General characteristics of DRG systems

DRG systems have been adopted by and evolved in countries differently. For example, half (50%) of the countries surveyed developed their DRG system locally, while the remainder imported or adapted it from elsewhere. Among countries that reported they adopt aspects of their DRG system from other countries, it is most common to import only the grouper, although several countries (Ireland, Italy, Norway and Portugal) import both DRG groupers and codes. In addition, each country has taken a different route and at different times to introduce a DRG-based payment system. For example, in England and France they were developed initially for the purpose of patient classification or to describe hospital activity, and later also for payment purposes (England, France, Germany, Netherlands, and United States). Other countries such as Ireland, Portugal, Spain (certain regions) and Sweden use DRGs principally for budget allocation or benchmarking purposes. Although the majority of countries define DRG cost-weight standards at the national level, the similarities end there. Countries differ with regard to the costing methods used to define tariffs and for those jurisdictions that define tariffs at the regional level (e.g. Italy, Spain, Sweden), variation exists within the country. Such differences are often the result of dissimilarities in the number of hospitals that participate in collecting standardized cost-accounting information and the level of detail required to meet national cost-accounting standards (Schreyogg et al., Reference Schreyogg, Stargardt, Tiemann and Busse2006; Busse et al., Reference Busse, Geissler, Quentin and Wiley2011).

Frequency of updates

Updates to the patient classification system or the payment tariff are not exclusively targeted at incorporating technological innovation into DRG-based hospital payment systems; rather, they are intended to ensure that the systems are continuously modified to reflect current practice patterns and treatment costs. Our survey suggests that the majority (71%) of countries update on an annual basis, although there are exceptions. Italy, for example, generally updates both the patient classification system (at the national level) and payment tariffs (at the regional level) every two to three years, although the latter can vary among regions. While Austria updates its patient classification system annually, the payment rates are adjusted only for specific DRGs when deemed necessary by policy makers. For countries that import their DRG system from other countries, the timing of updates is normally tied to that of the originator country. For instance, Ireland updates its patient classification system and tariffs every four years and annually, respectively, which informs updates to the Australian system. Given that updates are based on available data, several respondents suggested that DRG-based hospital payment systems exhibit a certain time lag, in terms of reflecting current practice patterns and treatment costs.

Use of short- and long-term mechanisms to support technology adoption

Concern has been expressed that DRG-based payment systems may pose particular disincentives for expensive breakthrough medical technologies, particularly those for which patient benefits are realized over a period of months or years and for which limited evidence exists to sufficiently inform either the development or updating of the system. As illustrated in Table 1, most of the countries (79%) surveyed use some form of short-term payment for specific technologies, which generally entails some type of additional funding to hospitals using certain medical technologies. Only Austria, Canada and Norway do not offer such payments. A majority (64%) of countries offer either separate or supplementary payments for new technologies, with half of those surveyed extending both types of short-term payments. Only Italy and Sweden provide cost-outlier funding.

Table 1

Overview of the use of short-term payment instruments in 14 countries

Sources: Survey data; Scheller-Kreinsen et al. (Reference Scheller-Kreinsen, Quentin and Busse2011).

Technologies eligible for the aforementioned payments typically meet one or more the following criteria: high costs and/or costs exceeding standard deviation of the respective DRG(s); technology used in defined specialist services/interventions or diagnoses; level of activity associated with technology is found in a number of DRGs (i.e. used in a sufficient number of patients); and a new innovation. Some countries also consider whether the technology provides meaningful clinical improvement or is more cost-effective compared with existing services, which we discuss further below.

