>> This situation is accompanied, on the one hand, by a decades-long investment backlog due to insufficient investment funds from the federal states, which is naturally exacerbated by existing overcapacity. The use of investment funds for hospitals that are not actually needed diverts these resources from the hospitals that are actually needed to provide high-quality patient care. On the other hand, the federal states' hospital planning, which is actually intended to lead to a needs-based supply of the population with efficient and economically independent hospitals, often amounts to merely continuing existing hospital structures or confirming changes that have already been made. Quality as a further criterion for hospital planning decisions was only incorporated into hospital legislation very late, but is hardly reflected in concrete planning.
Many small and tiny hospitals – often with poor equipment
In addition, there are a large number of small hospitals in Germany: In 2017, 666 of all German hospitals, or around one-third, had a maximum of 100 beds. A further 23% had between 100 and 200 beds (Federal Statistical Office 2018). In total, just under 57% of all German hospitals had fewer than 200 beds. This large number of very small and largely non-specialized hospitals contributes – it is often argued – to good accessibility of hospital care overall. However, on closer inspection, this turns out to be problematic because it simultaneously has a negative impact on the quality of treatment in two respects: Even patients with a necessary inpatient indication are often cared for in such small hospitals in Germany today. Although these hospitals are easily accessible, the majority of which – with a substantial proportion of their patients – are neither adequately equipped in terms of medical technology nor staffing. In 2017, 341,000 of the planned hospitals did not have a CT scanner, and 611,000 did not have coronary angiography (based on 1,329 general planned hospitals). This must be considered problematic, especially with regard to emergency care.
The consequence of inadequate staffing is that in these hospitals, patient care by specialists is often significantly delayed because only junior doctors are on call at night and on weekends, while specialists on call must first be summoned to the hospital. The supposed advantage of easy accessibility thus turns into a serious quality disadvantage, which could be avoided, in particular, by greater concentration and specialization in hospital care.
The fact that patients are often treated in hospitals that are neither adequately staffed nor technically equipped is also reflected in the high number of transfers: According to hospital statistics, a total of around 770,000 patients were transferred to other hospitals in 2017 (Federal Statistical Office 2018).
The reasons outlined above contribute to the paradox of staffing in Germany: There are more staff per capita in Germany than in other countries, but fewer per patient or bed-day (due to the high number of cases/bed-days). The nursing staff shortage in hospitals, currently the dominant health policy issue, thus represents a (late) indicator of ruinous hospital competition based primarily on economic incentives, ultimately impairing the quality of care.
High levels of overcapacity mean that German hospitals are often treating patients who do not require inpatient treatment. This is evident, among other things, in the fact that an increasing number of patients (now around 50% of all inpatients) are being admitted via emergency departments. This becomes clear over time: while in 2005, 60% of all inpatients (9.6 million) were admitted, in 2016 this figure fell to just 47% (8.9 million). At the same time, the proportion of patients admitted via emergency departments has increased from 34% (5.4 million in 2005) to 45% (8.6 million in 2016). In addition, case growth in these hospitals is particularly high for indications that are sensitive to outpatient care, even compared to neighboring European countries.
Perverse incentives in the G-DRG reimbursement system
Overcapacities in the hospital sector are not being reduced; instead, they are often filled with mild cases or cases that do not require inpatient treatment at all. This is also due to perverse incentives in the current G-DRG reimbursement system. It is undisputed that the reimbursement system contains significant incentives to expand the caseload, to split cases, and to perform non-indicated procedures, which – despite numerous detailed reforms – have still not been completely eliminated. This is evident, among other things, in the high number of hourly and daily cases (almost 700,000 cases in 2016). At the same time, the disproportionate profit prospects for certain services in the current DRG system – measured against average costs – such as technical procedures and joint replacement and back surgery, represents a clear incentive to establish "pseudo-centers" in hospitals that are not actually suitable for these services in terms of quality and equipment. Hospital operators repeatedly respond to financial difficulties in hospitals – especially smaller ones – with such measures in order to avoid their closure. Fundamentally, the case-based DRG system, which operates with average prices that are largely independent of the hospital and can only be influenced to a very limited extent by the individual hospital, means that the hospital can only influence its economic success or failure through its cost side and its service portfolio, i.e., the type and volume of cases treated.
