1.1. The forerunners

The first HTA agencies were established in the United Kingdom (NICE, 1999), Sweden (SBU, 1987), Denmark (DACEHTA, 1997) and Finland (FinOHTA, 1995). In addition, the Netherlands were very active in developing HTA as an academic discipline and a policy-making tool and repeatedly discussed throughout the 1980s and early 1990s creating a national HTA agency. However, no such body was created until 2006 (see Banta and Oortwijn, Reference Banta and Oortwijn2009). Taken in a global context, other countries active in HTA and establishing HTA bodies during the 1980s and 1990s include Norway, Australia, Canada and of course the United States, whose Office for Technology Assessment is usually described as the Ur-HTA agency (Banta and Jonsson, Reference Banta and Jonsson2009). Lastly, Spain has had a relatively long history of HTA at the regional level (Sampietro-Colom et al., Reference Sampietro-Colom, Asua, Briones and Gol2009), although competences of the national body responsible for HTA are de facto very limited to this day and Spain is naming the creation of a central HTA agency as one of its top priorities for health policy (Cappellaro et al., Reference Cappellaro, Fattore and Torbica2009).Footnote ²

These countries were the ones to lay the groundwork and experiment with different powers and designs of HTA agencies. The competences, jurisdictions and often also names of these agencies evolved throughout the years: for example, the Swedish SBU no longer exists as it was replaced and merged with other institutions. Similarly, there have been developments in Denmark and in Finland in the level of centralization and sharing of institutional responsibilities of HTA functions among different institutions.

The mainstreamers

The next big wave of HTA agencies’ proliferation began in the mid-2000s as the idea of an institution of public interest charged with HTA as a tool to aid coverage decision-making made its way to mainstream health policy in Europe. Seven EU member states set up national HTA bodies in this period: Belgium (KCE, 2004), France (HAS, 2004), Germany (IQWiG, 2004), Hungary (GYEMSZI, 2004), Poland (AHTAPol, 2005), Austria (LBI, 2006) and Ireland (HIQA, 2007), and with some delay Croatia, whose Quality and Accreditation Agency (AAZ) created an HTA department in 2009. A specific case is Latvia, which had an organizationally separate body for HTA (Health Economics Centre, VEC) from 2009 to 2011. In 2011, VEC was merged with a payment organization as a consequence of massive cuts in public administration provoked by the economic crisis (see also Mladovsky et al., Reference Mittermayer, Huić and Meštrović2012; Nacionālais veselības dienests, 2012). In addition, CVZ was created in the Netherlands in 2006 as somewhat of an outlier case for a country so active in HTA practice, yet doting itself of an HTA body relatively late. In Italy (and Spain), HTA was recognized politically at a national level in 2006 by national health plans, with ministries of health taking on supporting roles to develop and coordinate HTA at regional level. In the United Kingdom, Scotland established an HTA agency in 2002 (SMC).

The mainstreamers set up their HTA bodies following a relatively similar model, sometimes explicitly acknowledging their source of inspiration, for instance, NICE that serves as a model to the Polish AHTAPol and the Scottish SMC (Kolasa and Wasiak, Reference Kolasa and Wasiak2012). As a result, most of these HTA bodies have the formal status of independent institutes financed from public sources, with advisory roles only. Of course, the HTA landscape is more varied than a single national autonomous HTA agency (or regional ones in some decentralized health systems): in Ireland, for example, a university institute (National Centre for Pharmacoeconomics) acquired advisory functions to the ministry of health regarding HTA in 2006, one year before HIQA was established. Similarly, in Finland, the Centre for Pharmacotherapy Development ROHTO as a unit within the health ministry, albeit with some autonomy, was created in 2003 in addition to earlier HTA bodies. Nevertheless, HTA agencies created in the 2000s look surprisingly alike in their institutional set up.