Almost all countries that employ short-term payment arrangements indicated that they are determined through standardized processes, which usually involve some form of application and negotiation process, especially with regard to separate payments. Negotiations around the specifics of the short-term payment (e.g. payment amount, term of payment) typically involve individual hospitals, insurance companies and national bodies, such as Ministries of Health and those involved in setting DRG payments, or regional authorities in decentralized countries (e.g. Italy, Spain). The process for determining the payment amount is by and large informal, via negotiation and agreement between the relevant parties involved. The United States is one exception, where short-term payments (called ‘add-on payments’ for inpatient payments and ‘pass-through payments’ for outpatient payments) are paid based on pre-determined criteria. Add-on payments equal to the lesser of 50% of the amount by which the total covered costs of the case exceed the DRG payment, capped at 50% of the costs of the new technology (Clyde et al., Reference Clyde, Bockstedt, Farkas and Jackson2008). Pass-through payments are equal up to 100% of reported costs minus device costs already built into the base payment rate. This formula essentially requires Medicare and hospitals to share in the financial risk of providing costly new technologies. Survey respondents indicated that it generally takes between several months and one year or more to negotiate and finalize short-term payments.

In most cases where short-term payments are used, they are reviewed on an annual basis, during scheduled updates, to assess if the technology still meets the eligibility criteria or whether it can be adequately incorporated into an existing or new DRG. However, some countries, such as France and the United States, review less frequently, typically every two to three years. French respondents noted that growing funding pressures may result in more frequent reviews of its lists of reimbursed products (‘LPPR’) and supplementary payments; however, there has been some consensus that public authorities should try to ensure some degree of reimbursement predictability or stability for technology manufacturers. In Germany, the agreed short-term (‘NUB’) payment between hospitals and insurance companies applies only for one year, after which hospitals need to reapply if conditions warrant additional, separate reimbursement (Henschke et al., Reference Henschke, Baumler, Weid, Gaskins and Busse2010). England also employs a time limit for ‘innovation payments’; payments can only be agreed for three years. In all other cases, short-term payments cease when a technology is no longer eligible or it is included in the DRG system. For example, in Australia, short-term payments are phased out, with payments absorbed into DRG payments after (1) the full cost of the technology (e.g. a device) is accurately included in the relevant DRG and (2) the technology is well assessed and integrated into clinical practice. Table 2 compares and contrasts different national approaches to short-term separate and supplementary payments.

Table 2

Comparison of different national approaches to short-term separate and supplementary payments

DRG=diagnosis-related group; HRG=healthcare resource group.

Sources: Survey data; Medpac (2003); Department of Health (2011); Scheller-Kreinsen et al. (Reference Scheller-Kreinsen, Quentin and Busse2011); and Henschke et al. (Reference Henschke, Baumler, Weid, Gaskins and Busse2010).

In terms of long-term instruments, a minority (<35%) of countries have created a new DRG outside of a general update in response to innovative technology. In cases where this has occurred, it is because the technology is used at a high volume and/or has a significant cost impact. The example of coronary stents was noted by respondents in Australia, Austria, Canada, Germany and the United States. This has also been applied to date to transcatheter aortic-valve implantation (TAVI) in France and intracranial neurostimulators in Portugal. Otherwise, it was fairly common for the payment tariff to be recalculated when a given technology increased the cost of all services bundled in a DRG or of all hospital services, but this was usually done during the normal updating process, as previously discussed.

Table 3 presents examples of technologies, particularly medical devices, where countries have applied short- and long-term payment mechanisms. For example, in Italy short-term instruments – both separate and supplementary payments – are used in various regions for transcatheter aortic-valve implantation (TAVI). The Lombardy and Basilicata regions employ supplementary payments, although with different characteristics. Lombardy reimburses 95% of the average price paid by regional hospitals (∼€20,000), while Basilicata pays an additional €8000 over the relevant DRG tariff. In contrast, Campania and Puglia regions use separate payments to cover both the cost of hospitalization and the device itself (about €23,000 together), instead of reimbursing TAVI through the DRG system.

Table 3

Examples of short- and long-term payment instruments applied to devices

TAVI=transcatheter aortic-valve implantation; DRG=diagnosis-related group.

Sources: Survey data.

^a Either a separate or supplementary payment was made.

^b Principally relates to creating a new DRG code.

Role of evidence

With increasingly limited health care budgets, policy makers have sought to obtain greater value from investments in new technologies and other interventions. However, there has been limited evidence of whether DRG reimbursement tariffs for new technology are sufficiently linked to evidence of value, namely data on a technology’s therapeutic benefit and cost-effectiveness.

