Why we need to move beyond diagnosis-related groups and how we might do so

Introduction

Diagnosis-related groups (DRGs) are one of the core elements of New Public Management in hospitals across European health care systems. DRGs have become widespread in Europe since the early 2000s^1,2 and emerged as a key instrument to improve the activity and efficiency of services. This is said to result from introducing more transparent ways of paying hospitals and measuring hospital performance.^3–5 However, across the different health care systems in Europe, there are many examples that DRGs have adverse effects.^1,3,6–9 There is a danger that the unintended consequences of DRGs undermine the goals of health care systems. It is high time to take action. European countries need to fundamentally rethink hospital governance and consider new systems of budgeting and performance measurement. Central Denmark Region is currently piloting a system, based on the Triple Aim approach, ¹⁰ which may offer inspiration.

Programme theory of DRGs

DRGs were developed in the United States in the 1980s in response to significant increases in hospital spending. ³ The situation was similar in Europe some years later.^3,5 Although there are variations in the form and substance of DRGs across countries, the overall aim of implementing DRGs has been to increase the efficiency of hospitals. DRGs are a patient classification system that combines clinical and resource utilization needs, permitting reimbursement payments to be calculated. DRGs are also meant to provide a means for hospital managers and policy makers to monitor and control the use of health services.^1,2

Adverse effects of DRGs

DRGs were a response to the negative consequences of systems of hospital payment based on prospective global budgets and fee for service. Although DRGs have addressed some problems, such as waiting lists, they are associated with a new set of adverse effects. While a study in England found no such effects, ¹¹ there are many examples in other European countries of DRGs having adverse effects which are difficult to control. Adverse effects take three forms.

First, there are adverse effects that relate to the very goals DRGs are meant to pursue. A review of Scandinavian studies concluded that in terms of efficiency, only half of the results were positive. ¹² DRGs are based on a fixed value in relation to the specific patient group treated and the specific service provided. Although this includes different grades of severity, the system has difficulties accommodating the complexity of diagnosis and treatment. This gives hospitals an incentive to assign patients a higher grade of severity than is actually the case.^1,3,7 Service providers then have an incentive to maximize the income of hospitals. Many DRGs systems try to address this problem by including additional factors. For example, NordDRG in Nordic countries and Estonia ³ also classifies patients according to sex and age. However, this makes the DRG system more complex and jeopardizes its use for comparing the performance of hospitals nationally. ¹³ This is further challenged as the same treatment may be provided in different ways across hospitals and countries. ³

Second, DRGs may adversely affect the equity of the health care system. DRGs divide patients into discrete groups of average diagnoses and treatments. However, the needs of vulnerable patients often exceed the funding provided by a specific DRG value.^3,7,8 Hospitals thus have an incentive to give preference to patients who fall into less severe categories. This has adverse effects on the equality of access to hospital care, although the magnitude of the effect depends on the degree of competition among hospitals. Adjusting DRGs in different ways has helped somewhat. For example, the NordDRG system also includes the length of stay and the status at discharge. ³ While this reduces some of the variation in a particular DRG, the system still fails those patients who are already on the margins of the health care system and society.

The final set of adverse effects relates to the quality and effectiveness of health care. Although DRGs say little about the quality of care and health outcomes,^3,9 they may have an impact on both. A study of 35 OECD countries showed that DRGs slow down quality gains. ⁶ In France, competition for patients became skewed as DRGs did not cover the same cost items in public and private hospitals. Together with a lack of transparency, the DRG system offers an incentive to misuse the classification of patient cases.^3,8 Hospitals provide treatment as day cases rather than outpatient consultations, and the focus is on maximizing earnings rather than offering the most effective and highest quality of treatment. Similarly, across Europe, research shows that the quality of patient care is a compromise between financial considerations of hospitals and concerns for quality.^1,3,8

A new model for hospital governance

The adverse effects of DRGs seem to be difficult to control. There is a need to take action and to move beyond DRGs. However, to date only a few examples of alternative models of governing hospitals have been implemented. ³ A pilot currently being conducted in Central Denmark Region potentially offers inspiration for political and administrative decision makers. It is running for two years (2014–2015) and includes nine clinical departments covering inpatient care, day surgery and emergency care.

This new model of hospital governance is based on the Triple Aim approach which focuses on patient-experienced quality, health outcomes and cost containment. ¹⁰ Compared to DRGs, the model has three distinct characteristics: it separates the payment of hospitals from the management of performance; payment is based on a prospective global budget for individual departments; and the measurement of performance is based on locally defined goals.

In contrast to DRGs, the prospective global budgets in the new model abolish the financial incentive for hospital departments to misclassify patients and treatments; the model simply does without any classifications at all. The budgets of the individual departments reflect their current level of expenditure to minimize the differences among departments. Global budgets mark a fundamental shift in focus from financial considerations to the quality of health care. They give room for health care professionals to apply their professional knowledge and skills to ensure the best quality of care for individual patients.

The effect of global budgets at the level of individual departments is closely tied to the existence of goals for the measurement of performance of departments. The goals are local variations of the Triple Aim approach. The three aims are not independent, and a change in one can affect the other two. ¹⁰ The health care professionals in each department are responsible for balancing the three aims and for formulating goals. This contrasts with DRGs, which are a centralized tool of performance management by and for managers. Instead, the local goals are decentralized, designed and used by health care professionals. Studies show that including health care professionals in management has a positive effect on services. ¹⁴

Lessons for moving beyond DRGs

What can we learn from the pilot in Central Denmark Region? Although DRG systems have been introduced to increase productivity, the suspension of the DRGs can be a springboard to produce more quality from fewer resources. The Triple Aim approach suggests doing this by customizing services to the needs of individual patients. It adopts a partnership practice to secure both trust in hospitals, optimal treatment and effective use of resources. ¹⁵

Nevertheless, the new model of governance is not without its caveats. These need to be addressed so that the new model does not have the same adverse effects as its predecessors. It is possible that the model gives too much influence to health care professionals, whose paternalism waters down the focus on patient-centred care. Hospital departments may also doubt the true intentions of political and administrative leaders when it comes to abandoning DRGs and granting autonomy. Finally, the conditions for successful implementation remain unclear: how important is the quality of local leaders in hospital departments? To what extent is it necessary to protect departments from other, potentially competing influences?