The non-adopters

The rest of EU member states belong to the last category, non-adopters, who have so far not followed the trend of HTA agencification. Bulgaria, Cyprus, Czech Republic, Estonia, Greece, Latvia, Lithuania, Luxembourg, Malta, Portugal, Romania, Slovakia and Slovenia have not yet set up HTA agencies, although the topic has been on the agenda in many of these countries. Slovenia, in particular has been discussing the creation of HTA bodies since the mid-2000s (Turk and Tit, Reference Turk and Tit2008; Tit, Reference Tit2010). Romania has postponed creating an HTA body until 2012, when it introduced HTA into its pricing and reimbursement decision-making under pressure from the International Monetary Fund and the World Bank (Gulácsi et al., Reference Gulácsi, Rotar, Niewada, Löblová, Petrova, Boncz and Klazinga2014). The body in question is a unit within the ministry of health rather than a full-blown agency. A similar development followed in Lithuania, where HTA as a scientific discipline has had a relatively long tradition (Danguole, Reference Danguole2009). The country recently established an HTA unit within the ministry of health (Wild et al., 2015). In line with our reasoning that the agency model itself is significant, a recent report describes this unit as underwhelming (Wild et al., 2015).

This is not to say that some aspects of HTA as a method are not used in these countries. Health economics and selected elements of HTA have been reported in use by academics as well as some decision-makers (reimbursement committees of ministries of health or payers) (Sorenson et al., Reference Sorenson, Kavanos and Karamalis2009; Gulácsi et al., Reference Gulácsi, Orlewska and Péntek2012). In other words, there is probably no country in the EU today without any notion of HTA, but the decisive step of creating an institution whose main responsibility would be to perform HTA analyses has not been taken in the non-adopting countries.

It is not surprising that the forerunners correspond to the group of European countries that traditionally set trends in health policy (and which lead developments beyond the health sector more generally, in public policy, administration and management).Footnote ³ However, the distribution of countries between mainstreamers and non-adopters is puzzling as it to a large extent mirrors the Iron Curtain: most of the mainstreamers are from Western Europe, while CEE countries have been significantly more hesitant and form the majority of the non-adopters. Given that other trends in health care policy have spread to CEE just as across the old EU-15, the question therefore is why Western European countries adopted HTA agencies en masse in the mid-2000s but most CEE countries resisted their creation or postponed it – and why there are exceptions to this rule such as Hungary or Poland.

Resource limitations

Lack of resources, financial or human, is often put forward by practitioner-oriented capacity-building literature (cf. Moharra et al., Reference Moharra, Espallargues, Kubesch, Estrada, Parada, Vondeling and Raab2009). It is reasonable to assume that financial resources are needed to set up a new agency. A new public agency whose mission is to advise decision-makers needs qualified experts who need to be remunerated and new staff needs to be trained. Countries where a new HTA agency would require a significant additional investment that they cannot afford are less likely to establish one – or are more likely to lag in diffusion in comparison with other countries because the high costs of policy adoption make waiting for more evidence about the policy more attractive (see Brooks, Reference Brooks2007). At first sight this seems rather plausible when we look at HTA agencies in Europe but financial resources might have difficulties explaining why Portugal does not have an HTA agency and Poland does [the gross domestic product (GDP) per capita of Portugal has been about twice – or, further back in time, more – than that of Poland], or why Hungary has one and the Czech Republic does not (at roughly similar levels of economic development). Certainly, money is a condition for establishing new agencies but definitely not a sufficient one and maybe even not even a necessary one.