In summary, the high hospital and bed density in Germany, together with the high proportion of very small and very small hospitals with often inadequate medical equipment and staffing, leads to high and increasing case numbers and three interrelated problems:
(1) an inappropriate number of patients,
(2) who are often treated in hospitals that are not adequately staffed and technically equipped, whereby
(3) the staff-to-patient ratio or per bed-day ratio is on average rather low compared to neighbouring European countries.
High outpatient potential
A large portion of the overcapacity in inpatient care is avoidable. This is evident from a glance at the outpatient treatment potential, i.e., the cases that could be treated or operated on an outpatient basis. Busse and Berger show in their 2018 Hospital Report (Berger/Busse 2018) that in selected indication areas, this amounts to up to 35%.
of all hospital cases. In total, this amounts to around 5 million hospital cases annually, or 14,000 hospital cases per day. This alone corresponds to the patient volume of around 500 average hospitals, i.e., already one-third of the capacity of German general hospitals.
The methodological approach of "outpatient-sensitive hospital cases" (ASK) leads to similar results with regard to the extent of outpatient potential (cases that can be treated on an outpatient basis). Sundmacher et al. (Sundmacher/Schüttig/Faist 2015) published a list of relevant outpatient-sensitive medical conditions that were agreed upon by physicians in Germany as part of a Delphi study. The corresponding medical conditions are divided into 40 diagnostic groups, 22 of which form a core selection, covering 90% of the hospital cases recorded in the 40 groups and for which a preventability of over 85% was estimated within the study. Considering the proportion of preventable ASK for the 14 main diagnoses with the highest number of cases in 2016, this results in a proportion of preventable cases (ASK) of just under 38%. Overall, the application of the ASK approach results in around 4 million avoidable hospital cases.
Large differences in quality
In addition, there are currently significant quality differences in Germany, even for cases that are clearly eligible for inpatient treatment, depending on the number of cases treated (volume-outcome relationship), as recent studies clearly demonstrate (especially Nimptsch/Mansky 2017). For most of the treatment types examined, a clear correlation was observed between the annual caseload of the treating hospital and in-hospital mortality. Furthermore, Nimptsch and Mansky's calculations showed that centralization measures would contribute to quality improvements and, in particular, to reducing preventable deaths (more than 1,000 preventable deaths in acute myocardial infarction and chronic obstructive pulmonary disease, respectively).
Further studies show that the volume-outcome relationship can be demonstrated not only for in-hospital mortality but also for the frequency of reoperations for a number of treatment types (Pross/Busse/Geissler 2017). Overall, this shows that a significantly greater concentration and specialization of the German hospital landscape leads to a significant improvement in the quality of inpatient care – at least in terms of mortality and reoperations.
Against this backdrop, the Bertelsmann Foundation aims to contribute to the discussion on how the changes to hospital structures in Germany, which are considered necessary, can be achieved with this project. The first phase of the project served to develop a target vision for the changes to German hospital structures considered sensible and desirable. The aim was to identify and structure internationally and nationally discussed requirements and model approaches for a needs-based and resource-efficient hospital structure. Given the current situation described above, the focus was on approaches aimed at reducing excess capacity and increasing the concentration and specialization of hospitals.
This target image is described below.
The starting point is the current volume of hospital care in Germany, comprising 1,329 general hospitals with 398,000 beds, 35 university hospitals with approximately 45,000 beds, and 285 psychiatric hospitals with approximately 47,000 beds; a total of approximately 1,650 hospitals with >490,000 beds and 19.5 million cases (as of 2017/Federal Statistical Office 2018). The considerations for the target vision are based on full consideration of the aforementioned outpatient potential, the exclusion of non-indicated procedures and cases, and the systematic consideration of qualitative criteria. A further prerequisite is that, following such a restructuring of the German hospital landscape, safe, high-quality care is available 24 hours a day, 7 days a week (24/7) in every region.
