comment
In everyday language, safety is perceived as an objective state. Politicians talk about safety, we discuss drug safety, and we use seat belts in road traffic. From a scientific perspective, however, safety is not an objectifiable state: safety is the subjective perception of objective risks (2, 3). With the safety loop, we have illustrated (Fig. 1) that there is a significant relationship between the subjective perception of objective risks, i.e., "felt safety," and the effect of this subjective perception on changes in objective risks (4, 5). Fig. 1 explains that communication (verbal or non-verbal) is the most significant factor influencing the perception of objective risks. This clarifies the importance of the safety loop for the discussion of business models in the healthcare system. Since the mere promotion of a business model conveys "felt safety," and this perception influences the patient's objective risk, it becomes clear that health risks increase if the claims of a business model are inaccurate.
The resulting dilemma stems from the lack of an accepted method for demonstrating real-world effectiveness. Internationally, randomized controlled trials (RCTs) are considered the best method for collecting valid data in healthcare systems. This statement holds true for demonstrating efficacy, i.e., the proof of principle, but not for demonstrating real-world effectiveness. Under the experimental conditions of an RCT, all known confounding or risk factors are deliberately avoided through the definition of exclusion criteria. While this allows for proof of principle, it does not reflect real-world healthcare practice, because confounding and risk factors are inherent in everyday healthcare.
Over the past decade, we have collaborated with various international groups to develop a system called the Pragmatic Controlled Trial (PCT), which allows us to compare the effects of different interventions under everyday conditions (without randomization) (6, 7, 8). This method is based on Bayes' Theorem, which uses available information to confirm a hypothesis step by step. Battles were won and a secret code (Enigma) was deciphered during World War II by applying this theorem (9). Thus, the theorem of the Presbyterian pastor and mathematician Thomas Bayes (1701–1761) was the first method, long before the development of randomization, to demonstrate that coincidences are indeed suitable for inferring causal relationships through intelligent use (10).
These non-experimental methods are essential for analyzing "big data." When predictions about analyses are disseminated, it's often forgotten that "big data" simply represents a "charming natural chaos": every patient is different from every other, and virtually no two therapies are absolutely identical. Therefore, the first challenge in analyzing "big data" is to develop algorithms that can group patients and interventions into largely homogeneous groups for later comparisons. This requires translating a precise research question into a language that computers can understand.
In February 2018, we complained in the German Medical Journal (11) that the decision-makers at the Institute for Quality and Efficiency in Health Care (IQWiG) do not recognize the importance of the effectiveness of healthcare services in everyday practice (12). In their response to our commentary, factual arguments would have been helpful in reconsidering our own statements (11). The Institute of Clinical Economics (ICE) e.V. is dedicated to developing and disseminating methods for increasing and describing the added value of healthcare services under everyday conditions (www.clinical-economics.com). We are seeking partners who can support us with sound scientific contributions so that we can provide scientifically based comments on the healthcare system master plan before economic concepts are discussed. To this end, we have compiled the differences between science and economics from our perspective.
The difference between science and business
Economics and science are synonymous. Economics will
Justified by science, science is financed by the economy. This interdependence is unavoidable, but not without its problems. Reversing these asymmetrical relationships might achieve economic goals, but not healthcare. Unless otherwise indicated, we have used our own definitions in Table 1 because there are no universally accepted definitions for the terms used.
The goal that has not yet been achieved should be to find different terms for defining different content.
Two recent findings support the urgent recommendation to back up economic concepts with reliable data: Not everything labeled "science" actually is. John Ioannidis of Stanford University has written a whole series of noteworthy papers on this topic. His 2005 work, "Why most published research findings are false" (13), is among the most cited publications. The magazine "Stern," in its September 5, 2019 issue (No. 37), published an appeal by physicians against the dictates of economics in our hospitals, entitled "Save Medicine," which was initially signed by 215 individual physicians and 19 organizations. The appeal contains three demands:
• „The DRG system must be replaced or at least fundamentally reformed.”.
• The economically driven, dangerous overtreatment and under-treatment of patients must be stopped. We acknowledge the necessity of economic management.
• The state must plan and properly equip hospitals where they are truly needed. This requires a master plan and the courage to merge two or three clinics in some places into larger, more efficient, and better-staffed centers.“
This appeal can only be implemented under socially acceptable conditions. No one will support it if they are made to feel insecure. The appeal will be fully supported by all patients with whom a trusting doctor-patient relationship exists. In addition to these conditions, the prerequisites already addressed in the following seven discussion points (1) must be established. We support some of these statements and explain our position where it differs.
The seven discussion points of the BMC
1. Reimbursement for digital applications
If similar requirements are to apply to pharmaceuticals and digital health applications, at least with regard to the reimbursement of costs for the latter, proof of the added value resulting from the availability of this health application must be required.
2. Extension of access rights to the telematics infrastructure
This demand should be supported if the opportunity costs of the additionally financed services, e.g., from management companies or patient education, not only promise but demonstrate the expected added value.