Human resources could provide another explanation. If a country has to train 15–50 health economists and other experts (for the size of a Western European HTA agency, see table 5.1 in Garrido et al., Reference Garrido, Kristensen, Nielsen and Busse2008: 90) ex nihilo, it is probably not going to establish an agency for them. There is evidence, however, that countries without trained health economists first set up HTA agencies and then sought to fill in the lack of human resources: Poland set up a twinning program with French experts to train more specialists for its new HTA agency (Nizankowski and Wilk, Reference Nizankowski and Wilk2009), Croatia used pre-accession EU funds for similar goals (Mittermayer et al., Reference Mladovsky, Srivastava, Cylus, Karanikolos, Evetovits, Thomson and McKee2014). In Romania, HTA has been in the making – even before the International Monetary Fund’s and World Bank’s intervention – with the help of NICE International, as well as other HTA bodies from Western Europe supported by a World Bank project (Swiss Tropical and Public Health Institute, 2013). Lithuania benefitted from knowledge transfer from Swedish experts and has trained a number of health economists since the early 1990s (Danguole, Reference Danguole2009), and similarly, Slovenia is a leader of a EUnetHTA work package, which shows commitment to HTA as a field in the country, yet neither of the countries established a formal HTA agency to date. Lack of human resources, therefore, does not look like it has major explanatory power regarding HTA establishment (although financial and human resources may become hurdles when it comes to implementation of a diffused idea).

Centralized vs decentralized health system governance

There is recognition within studies on HTA that the degree of centralization of a country’s health care system influences the practicalities of HTA (Cappellaro et al., Reference Cappellaro, Fattore and Torbica2009; Drummond and Sorenson, Reference Drummond and Sorenson2009; Ciani et al., Reference Ciani, Tarricone and Torbica2012), although how centralization promotes or impedes the initial introduction of HTA is ‘open to debate’ (Sivalal, Reference Sivalal2009: 286). Neumann (Reference Neumann2009: S45) suggests that ‘the idea of establishing public HTA organizations has been a more natural fit with the more centralized, government-funded and administered health care systems of Europe’ as opposed to the decentralized and more privately funded system in the United States. This is far from a naïve assumption, given that health care reform in general is thought to be more likely to succeed in countries where the authority of the central government is strong compared with other actors in the sector (Immergut, Reference Immergut1992; Freeman, Reference Freeman2000). It also goes in line with Wild and Gibis’s (Reference Wild and Gibis2003: 188) argument that corporatist health care governance, notably the larger professional autonomy of physicians that accompanies it, ‘can be seen as an obstacle in the diffusion of HTA’.

A quick look at the distribution of EU countries between the mainstreamers and non-adopters suggests, however, that this hypothesis is not likely to be confirmed. Of course, operationalizing what constitutes ‘strong central decision-making’ is a tricky exercise in any given country; comparing power of institutions across all the 22 mainstreamers and non-adopters is even more problematic. There is, however, one critical case whose positioning on the strong–weak decision-making continuum raises little doubt: Germany. With its corporatist decision-making in the health sector, which typically involves the government on a par with health insurance funds and health professionals, Germany is a country where few would doubt the relatively weak role of the central decision-making power (in this case, the Federal Ministry of Health). By this logic, Germany should refuse to set up an independent HTA agency. In spite of this, IQWiG was established in 2004. When we look at other countries in the mainstreamers’ category, we can reasonably suppose that Germany is not an outlier: at the very least Austria and Belgium could be characterized as countries with weak central institutions, especially in comparison with some of the non-adopters. The fact that Germany created an active HTA agency suggests that a high degree of centralization of health policy decision-making is neither a sufficient, nor in fact a necessary condition to establishment of HTA agencies.

Bismarck vs Beveridge

The Bismarck vs Beveridge division constitutes another plausible theoretical explanation for the distribution of countries between the mainstreamers and non-adopters. Although it has been forcefully argued that the Bismarck and Beveridge distinction has been increasingly blurred to the point of becoming irrelevant (see for instance, Kutzin et al., Reference Kutzin, Cashin and Jakab2010; Schmid et al., Reference Schmid, Cacace, Gotze and Rothgang2010), it is still worth investigating whether the organization of a health system’s financing influences the establishment or not of an agency whose primary objective is to advise what to finance. However, despite the fact that the Bismarck vs Beveridge divide is widely used as a major criterion for case selection in health policy studies, literature exploring its relationship to HTA seems limited. Although Cavazza and Jommi (Reference Cavazza and Jommi2012) look at how a country’s health care system (in addition to broader governance traditions) affects the concrete work practices of its HTA agencies, only Wild and Gibis (Reference Wild and Gibis2003) draw a clear link between a country’s health care financing model and its willingness to engage in HTA activities and establish formal HTA institutions. According to them, Beveridge systems (in which care is financed by the state through taxation) are more likely to establish HTA agencies, whereas Bismarckian systems (where care is financed by third-party payers, typically insurance funds, based on employers’ and employees’ mandatory contributions) are likely to resist. This is because of four main reasons: the ‘unduly strong’ (Wild and Gibis, Reference Wild and Gibis2003: 189) role of physicians in Bismarckian corporatist decision-making, their resistance to ‘evidence-based medicine’, limited opportunities for funding of public health research such as HTA, and the lack of interest of the government in health policy tools, including HTA.