Target image for the future German
Hospital landscape
The overcapacities in the German hospital sector and
The directly associated, ruinous, ultimately impairing quality of care, primarily based on economic incentives, and the perverse incentives in the current G-DRG reimbursement system, as well as the perverse incentives in the current G-DRG reimbursement system, lead, as shown, to the treatment of a significant proportion of patients in hospitals even though these patients do not require inpatient treatment. Furthermore, a lack of concentration and specialization leads, among other things, to insufficient technical equipment and low staffing levels per case or bed-day. All of this results in significant quality losses, which can be remedied by reducing capacity, concentrating, and specializing care in fewer, but significantly larger hospitals or locations. These "new standard care" hospitals are then in a position to guarantee the medical treatment quality required today, even in emergencies, across the breadth of the specialist departments represented there, with adequate medical-technical equipment and 24/7 staffing, including specialist physicians.
This inevitably results in a departure from the current three- and sometimes four-tiered German hospital care model (basic care, standard care, supra-local specialized care, and maximum care, usually provided by university hospitals; specialist clinics are also permitted), which often results in very small to very small hospitals, particularly in the area of basic and standard care. The future hospital structure developed and presented here instead envisages a care system with "new standard care" (with the option of treating heart attacks, strokes, etc.) in medium-sized centers and "maximum care" in large and regional centers – particularly for quality reasons, which are explained in detail below.
We believe that around 360 hospitals (new standard care and specialty care) and around 50 maximum care providers are sufficient to provide efficient care for the population in Germany (at the end of 2017, Germany had around 82.7 million inhabitants), as presented in the target scenario (see table “Target scenario for the future German hospital landscape: initial situation and target scenario”). Each of these 410 hospitals will therefore have to provide standard care for around 200,000 inhabitants2; within the maximum care framework, one maximum care provider will be responsible for around 1.65 million inhabitants. In particular, the “new standard care” will be provided by hospitals that are significantly larger and better staffed and of better quality than today’s typical German district hospitals providing basic and standard care. This is particularly evident in the quality requirements presented below.
Not all cases that can be treated as outpatients will necessarily have to be, or can be, cared for within the current outpatient medical care structure. Instead, the establishment of outpatient surgery centers at future "new standard care" hospitals should be considered – particularly for outpatient operations. The same applies to the conversion of small, current general hospitals that no longer function as acute care hospitals into multidisciplinary "intermediate/integrated treatment and care centers." These could, for example, combine the functions of current medical care centers and outpatient surgery centers and also take over post-hospital care and observation in cases where inpatient treatment is suspected. In addition, the option of highly specialized outpatient treatment in hospitals should be significantly simplified and expanded.
Qualitative and structural requirements for the future German hospital landscape
In a sustainable hospital landscape, several key qualitative criteria must be met. These criteria include, in particular:
• Better technical equipment: The ability of all hospitals in the "new standard care" to provide comprehensive acute treatment for heart attacks and strokes. Heart attacks and strokes are the main life-threatening conditions that require treatment in hospitals. At the same time, the equipment required in hospitals is of key importance, such as medical technology for comprehensive diagnostics. In the case of (ischemic) stroke, the existence of a stroke unit is also necessary for effective treatment either with thrombolysis or, in suitable cases, with thrombectomy. Primary percutaneous coronary intervention (PCI) is now the treatment of choice for heart attacks. For this, a cardiac catheterization laboratory is also required in addition to comprehensive diagnostic equipment with appropriate imaging procedures.
• More and higher minimum quantities: The fulfillment of internationally scientifically recognized minimum quantities, in particular for the services of transplants, care of premature and newborn babies, births, knee endoprosthetics, hip endoprosthetics, back surgery as well as cancer surgery and coronary surgical interventions, by all hospitals in the new care system outlined here.

The concept of minimum quantities is based on the idea that the quality of the treatment outcome is directly related to the quantity of services provided and that there is therefore a relationship between the probability of treatment success and the routine provision of services by the key players involved in the treatment process. Minimum quantities are now recognized in international scientific literature for a wide range of medical services. In Germany, however, minimum quantities have so far been used only very hesitantly. Internationally, there are prescribed minimum quantities for many more indications; moreover, the minimum quantities used are often significantly higher internationally than currently in Germany. The increased use of this quality instrument, based on internationally recognized minimum quantities, will lead to significantly improved quality of care in the affected indications, but also requires a corresponding concentration of the affected services in significantly fewer hospitals.
