Quality and patient safety in light of the SARS-CoV-2 epidemic

Many issues that have dominated the health policy debate in Germany and internationally in recent years have been pushed into the background by the coronavirus crisis1.
Quality orientation as a central paradigm – who today still remembers the intensity with which the quality initiative was discussed in the middle of the last decade, with patient safety issues at the forefront, with competitive elements like pay for performance raising emotions, patient autonomy being held in high esteem, and patient-reported outcomes being celebrated as cornerstones of quality measurement? Perhaps the most difficult task today is to restore continuity between the present, shaped by the coronavirus, and the issues of this seemingly distant past. Not everything can have been wrong, but on the other hand, it must be recognized that, due to the coronavirus, some of the assumptions and certainties of just one and a half years ago that were considered unshakable require intensive discussion and revision. This text can only outline the problem and present a work program using a few examples. A detailed derivation of the conceptual consequences based on the findings from the coronavirus pandemic to date and the technical and scientific foundations of the quality and safety debate of the last decades must be reserved for in-depth discussion in the coming years and will take up a considerably larger scope.

Quality and safety: the starting point
>> Assessing the impact of a momentous event such as the SARS-CoV-2/COVID-19 pandemic requires a precise description of the initial situation before the coronavirus pandemic. Summarizing the insights gained from the highly contentious discussions on quality and patient safety in recent years, the following points are particularly noteworthy:
• the neutral definition of quality through the “requirements” to be negotiated, which result from objective needs, perspective of observation and structural relevance, and with regard to which “characteristics” are assessed [1],
• the normative definition of patient safety based on the “absence of adverse events” or – better – on the willingness of institutions and systems to perceive, change and ultimately innovate [2],
• the complex nature of healthcare, which is also reflected in the emergence of patient safety-relevant events [2],
• the differentiated methods of recording quality and safety, which are clearly distinguishable from one another in terms of statistical requirements, ranging from the scientific approach (experiment) to the quantitative recording in the everyday treatment of large collectives and the monitoring approach of the indicator concept to the generating procedures that serve to gain knowledge in the area of the so-called unknown unknowns (e.g. CIRS or surveillance methods in the pharmaceutical sector) [2],
• the central use of patient-reported outcomes, i.e. treatment results that can only be reported by patients (in connection with patient-centered treatment) [3],
• the rule-based identification, evaluation and synthesis of scientific studies (evidence-based medicine and the methodologies of health services research based on it) [4], and
• the comprehensive body of knowledge on the effectiveness of interventions that can achieve quality improvement, are themselves complex in nature in a complex context and can therefore only be effective if they are composed of several carefully coordinated individual interventions (so-called complex multicomponent interventions) [5].
Corona – the ultimate challenge
While the aforementioned points can be considered consolidated, they by no means represented a "dormant situation" even pre-coronavirus. Nevertheless, the current challenges posed by coronavirus are of such a powerful nature that some previously vigorous debates, such as the importance of billing data in assessing the quality of care or the relevance of patient-centeredness from today's perspective, could almost seem irrelevant. Of course, it cannot be overlooked that the profound significance of coronavirus cannot be discussed solely in the field of quality and patient safety, as conflicts are being played out that already existed and, in their interaction, have now acquired such relevance that one can practically speak of a change of era (for political theory, see [6]). However, an aspect inherent in the medical-biological field is easily overlooked: the fact that an epidemic is caused by an infection caused by an infectious agent and manifests as a transmissible infectious disease. This circumstance is of unique importance from the perspective of the quality (and safety) of care: infections represent the only quality and safety problem that is communicable, i.e., has its own biological activity. Epidemically occurring infectious diseases affect the interaction between patient and society to a much greater extent than other diseases. The differences in how societies deal with them range from Australia to Sweden and ultimately provide insight into the inner workings of a society. It should be noted, however, that quality problems associated with an infectious disease have a particularly strong impact not only on the individual perspective but also on the population and societal perspective, not only indirectly (e.g., through costs), but directly: through the spread of the disease and the resulting burden of disease.
Concept of quality under stress
