Health care research is absolutely necessary from a medical and regulatory perspective

Many scientific innovations are developed in the field of basic research and then, based on the results of in-vitro research, are so solidly processed through small animal experiments that the path to applying these processes to humans is cleared. Further clinical research deals with measuring the effect of the therapeutic procedure and the side effects and at the same time lays the foundation for the approval procedures of the relevant authorities in North America and Europe. In order to be able to measure the effectiveness of therapeutic procedures, special patient groups are often put together in order to measure the effect of the new procedure/medication on a placebo procedure or on other, often older, medications or medical procedures/procedures in prospective randomized studies. A fundamental bias in these studies is often a strictly selected patient population that only slightly corresponds to the overall “real world” population addressed under field conditions.

>> These strict inclusion and exclusion criteria are intended to ensure that the effect under investigation can be measured quickly and clearly presented. The effect of such patient selection was the subject of a recent publication in the German Medical Journal: Nüesch et al. demonstrated in a recent study that the exclusion of patients in randomized controlled trials often leads authors to misinterpret the therapeutic effect. Due to the inclusion and exclusion criteria used, the recording of side effects in interactions with other medications is often not evaluable and is frequently not addressed in these phases. The interpretation of the results of randomized trials with a selected patient population is further complicated by the consideration of surrogate parameters, which often fail to demonstrate a relevant positive effect of the interaction in the medium and long term.
Specific population groups, such as our pediatric patient population or geriatric patients, are often not addressed by such studies in principle.
Meta-analyses and systematic reviews often fail to provide conclusive evidence regarding the positive or negative effects of a procedure. Here, too, the work of Nüesch et al. demonstrates a distorting effect of patient exclusion in meta-analyses on the assessment of treatment efficacy.
The problems with creating meta-analyses and the issue of study comparability are outlined here as follows:

• Difficulty in identifying all relevant studies
• Negative (non-significant) results are often not published.
• inadequate presentation of related methods
• Variations in study design
• retrospective studies
• Variations in the influencing criteria
• poorly coordinated criteria for quality measurement
• different measurement methods
• Variations in the method of analysis
• Different uses of statistical methods
• Variation in the presentation of results (e.g., survival at different endpoints, variations in study design)
Röhrig et al. presented a differentiated analysis of the choice of different study types in the German Medical Journal (Deutsches Ärzteblatt), focusing on scientific quality and validity. The authors also describe the common types of secondary research, but do not address health services research.
In contrast, in 2006 Willich pointed out the limitations of randomized studies in the German Medical Journal and highlighted the importance of health services research for the quality of medical care.
The fifth function of health services research proves to be both a necessary complement to experimental basic research, which generates potential for clinical innovations, and a complement to epidemiological and clinical research, in which etiological research approaches (aimed at understanding causal mechanisms or the effectiveness of interventions under ideal study conditions [efficacy]) are the main focus. Under the health services research paradigm, hypotheses about effectiveness under everyday conditions (effectiveness, relative effectiveness) and with regard to efficiency in the real healthcare setting are tested.
Health services research explicitly addresses the realities of healthcare provision in order to compare the effectiveness of different treatment measures. Health services research is not a replacement for the currently established primary and secondary research, but rather, in the authors' view, a necessary complement.
The following section presents exemplary cases of procedures and medications where side effects were only discovered after approval and use in real-world healthcare settings. The resulting consequences from a regulatory perspective will then be discussed.
In 2000, the results of the VIGOR study (Vioxx Gastrointestinal Outcomes Research) were published in the NEJM. This randomized controlled trial showed that rofecoxib (cyclooxygenase-2 inhibitor, Vioxx) has fewer gastrointestinal side effects than naproxen, the standard anti-inflammatory drug.
Unexpectedly, the VIGOR study showed a fourfold higher incidence of myocardial infarctions in the rofecoxib group. This observation clearly indicates a significant risk potential of the drug (source in print). More recently, the comprehensive analysis of Vioxx administration revealed substantial side effects that necessitated a change in the drug's indication.
Despite its initial success on the market, the cardiovascular side effects of Refecoxid were only demonstrated in 2004 – four years after its approval – through a placebo-controlled study. By then, more than 100 million prescriptions had been issued, and 80 million patients were taking Vioxx.
The manufacturer withdrew the drug from the market due to the excessive risk of stroke and heart attack. The Food and Drug Administration (FDA) is struggling to find an explanation.
Drug-eluting stents (DES) for the treatment of coronary artery disease are currently implanted in the majority of the more than 2 million percutaneous coronary interventions (PTCA). The FDA's approval of DES was based on the significant reduction in restenosis after PTCA, as demonstrated by randomized trials comparing DES to bare-metal stents (approximately 101 TP3T vs. 301 TP3T). Stable coronary artery disease, single, non-complex stenoses, and de novo lesions were the indications for PTCA in most of the patients studied here. After initial controversy and differing data on patient safety, the new method (DES stent) can be considered comparable to and safer than the older bare-metal stents. With unchanged mortality rates among patients, the lower restenosis rate was ultimately the only factor leading to a significant increase in the frequency of implantation. The implantation of drug-eluting stents (DES) was frequently and uncritically extended to off-label indications, even without supporting studies. In addition to the higher cost of the implant and the increased follow-up costs for dual antiplatelet therapy, DES thus became an enormous financial burden for the insured community. Many of the indications could not be scientifically verified due to the complexity of randomized controlled trials. Increased collection of patient data in registries or within the framework of health services research would have been desirable in this regard.
But what is the situation with the
different study formats
randomized vs. comparative presentation?
Kirtane et al. posed the interesting question of whether the widespread use of DES in the "real world" yields the same results regarding efficacy and safety compared to randomized controlled trials. Comparative and randomized controlled trials comparing DES vs. BMS with more than 100 patients from February 2008 onwards, using mortality as the endpoint and a cumulative follow-up of at least one year, were identified.

