for another. Inclusive evaluation therefore requires consideration of diverse users rather than assuming that a single average user represents the entire population.
The human-centred approach also recognizes that technology should be adaptable. Human needs may change over time, and technologies may be introduced into environments different from those originally anticipated by developers. Continuous user feedback can therefore contribute to iterative improvement.