This hypothesis is to some extent confirmed by the empirical division between the forerunners and the mainstreamers, which was the only significant one when Wild and Gibis wrote their article (see Table 2). Sweden, Finland, Denmark, United Kingdom and Spain all had tax-based Beveridge health systems, whereas Bismarckian systems lagged behind with regard to establishing HTA agencies when the forerunners did. To explain further development, however, especially the division between mainstreamers and non-adopters, this hypothesis seems to be of little help. Positioning a country on one extreme of the Bismarck vs Beveridge divide is, once again, not so simple: Hungary, for instance, has a system based on work-related social contributions but with a single national insurance fund. Similar mixed elements of the two ideal-types can be found in most countries, especially in the post-Semashko post-communist ones. In fact, only Latvia would qualify as following a Beveridge model among post-communist countries. This means that once the forerunners established HTA agencies, there were relatively few Beveridge-type countries left in Europe to show a clear trend: Ireland, Italy and Latvia followed in the 2000s, whereas Portugal, Greece and Malta have not to this date. More importantly, this division based on the type of health care financing does not explain why some of the Bismarckian countries established HTA agencies and some did not.

Table 2

Bismarck vs Beveridge explanations for diffusion of health technology assessment agencies in Europe

Source: own compilation.

Beveridge health systems in bold italics.

^aOutlier, specific case – see description above for short discussion.

General trends in agencification

Leaving aside explanations specific to the health sector, it is also possible that the non-adopters have so far not established HTA agencies because they are reluctant to create independent agencies and other forms of QUANGOs in general. Surprisingly, little is known about the patterns of agencification in CEE in particular – most studies focus on the Western World (particularly EU-15 plus Australia and New Zealand: see for instance Pollitt et al., 2004; Christensen and Lægreid, 2007; Gilardi, 2008; Bouckaert et al., 2009; Verhoest et al., 2010). Thiel (Reference Thiel2011) concludes from an expert survey in Western European countries and Hungary, Estonia, Lithuania and Romania that CEE countries (with the exception of Hungary) have delegated more tasks to agencies than their Western European counterparts and created more agencies, although typically without legal independence. According to her conclusions, most agencies in these countries are semi-autonomous, directly depending on the ministry for budget, personnel, etc., but these patterns are in a way similar to the patterns of independence of agencies in some Western European countries. Studies of independent agencies in individual sectors, such as for instance anti-corruption or gender equality agencies (Batory, Reference Batory2012) suggest that some autonomous (or semi-autonomous) bodies exist in almost all CEE countries. Randma-Liiv et al. (Reference Randma-Liiv, Nakrošis and Hajnal2011) underline that although agencies as such existed during the communist regimes as well, there has been an increase in their numbers in some CEE countries after 1990, followed by a decrease in the mid-2000s, owing perhaps to global de-agencification tendencies (see Elston, Reference Elston2014).

In short, empirical accounts of agencification in CEE from political science scholarship, where most of the HTA non-adopters come from, do not point to any general unwillingness to create semi-autonomous public bodies. If we accept Randma-Liiv et al.’s arguments (based on a limited number of countries) about increasing de-agencification in the region as true, it might mean there the trend toward setting up new QUANGOs indeed passed or peaked around 2005 and countries which have not set up HTA agencies by that time will remain unlikely to establish them at all. Ireland and Latvia are here the two notable exceptions; the Romanian and Lithuanian case of creating a unit within the executive might be an illustration of a move away from agencies to a different set up, but the hypothesis of refusal to create HTA agencies as a consequence of resistance to agencification in general does not seem plausible.