Randomized controlled trials and comparative studies were analyzed separately for this purpose.
The randomized trials found no differences in long-term survival rate or incidence of myocardial infarction between DES and BMS.
Real-world observations with a larger number of patients, however, showed the superiority of DES, with a reduced mortality rate and a lower incidence of myocardial infarction. The authors conclude that DES can be used safely and effectively even off-label. Furthermore, the authors point out the differing results of randomized versus observed trials comparing DES and BMS. Whether this conclusion will stand up to further investigation remains to be seen. The authors also emphasize that neither randomized nor observational studies can be relied upon completely. These findings underscore once again the urgent need for health services research to evaluate therapies in real-world settings.
In general, it must be weighed whether a procedure can be recommended based on the expected risk, or whether the risk of side effects and complications in "real life" is considered higher. The answer to this question should also be followed up by the regulatory authorities after approval has been granted, as is done with the DES registry.
The challenge lies in striking a balance between promoting innovation and avoiding unnecessary administrative over-control, which would hinder necessary innovation.
Ordopolicy perspective
Any intervention in the existing healthcare system, whether politically, legally, or economically motivated, results in altered conditions for the users of that system. The effects of such changes must be scientifically investigated in order to counteract negative developments on the one hand and to identify and utilize positive effects on the other. Such studies under everyday conditions form the core of healthcare research.
Health services research encompasses various tasks, such as describing and analyzing the current state of healthcare provision (both "as-is" and "deficit" analyses) and developing healthcare concepts. Furthermore, it includes providing scientific support for the implementation of new healthcare concepts and evaluating both new and existing concepts under real-world conditions at the level of the general population or relevant subpopulations.
An important aspect of health services research is the investigation of undesirable effects, such as side effects of medications. Reviewing the use and safety of drugs after approval is only possible through observational studies under everyday conditions, which can provide timely and valid results for such questions.
In 2004, Vandenbroucke pointed out: "The average randomized drug trial is too small, and does not have sufficient follow-up to detect adverse effects that are fewer than about one per 200 per year, or that take longer than 1 year to develop. To investigate adverse drug reactions, either case-control studies or large-scale observational follow-up studies are needed."„
To detect rare risks, large patient numbers are regularly required. This relationship is illustrated in Table 1 below: It lists the minimum number of participants required per group, depending on the frequency of the outcome in the control group and the level of risk increase being investigated in the respective study.
To enable regular monitoring of the efficacy, safety, and use of pharmaceuticals after their approval, a representative data pool is required. This is still lacking in Germany, even though the establishment of such a joint data pool for statutory health insurance funds (SGB V § 303a-f) was enshrined in the Social Code as early as 2004 with the Health Care Modernization Act (GMG).
The German Council of Experts on Developments in the Healthcare System (SVR) already determined in its 2000/2001 report that Germany lags considerably behind in healthcare research compared to international standards. Subsequently, the SVR has repeatedly called for an intensification of healthcare research in Germany in its reports. In parallel, policymakers, healthcare payers, and providers have increasingly recognized the importance of healthcare research and launched a number of initiatives and projects.
In 2005, the 108th German Medical Assembly resolved to implement a framework concept for promoting healthcare research through the German Medical Association. A year later, the German Healthcare Research Network was founded as an interdisciplinary network open to all institutions and working groups involved in ensuring medical care from scientific, practical, or health policy perspectives. Also in 2006, the Federal Government and the social security institutions concluded an initial agreement on promoting healthcare-related research. This alliance for healthcare research included the Federal Ministries of Education and Research, Health, and Labor and Social Affairs, the German Pension Insurance, the umbrella organizations of the statutory health insurance funds, and the Association of Private Health Insurance Companies. It was agreed to jointly support research in the area of "chronic diseases and patient-centered care" and to utilize the transfer of knowledge in healthcare research to improve patient care. A total of approximately €21 million was made available for this purpose over six years.
The current governing coalition also emphasizes the importance of healthcare research in its 2009 coalition agreement. Under the heading "More Research in Healthcare," the agreement states: "Health research contributes to improving the quality of life for people from all walks of life through innovation, while simultaneously ensuring the financial sustainability of the healthcare system. Insights into healthcare delivery under everyday conditions are particularly important so that the quality and efficiency of healthcare can continue to increase despite limited resources. Therefore, we will systematically expand healthcare research."„