Overall, in practice it appears that evidence is only used in a minority (43%) of countries surveyed, including Australia, England, France, Finland and Sweden to support determinations of or updates to DRG payments (Table 4). Of those jurisdictions that use evidence, Finland and France consider data on therapeutic value only, while Australia only considers costs. The remaining countries (England and Sweden) attempt to consider both therapeutic benefit and costs.

Table 4

Role of evidence in determining DRG tariffs and short-term payments for new technologies

DRG=diagnosis-related group.

Sources: Survey data.

^a Eligibility and terms of short-term payments negotiated at the local level, so consideration of therapeutic benefits and costs may vary across hospital district, region, or county and technology.

It is slightly more common in many countries for the available evidence on the technology to play a role in determining short-term payment arrangements, in particular separate or supplementary payments. The majority (64%) of those countries using short-term payments consider evidence when negotiating or determining payments (Australia, England, France, Germany, Italy, Portugal, Spain, Sweden, and United States). However, in England, Italy, Spain and Sweden, short-term payments are negotiated locally, typically at the hospital level, so the role and use of evidence may differ across hospital and technology. In a few countries (Australia, England and France), additional evidence collection is also required as part of the short-term payment agreement.

Since we were interested in ways to strengthen the link between evidence of value and DRG-based hospital payments, we surveyed respondents on their perceptions of the key challenges in using evidence to determine payments. The main challenges identified included a lack of clear and standard processes for collecting evidence on technologies and subsequently how to obtain access to available data and apply it when determining DRG codes and tariffs. In addition, respondents noted that a lack of clinical data for new medical technologies can hinder evidence-based payment decisions. Analysts have noted several challenges in ensuring a robust evidence base, in particular for devices (Drummond et al., Reference Drummond, Griffin and Tarricone2009; Sorenson et al., Reference Sorenson, Tarricone, Siebert and Drummond2011). For instance, the relative lack of controlled and longer-term clinical studies [e.g. randomized controlled trials (RCTs)] is a particular problem and one that is most prevalent for devices compared with other technologies. RCTs for devices are often small, because the initial patient population is not as large as many target populations for drug treatment. Moreover, for many devices, uptake into clinical practice occurs rapidly and, at times, at a point much earlier than the completion (or initiation) of clinical studies. Of particular importance to hospitals, available evidence on the cost-effectiveness of devices also often overlooks the broader organizational implications (e.g. surgeon/end-user training, associated capital investments) of adoption.

Impact of hospital payments on the adoption of new technologies

As highlighted in our previous discussion, the evidence on the impact of DRG-based hospital payment systems is mixed, with some studies suggesting they serve to support technology adoption, while others proclaim a hindering or neutral effect. When queried about the impact of DRG systems on technology uptake, the majority (70%) of respondents believed such systems do indeed have an impact on adoption. Of these, 87% indicated that DRG systems can act both as an incentive or deterrent to the uptake of new technologies, depending on the circumstances. For example, one Australian respondent indicated that the design of the patient classification system can have a marked influence on the integration of a technology into the system, where the use of broad categories enables hospitals to compensate cost increases (related to the technology) in one area and quickly adopt a technology, while the opposite is true for systems that define DRG narrowly. Others emphasized the importance of setting the tariff at a level appropriate to reflect the costs of new technologies, which would provide the necessary incentives to adopt such technologies into practice. Very few (8%) considered DRG-based hospital payments a deterrent under all circumstances. In these cases, the main deterrents identified were the time lag between introduction of a new technology and integration into the DRG system and the fact that evidence of value for a particular technology is not – in most systems – formally linked to the payment system. Any disincentives for adoption were seen to predominately affect cost-increasing, quality-enhancing technologies, at least temporarily until they could be appropriately reflected in the existing DRG system.

All of the respondents that use short-term payments in their respective country believed that they could be employed to address some of these issues. However, several respondents highlighted that their effectiveness in facilitating the adoption of new technologies was contingent on successful negotiation of the terms of the payment arrangement, especially with regard to separate payments, and some limitations placed on the duration of the payments.