Bulletin of the World Health Organization

Ethical analysis to improve decision-making on health technologies

Samuli I Saarni a, Bjørn Hofmann b, Kristian Lampe a, Dagmar Lühmann c, Marjukka Mäkelä a, Marcial Velasco-Garrido d, Ilona Autti-Rämö a


Health technology assessment (HTA) is a multidisciplinary field of policy analysis. It studies the implications of the development, diffusion and use of health technology. Its power lies in providing a joint basis for policy discussions about health care, instead of each party bringing its own calculations and then disagreeing on who is right. However, it is also a fundamentally value-laden enterprise.

HTA started in the 1970s with the primary interest of ensuring the effectiveness and safety of new health technologies.1 The cost and comparative cost-effectiveness became increasingly important as methods of assessment developed parallel to methods in health economics. More recently, the effects of health technologies on organizations, as well as legal, societal and ethical aspects of technologies, have come under the scope of HTA.2 This reflects the aim to increase the relevance and applicability of the assessments and the realization that health technologies are always applied in a social context.35 However, a shortage of accepted and practical methods for incorporating these considerations within HTA has been recognized.6,7

This paper describes a flexible, easy-to-use model for incorporating ethics into HTA. The aim is to make HTAs more internationally transferable and relevant to policy-makers in different health-care settings and cultures.

Ethics and HTA

The importance of considering technology’s impact on “social, ethical, legal and other systems” was recognized early1 and has subsequently been generally accepted.4,6,7 The importance of ethics in HTA is based on three insights.5,8 First, implementing health technologies may have moral consequences, which justifies adding an ethical analysis to a “traditional” assessment of cost and effectiveness. Second, technology also carries values and may challenge prevalent moral principles or rules of society3,911 that should be addressed by HTA.

Third, a more fundamental insight, is that the whole HTA enterprise is value laden. The goal of HTA is to improve health care, and as health care is value laden (in trying to improve the well-being of people), then HTA is value laden too. The conviction that health care and health policy should be evidence-based and decisions should be transparent is a generally accepted value-base within HTA. Important value-decisions are often made implicitly in HTA methodology: when choosing which technology to assess; interpreting research results; deciding on what counts as evidence; and whose view decides the rationality of implementing a technology.8,1114 Considering a particular HTA, the formulation of the problem, the choice of outcome measures and comparative technologies also reflect values and determine the possible outcomes of the assessment. In summary, this approach to ethics aims to uncover and justify the underlying normative structure of HTA, to assure the usefulness of the assessment.13,15

Decisions to implement technologies imply resource allocation. Choosing a technology may imply devaluing or abandoning other technologies, but may also lead to reallocation of resources within health care, or between wider sectors of society.16,17 Ideally, political decision-makers are expected to balance individual and wider societal interests, taking into consideration all values at stake. Ethical analysis within HTA can provide insight into these issues, and assist decision-makers in interpreting information in a policy-relevant way.18

Despite the general agreement on the importance of integrating ethics in HTA,5,8,14,15 ethical issues are still rarely addressed in HTA reports.1921

EUnetHTA project

The objective of the EUnetHTA project, launched in 2006, is to connect public HTA agencies, research institutions and health ministries and to enhance the exchange of information and support policy decisions. A total of 59 partner organizations participate from 31 countries.22 The ethics model belongs to a part of the project in which 25 HTA organizations are developing a generic “HTA core model” to guide future assessments.

The basic idea of the core model is to structure the contents of an HTA into pieces of information. These “assessment elements” are formulated as questions (e.g. the impact of technology on mortality). The first version contains 163 assessment elements divided into nine domains (Box 1), each with recommended methods of assessment.23 This enables a consistent structure of HTA, allowing users to find relevant and transferable information easily.

Box 1. The domains of assessment in the HTA core model

  • Health problem and current use of technology
  • Description and technical characteristics of technology
  • Safety
  • Clinical effectiveness
  • Costs and economic evaluation
  • Ethical analysis
  • Organizational aspects
  • Social aspects
  • Legal aspects

HTA, health technology assessment.