To sustainably expand and permanently establish health services research as an independent scientific discipline in Germany, additional funding is needed for the training of early-career researchers and the development of scientific structures, in addition to project-based funding. Furthermore, high-quality research requires funding that is free from the interests of manufacturers and providers. Against this backdrop, the German government's objective of strengthening health services research with high priority and providing it with long-term, sustainable support is expressly welcome. The current framework program for health research includes four measures for this purpose. Funding is to be provided not only for studies but also for study structures, junior research groups, and centers for health economics research. In February 2010, an initial €54 million was allocated for studies in health services research and for the establishment of centers for health economics research.
In addition to research funding from tax revenue, further sources of financing should be considered. For example, in its 2009 report, the German Council of Experts on Health Care (SVR) proposes that statutory health insurance funds should be legally permitted to finance health services research studies with a fixed percentage of their benefit expenditures. Health insurance funds require detailed information about healthcare provision for their contractual agreements, which can often only be obtained through comprehensive studies. From the health insurance funds' perspective, the primary goal of health services research is to develop valid, pragmatic answers to concrete, healthcare-relevant questions as quickly as possible and in a patient-centered manner, answers that can serve as a basis for decision-making processes. Specifically, the SVR calls for "statutory health insurance funds to be legally permitted to finance such studies with a fixed percentage of their benefit expenditures (e.g., 0.1 %)." If this proposal were implemented, over €160 million per year would be available to statutory health insurance funds for health services research studies.
An alternative proposal suggests involving not only statutory health insurance funds (GKV) but all potential beneficiaries of such studies in their financing. An example of this is the arrangement at InEK (Institute for the Hospital Remuneration System), where costs are distributed among all affected payers based on defined service cases. The statutory health insurance funds have therefore proposed levying a system surcharge (research cent) to finance healthcare research.
To fully exploit the vast potential that healthcare research offers for the further development of the healthcare system, the needs of all those affected must be considered as comprehensively as possible. Only under this condition will the necessary acceptance among stakeholders and the general public for long-term expansion be achieved.
In healthcare research projects, which typically involve highly sensitive health data, adherence to all data protection regulations is of paramount importance. This concerns not only the data protection interests of patients but also those of the healthcare providers involved. Regarding the diagnostic data of hospital patients, Schoffer states: "For a patient, as the data subject, the primary risk is that the reason for their hospital stay—the main diagnosis—can be discovered by individuals (e.g., employers) who might misuse this information. This can lead to negative consequences for the patient. For a hospital, as the data subject, the risk is that information about its financial situation becomes public, which could be used by health insurance companies, competitors, planning authorities, etc., and lead to negative consequences. Generally, the potential harm caused by deanonymization is considered greater for hospitals than for patients."„
Against this backdrop, research data centers have developed the procedure of 'controlled remote data processing'. With this procedure, data users do not have direct access to the anonymized individual data, which are held exclusively by the statistical offices. Researchers receive only so-called structured datasets (dummy files) that are identical to the original data in structure and characteristics. Using these dummy files, researchers create an analysis program, which they send to the responsible research data center. There, the original data is analyzed using the analysis program, and the results are then checked to see if they raise any data protection concerns. Only if no such concerns are the results transmitted to the researchers. Under no circumstances are original datasets of individual data subjects shared; only aggregated results are ever provided.
In the future, healthcare research projects will increasingly focus on linking data from various sources, such as routine data from statutory health insurance with data from cancer registries. So far, such projects in Germany—unlike in Scandinavia, for example—often fail due to data protection regulations. The German Council of Experts on Healthcare (SVR) therefore calls in its current report: "Practical methods should be enabled to allow for the simplified merging of healthcare information for scientific purposes."„
The controlled remote data processing method described above offers, in principle, the possibility of conducting such studies while safeguarding the interests of all those affected.
Summary
Even when using prospective randomized studies or meta-analyses, research development is limited in its ability to draw conclusions about broad application to the population.
Many examples show that substantial side effects are only observed in the application to the general population after the implementation of therapies or medications.
The next logical step must be the expansion of healthcare research, which can extend the limitations of the current concept for evaluating new procedures.
Health services research holds great potential for the further development of the healthcare system in Germany. To unlock this potential, the further expansion of health services research is urgently needed. A prerequisite for the sustainable development of this young research discipline is secure funding. Furthermore, innovative solutions for data protection issues must be developed.

Definition according to H. Pfaff, ZVFK (2003)
„"Health services research can be defined as basic and problem-oriented interdisciplinary research that examines the framework conditions of healthcare and the provision of medical care."
• describes,
• explained causally and building upon that
• Developed care concepts,
• their implementation is being researched and/or
• evaluated under everyday conditions.“