Influence of international actors

In Europe of mid-2000s onward, the EU is clearly the most important international actor with the possibility to influence adoption of new policy trends, both for CEE and old EU member states. Until recently, the EU had very limited leverage to promote establishment of HTA agencies in countries that did not have them – despite its longstanding interest in HTA, more ambitious EU-level collaborations started at about roughly the same time that the mainstreamers were founding their institutions, with the EUnetHTA project launched in 2006. In October 2013, with the entry in force of the EU Cross-Border Healthcare Directive (Directive 2011/24/EU), a ‘voluntary network connecting national authorities or bodies responsible for health technology assessment’ was created, to which all EU member states, even those without HTA agencies, nominated representatives. These representatives were, however, often from ministries of health rather than HTA institutions (DG SANCO, 2014) – the Directive did not give the EU a mandate to require countries to establish HTA bodies if they did not have one. In other words, the influence of the EU was, before the Directive, limited to encouraging capacity building through socialization within EUnetHTA and occasional provision of funding (such as the Polish Twinning project) – and not so greatly enhanced after the Directive. Although a lack of means of coercion per se has not necessarily limited the EU’s influence in many policy areas, it can also be said that, despite its interest, HTA has not been high enough on the priority list for the European Commission (Author reference 1) – certainly not enough to warrant the EU promoting a particular organizational model.

Similar conclusions can be drawn for the World Health Organization (WHO), which also focused on awareness raising, socialization and capacity building (Garrido et al., Reference Garrido, Kristensen, Nielsen and Busse2008; World Health Organization, 2014). In contrast, the International Monetary Fund and the World Bank have had at their disposal hard conditionality linked to financial incentives such as international loans. The Romanian case mentioned above suggests that this coercive power might indeed lead to establishment of HTA institutions (although not necessarily an agency in the forerunners’ or mainstreamers’ sense) – but it is also exceptional, depending on the macro-economic and political context of a country with deep economic problems, giving the international actors extraordinary leverage.

These circumstances were not present for most other countries (or if they were, such as in Latvia, they did not necessarily lead to a strengthening of HTA, much less so in an agency form). At the time when the mainstreamers founded their HTA agencies, international actors were only starting to consolidate their interest in the field – in this sense, the forerunners can be conceptualized as uploaders of their policy innovation to the international level. However, the interplay between the international and national levels becomes more complex past the mid-2000s, with no clear top-down influence of the EU or WHO over either the mainstreamers or the non-adopters. In any case, although international actors are likely to be influential in promoting awareness of and capacity in HTA as a field, they do not push for an agency model that we observe across the mainstreamers but not the non-adopters.

To sum up, none of the macro explanations taken from the ‘usual suspects’ in the health sector or policy sciences explains sufficiently why concrete countries did not set up their HTA agencies and why most of those countries are in Eastern Europe. A brief reflection of their potential conjunctural causation is not more promising. Given the unclear role of international actors and the general attitude to agencification, an obvious combination of variables to explore further would be resources coupled with the centralization of health care system or the Bismarck/Beveridge divide, or both. A cursory consideration of these combinations, however, still leaves many cases unexplained (for instance, a rich, Bismarckian Germany with its weak central decision-maker contrasted with a not-so-rich, Bismarckian Hungary with a strong center, etc.).

Attention should therefore be given to actor-centered explanations as suggested by policy diffusion scholars: interest group politics or the variety of public policy meso-theories such as policy communities, iron triangles, policy networks, advocacy coalitions and the like, should be better equipped to explain the distribution of HTA agencies in Europe. The remainder of this text develops a theoretical model around domestic epistemic communities – a group of actors particularly relevant for the diffusion of HTA.