Fulfilling the qualitative criteria mentioned here precludes future provision of care in (small) primary and standard care hospitals, which typically only have internal medicine, surgery, and, in some cases, obstetrics departments. Because such mostly small hospitals cannot guarantee either acute care for heart attacks and strokes or compliance with minimum volumes. Furthermore, if "genuine hospital cases" (such as heart attacks or strokes) disappear, given the current framework for inpatient care in Germany, there is a very high risk that small hospitals will fill their beds with cases that must be considered outpatients – a behavior that is already evident in many places under current conditions.
In addition, the hospitals participating in the care in the future should meet a number of structural requirements:
• Accessibility for hospitals of the “New Standard Care” for the population to be cared for: usually max. 30 minutes3
• Accessibility for maximum care providers: max. 60 minutes
• In all departments and specialties central to emergency care at all levels of care, a 24-hour availability of specialists directly in the hospital
• Minimum equipment of the “New Standard Care” hospitals with all specialist departments or the corresponding interdisciplinary medical centers required for care at this level, and in particular emergency care.

Based on the above specifications, the target scenario, taking into account ASK cases (approximately 4 million), transfers (just under 0.8 million), and hourly and daily cases (approximately 0.7 million), results in a caseload of only around 14 million. This is a rather conservative estimate of the proportion of reducible hospital cases, as several other potential reductions in the caseload are not taken into account (such as hospital cases that do not require treatment, surgeries, procedures, and inpatient cancer treatments that can be performed on an outpatient basis and are not covered by the ASK approach, or cases resulting from medically non-indicated case splitting).
The reduced caseload of 14 million results in an average caseload of approximately 30,000 per year for general hospitals and just over 60,000 per year for maximum care hospitals. As already explained above, this scale also guarantees the necessary staffing for all levels of care to ensure the round-the-clock staffing of specialists in the areas central to emergencies, which is essential for quality, particularly for emergency care, since the current clinical staff of all hospitals will be retained and concentrated. At the same time, the caseloads achievable in this way not only ensure good medical quality – among other things, by meeting minimum volume requirements – but also enable sustainable, economic management.
Table 1 provides an overview of the quantity structure of the initial situation and the target scenario.
International comparison
The fact that the German hospital system has considerable overcapacity is also supported by a comparison with the European average: Compared to the EU-15 average, Germany has 65% more beds. This drastic overcapacity translates into 50% more cases and 70% more bed-days per capita in Germany compared to the EU-15 average. The capacity reduction depicted in this target scenario would mean that Germany would correspond to today's European average in 2030.
We assume that the initiated structural change would definitely result in a comprehensive outpatientization of medically necessary but still feasible surgeries, cancer therapies, etc. The extent of current German hospitalizations can also be illustrated by international comparisons, particularly at the European level (EU15). This shows that lower hospitalization rates in other countries are largely due to a more extensive exploitation of outpatient potential (e.g., larger proportions of outpatient surgeries), without any loss of quality.
Based on the currently best European nations (especially Denmark) – measured primarily in terms of efficiency and length of stay – an even more radical scenario would emerge compared to the target scenario presented above, in which perhaps 300 care hospitals, each with around 500 beds and 45 maxi-
mal care providers with around 1,100 beds each would be required. As the example of Denmark in particular shows, such a radical reduction in inpatient hospital capacity, combined with simultaneous specialization and concentration, can lead to a demonstrable improvement in the quality of inpatient treatment.

The authors would like to thank the Review Board (Prof. Dr. Boris Augurzky, RWI – Rheinisch-Westfälisches Institut für Wirtschaftsforschung; Prof. Dr. Andreas Beivers, Fresenius University of Applied Sciences for Business and Media GmbH; Prof. Dr. Max Geraedts, Philipps University of Marburg; Dr. Matthias Gruhl, Authority for Health and Consumer Protection of the Free and Hanseatic City of Hamburg (until November 2, 2018); Prof. Dr. Bernt-Peter Robra, University of Magdeburg; Prof. Dr. Jonas Schreyögg, Hamburg Center for Health Economics) as well as Dr. Stefan Loos, Dr. Martin Albrecht and Karsten Zich from the IGES Institute Berlin for their cooperation in the project and in creating the target image.