The EUnetHTA model on ethical analysis (available at: http://www.eunethta.net/Work_Packages/WP_4) aims to improve the international transferability, quality and usefulness of HTA to decision-makers by considering ethical issues relating both to the technology evaluated and to the HTA process itself. Thus, the model is not a separate exercise on ethical aspects of a technology, but integrates ethical reflection and value-awareness into the HTA process from start to finish.

A challenge in integrating ethics into HTA has been that although there is a multitude of philosophical approaches to ethics in HTA there is lack of consensus among philosophers, and a lack of methods applicable for non-philosophers.3,10,13,2426 The model does not purport to solve the philosophical debate but to offer a tool usable by HTA organizations, irrespective of their resources (material, time and knowledge). Thus the model has three elements: a question-based approach26 that covers issues essential for ethical analysis within HTA; a brief explanation of methods that can be used to approach the issues; and a discussion on the integration of ethical analysis into the process of HTA. Key issues and examples of the model are presented below.

Topic selection

Key ethical issues to consider before starting an HTA include analysing whether there are morally relevant reasons for performing an assessment on the topic or not. The value-ladenness of a technology depends on the cultural context where is applied. Assessment resources are always limited and should be used in ways that potentially benefit public health.16 This may not always coincide with commercial or political interests. An upcoming issue is global responsibility: when HTAs can be made internationally transferable, assessing globally relevant topics becomes a priority.

Planning the assessment

The selection of comparison technologies and outcome measures are essentially value-decisions that determine the results of the assessment. Also the moral value of the comparator should be considered, even if it is already widely used. “Hard” outcomes such as mortality should not automatically override relevant patient-reported outcomes,27 and difficult-to-quantify societal and organizational effects should be included.

The set of questions

The set of questions aims to increase standardization, transparency and the international transferability of the assessment. They are especially useful for identifying and cataloguing the relevant ethical considerations, allowing for several methods to weigh and balance the issues.26 The questions highlight the interwoven nature of ethics and HTA, e.g. medical, safety or economic issues also yield important input for ethical analysis.

The model has 16 questions covering the core issues of ethical analysis. These were chosen from a comprehensive list of issues by a consensus procedure so that only issues that were considered both important and internationally transferable were included. These include, for example, principal questions, such as whether the technology can challenge moral, religious or cultural values of a society. The risks of technology with respect to patient autonomy, human dignity or integrity must be addressed, especially considering vulnerable patient populations with special needs for information and support. Issues of basic human rights must be included – will the technology help in realizing these, or threaten them? A key issue is to include all stakeholders in considerations of benefits and harms. This leads to assessing the effects of the technology on the justness, equity or fairness of health care: Who will get access to the technology? What has happened to related technologies before? Is legal regulation needed? Sample questions with clarifications are presented in Table 1.

Methods for ethics

The best method for answering ethical questions varies depending on the technology, the resources available, and the national and organizational values and culture. Novel technologies may raise new ethical dilemmas (e.g. pre-implantation diagnostics) and old technologies can become controversial in new contexts (e.g. male circumcision), requiring specific emphasis on ethical analysis. Expansions of technologies considered ethically unproblematic (e.g. new antibiotics) often make thorough ethical analysis less useful. It is important to see the technology in context: societal and ethical effects of a certain technology are likely to vary depending on the cultural norms and the structure and functioning of the health-care system. Thus the model describes several methods that can be used to conduct the analysis, but leaves the selection of the most suitable method to the discretion of the user. Table 2 lists methods based on philosophical theories actually used and that have proven useful within HTA, as identified by the International Network of Agencies for Health Technology Assessment International ethics working group. Local applications are also presented.