3. Continuation of the Innovation Fund
This point criticizes the fact that limiting funding to 15 projects follows an unclear logic: smaller projects are supposed to be able to achieve comparable insights to larger ones. The statement that the targeted number of cases is frequently not reached and that the lack of a description of the solutions achieved is considerably more significant.
The statement that a different study design would require fewer cases would have been supported by concrete data if the authors of this statement had been aware of its implications.
We expressly support the proposal not to limit the number of open-topic contributions or the duration of the projects. From ICE eV's perspective, each proposal should be described – in a way that is understandable to a subject matter expert – on a maximum of one page.
a) which supply problem will be solved by the proposed project,
b) which data and methods are needed for this and
c) by which evidence the feasibility of the project can be confirmed.
Facts must be distinguished from expectations based on literature citations. Using well-known literature citations reduces processing time by eliminating the need to check their validity.
However, if approved projects fail not only in exceptional cases, e.g. due to insufficient recruitment, deficiencies in the review process should be questioned, because a professional reviewer can be expected to recognize these deficiencies, which almost always lead to ethical and economic problems in the event of a student dropping out.
A significant flaw in the review process is perceived when projects that have been meticulously designed are rejected without scientific justification. This raises serious doubts as to whether a peer review process was actually applied. Furthermore, unjustified rejections reduce the motivation to participate in future calls for proposals.
If a continuation of the investment fund is planned, the following should be considered:
• For each project, the above-mentioned points a – c must be assessed by a subject matter expert.
If no deficiencies are found, the project can be added to the list of potentially funded projects:
• If defects exist, it should be noted whether they are correctable.
• In cases of irreparable deficiencies, these will be cited as the reason for the scientific rejection of the project. A comprehensible justification precludes funding during the current application phase. The rejection, with justification, can be issued immediately.
• If correctable defects are identified, suggestions for optimization will be provided. If these are submitted on time, the project will be added to the aforementioned list; otherwise, it will be rejected.
Both decisions will be published with a statement of reasons. Potentially eligible projects will be ranked in a list based on existing conflicts of interest. Conflicts of interest are easily identifiable and identifiable in the field of health services research.
4. Responsibilities of the office
Health services research is a relatively young, evolving field of research. The office should therefore be able to identify errors in the design or implementation of a project, as well as in the evaluation and interpretation of the data. If an idea appears worthy of funding, the efficiency of the entire process could be improved.
The funding program could be significantly increased if all available expertise were mobilized to eliminate identified shortcomings while they are still correctable. This would create a completely new culture of project funding, one that would utilize the resources already invested in the application process.
The current practice of simply informing unsuccessful projects of the negative decision without providing a reason does not do justice to the effort that the project groups have invested in the application.
Consideration could be given to funding 20 projects over two years instead of 15 projects funded over four years. After two years, the 10 projects that have achieved the most of the precisely defined and consecutively completed milestones could be selected. These 10 successful projects would then receive funding for another two years. The remaining 10 projects could receive a funding recommendation based on their progress and, if they demonstrate solid results, would very likely find a sponsor.
The relatively small field of health services research in Germany could become significantly more transparent and efficient through a professionalization of its funding strategy. If scientists whose previous professional experience involved science but little business model development are tasked with selecting the most suitable business models, there is a risk that this selection process will result in funding for some good scientific projects but very few effective business models.
5. Standardization of the evaluation
The standardization of the evaluation called for here accurately addresses the core of the problem, but dilutes it with disorganized proposals regarding the details. Every innovation begins with an idea. This idea must be translated into a concrete question. Only when the majority of readers recognize the same project objective from the chosen wording of a question is the question formulated with sufficient precision. Suitable endpoints, which are then measured in the project, must be derived from precise questions. Anyone who measures endpoints without considering the risks that can distort one or more of the measured endpoints in everyday healthcare practice will despair at the inherent chaos of daily healthcare.
Choosing the endpoints is where the challenge of finding the appropriate detection method begins. It would go beyond the scope of this discussion to differentiate that the sensitivity and specificity of the detection method are only the first two criteria for selecting the appropriate test. The clinical relevance of the test result becomes clear when it can be demonstrated that, in everyday clinical practice, further consequences are derived independently of the test result.
6. Transfer of innovations into standard care
These decisions are always influenced by differing values and therefore cannot be solved by science alone – but certainly not without it either.
7. Avoiding conflicts of interest within the system
The promotion of healthcare innovations should be subject to the same conflict-of-interest scrutiny as the debate surrounding a "free lunch"; it makes no sense to publicly discuss the consequences of a questionable arrangement while sweeping under the rug the misguided decisions involved in funding unfeasible projects worth millions. Action and accountability are also necessary when allocating subsidies.
In all of this, the following must be considered: The optimization of healthcare is not measured by successful business models, but by the reduction of health problems (Patient and Society Reported Outcomes).