The process of ethical analysis

The key to successful ethical analysis is integrating it into the HTA so that ethical issues are considered reflectively during the whole assessment process, starting from the planning stage; this contrasts with conducting a separate ethics exercise after all data have been gathered. The role of experts of the technology in providing the material and contextual input is emphasized. Expertise in methods of ethics can be of great benefit but is neither necessary nor sufficient. Literature searches may have to be conducted several times as new ethically relevant issues are being identified. All stakeholders must be identified, and a party responsible for the ethical aspects nominated. As HTA institutions vary greatly in their resources, mandates and organizational structures, the optimal process must be locally tailored.


The way to report the ethical issues, especially the need to make normative conclusions, depends on the mandate of the HTA agency: some agencies officially give guidance, while some just summarize evidence. Ideally, the legitimate decision-makers make the value-decisions, and thus values and assumptions underlying the assessment should be presented as openly and transparently as possible. The model includes a structure for considering and reporting outcomes of different implementation alternatives to all stakeholders separately. This increases transparency and has proven useful for presenting results to decision-makers.18,24 Detailed presentation is essential for international transferability.


We have described a model for incorporating ethics into HTA. The work is based on the insight that ethics seen as an “add on” to solve the moral issues of a technology is likely to have little effect on the implementation of the technology. Ethical analysis performed in isolation of the HTA process appears to be too narrow and comes too late. The model addresses and guides the whole HTA process, from planning, through to analysis and to preparations for decision-making, considering all stakeholders. The model is flexible to different HTA settings and easy to use, and can help make assessments more transferable between countries and increase the relevance of HTA to policy-makers.

Practical relevance of the model

The first version of the model was published in 2007, and a pilot HTA was carried out in relation to drug-eluting versus bare-metal stents in coronary artery disease.30 Although more HTAs using the model on different settings are needed, the ethical analysis faced several problems which supported the feasibility and confirmed the key principles of the model. The pilot emphasized the importance of interactive, close and reflective cooperation between ethics and other domains of assessment, which was difficult within an international collaboration. Ethical considerations could not influence the selection of topic (which showed little concern for public-health impact or global perspective in HTA), the formulation of the research questions, or the organization of the whole assessment process.

The “patient-intervention-control-outcome” structure, commonly used to formulate research questions in HTA, applies poorly to ethics. This structure asks for facts that can be analysed in an experimental setting, whereas research questions for ethical analysis are wider and require different types of information, typically the valuation of different qualitative and quantitative issues from different stakeholders’ point-of-view.

Systematic searches of the scientific literature for publications reflecting the ethical or moral implications of the technology are difficult. Few articles discuss the ethics of specific health technologies. This emphasizes the inclusion of sources of information other than scientific papers and the need for an international system of transferable ethical analysis, such as the presented model.

Ethical analysis also challenges traditional HTA methods to assess the validity of collected information. As the data range from randomized controlled trials and qualitative studies to direct focus-group hearings, balancing reliability and validity becomes a challenge. Further, balancing the values and views of different stakeholders always requires normative assessments.31 Thus, the requirement to be as transparent about the methods of summarizing and reporting the evidence is fundamentally important. A peer review of the ethical analysis not only helps to keep the analysis as neutral as possible, but also increases the commitment of stakeholders to act on the results of the HTA.18

As expected, not all of the 16 core questions were relevant, but going through them all helped to identify the relevant ethically problematic areas (in the stent case, especially the widespread off-label-use challenging the autonomy of particularly vulnerable patients.) The model was easy to use and did not require a professional ethicist.

Is the model internationally transferable?

A special challenge for HTA is – in accordance with the goals set out in the Ethics of Science and Technology Programme of the United Nations Educational, Scientific and Cultural Organization (UNESCO) – to take the global perspective and needs of the developing world into account.32,33 The model facilitates each of the two theoretical methods for doing this: to make HTAs conducted in developed countries more relevant to other countries, and to encourage local HTA creation by providing a flexible HTA model.

The question-based structure provides a low-threshold tool that does not necessitate ethical expertise. The model highlights the ethical aspects of “professional topics” and both involves professionals in ethical issues and engages ethicists in technical matters. Hence, the model can weaken the artificial borders between scientific evidence and moral goodness. The questions are open to contextual interpretation and do not limit the scope of HTA to high-tech devices or specific types of health-care organizations. Structuring information into detailed assessment elements allows users to extract exactly the information relevant to them.

However, there are also theoretical problems that must be addressed. First, does the model represent and promote a “western”, individualistic perspective that fits only certain types of health-care organizations but is inappropriate to others? Further, modern health technology (and HTA as part of it) can be seen as an “imperialistic” approach towards other health-care ideologies, emphasizing as it does rational, mechanical, non-intuitive and non-emotional ways of thinking. However, even if we accept this, integrating ethics and HTA does not necessarily have to enforce this way of thinking. On the contrary, it may be that the needs of the developing world would have been better addressed if ethical aspects had been taken onboard earlier in the HTA development.

Second, will the methods and issues of ethics be transferable between countries and cultures? It is quite clear that many moral issues and values are contextual – even more than scientific and technological issues – so is it not counterproductive to try to “transfer” an ethical model for HTA? This is important and highlights both advantages and obstacles for the model; it explains one reason for not trying to stipulate a certain, fixed method for ethical analysis within HTA. Further, the questions of the model are open to interpretation and contextual implementation. They can be rephrased, re-emphasized or give rise to local adaptations in other countries (e.g. the issue of autonomy may have quite a different meaning in Ethiopia than in England). The questions are not meant to be absolute, but are themselves subject to revision as a result of contextual use; application of the EUnetHTA model in different contexts can be highly fruitful for future development of the method itself. More than a binding framework, the model should be seen as a starting point for people wanting to begin their own, local HTAs. Finally, one can argue that some of the questions strike such basic moral aspects of health care that they will be relevant to all health-care settings (e.g. health-care interventions must benefit and not harm people, and the technologies can challenge religious, cultural or moral beliefs). In this respect, the model was created as an international collaboration, and includes only those issues that were considered both important and transferable by participating organizations. Although this aimed at making the core model more transferable, many of the excluded issues may be of importance locally.

Third, it can be argued that ethical issues are not likely to be prioritized in developing countries due to scarce resources, lack of expertise, etc. However, this presupposes a traditional HTA perspective driven by organizational resources. Actually, addressing ethical issues may fit HTA better in developing countries, because the role of technology is not yet fixed and is more open to shaping according to social needs. The ethical, social and organizational implications of a technology may be of greater importance than effectiveness in countries where resources are scarce.


The openness, transparency and flexibility of the method, as well as the fact that it does not subscribe to a particular theory in ethics, make the model suitable for wide use. Fundamentally, this is a pragmatic choice addressing the variable resources and mandates of HTA agencies, and a value-judgement that a basic consideration of ethical issues is better than no consideration at all. Still, in a sense, the EUnetHTA ethics model is very ambitious, as it not only tries to add ethics into HTA, but also to integrate ethics and HTA.

The work is based on the idea that HTA organizations have a responsibility to produce assessments that are as useful as possible to their users, ultimately benefiting public health and well-being. Ethical analysis both specifies this responsibility to the HTA organization and functions as a tool to help fulfil it. HTA is never value-free, so making values explicit is the key to increasing the international transferability and policy relevance of HTA. ■


The EUnetHTA project is supported by a grant from the European Commission. Samuli I Saarni acknowledges the Department of Mental Health and Alcohol Research, National Public Health Institute and the Department of Psychiatry, Helsinki University Hospital, Finland. Bjørn Hofmann acknowledges the Norwegian Knowledge Centre for Health Services. Ilona Autti-Rämö acknowledges the Social Insurance Institution, Finland.

Competing interests: None declared.



  • Finnish Office for Health Technology Assessment, National Research and Development Centre for Welfare and Health, Helsinki, Finland.
  • Faculty for Health Care and Nursing, University College of Gjøvik, Gjøvik, Norway.
  • Institute for Social Medicine, University of Lübeck, Lübeck, Germany.
  • Department of Health Care Management, Technische Universität Berlin, Berlin, Germany.