It is therefore beyond doubt that the coronavirus epidemic provides a reason to include not only the individual but also the regional, population, and societal perspectives in the formulation of the constitutive concept of "requirement." However, existing quality concepts are not able to adequately reflect the diversity of perspectives. The purpose here cannot be to question Donebedian, and bidimensional concepts such as those of the OECD and IQTIG [7] are still valid (albeit not satisfactory). However, the current discussion about coronavirus shows that more differentiated concepts are urgently needed that explicitly consider the societal perspective in addition to the individual and institutional perspectives. In "Quality 2030" [8], the author presented a concept that explicitly addresses the question of perspective as one of three dimensions (see Fig. 1). A similar extension should be made with regard to the outcome consideration in health care research: what really matters is what reaches the patient, but what is realized at the population level, regionally or as defined as a state structure, also counts [9].
The population or societal perspective is also crucial in assessing the course of the disease, because the regional impact of the coronavirus varies enormously and fluctuates over time. While it was once the metropolises, as "aerosol hells," that appeared responsible for the spread, only a few weeks later the rural regions were the most concerned. While it was once the "champions" (for example, Ireland) who had everything "under control," the epidemic erupts there again with full force just a short time later. How this development can be explained in detail will be left to later analyses; linear models are unlikely to provide adequate explanations.
Other perspectives on the concept of quality have receded into the background, such as the benefits of individual interventions, although this topic has dominated the debate over the past decades. One could almost speak of a crisis in evidence generation, which played such a significant role in benefit assessment; in many places, a relapse into the pre-scientific world has been observed. Randomized trials, the evaluation of studies according to their methodological strength, and comprehensible syntheses of knowledge – none at all. Instead, opinions dominate, individual scientists are elevated to authority, pre-publication servers become important sources of information, and methodological discourse gives way to scandalous reports. In some cases, there has even been open polemics against institutions of evidence-based medicine, and several urgent appeals were necessary to prevent the complete abandonment of a critical reception of knowledge and experience. The same applies to patients and patient rights, which had been strongly emphasized by law just a few years ago (see the Patient Rights Act of 2013). Of course, everything possible is being done to save patients' lives during this difficult time, but restrictions on visits and freedom of movement were still the order of the day, even though older people, for example, explicitly expressed their desire to maintain contact with the outside world, even at the risk of infection (for the question of information before COVID-19 vaccination, see [10]). Health and care workers are unwittingly playing the role of mouthpiece for "patient-reported outcomes" when they report increased, sometimes dramatically increased, aggression and despair among patients, not to mention the fate of the dying.
On the use of quantitative measures
The coronavirus epidemic has brought one of the fundamental principles of quantifying complex healthcare problems into sharp relief: If one wants to measure the rate of nosocomial intravascular catheter infections in an intensive care unit, one must quantitatively record these events and relate them to the total number of patients at risk. However, if one wants to investigate the complex question of how
To determine how good the care in one or more intensive care units is overall, the rate of catheter infections can also be used, but as an indicator, i.e. as a predictor of the occurrence of quality problems (for a detailed discussion [2]).
We have also experienced such a change in meaning with Corona: While the notification rate (German Mel-
Initially intended as a value to record the number of infected people (also inaccurately referred to as "incidence" [11]), its use changed at the beginning of 2021, and it was now referred to as an "indicator." This term was also used by the Federal Constitutional Court in the pending proceedings to clarify the constitutionality of the lockdown. In a scientific expert opinion on one of these proceedings, the author explained that it only makes sense to speak of a relevant indicator if it [12]
a. is statistically adequately presented and specified,
b. can be measured reliably,
c. is valid in its prediction function (which verifiably predicts problems),
d. and is not influenced by third factors (so-called confounders) (known or unknown) that distort the prediction process, or it must be corrected accordingly (risk adjustment).

These conditions are not met in the case of the "7-day incidence" used in the coronavirus epidemic, leading to many of the problems encountered in managing the epidemic (lack of credibility of political action, unexpected developments that could not be predicted, etc., even the vaccination rate could apparently not be measured). Above all, however, it should be noted that unreliable indicators can never be valid in their predictive function.
However, it must be emphasized that this situation is self-inflicted and could have been easily prevented by taking into account the state of knowledge that has long been established, for example, in quality and safety research. One possibility is a cohort study, in which a true statement on the incidence rate could easily be made longitudinally in a randomly selected population with regularly conducted tests. Another option would be the establishment of indicator sets, such as those used in other healthcare areas (detailed overview and discussion [2]). An initial proposal for the coronavirus epidemic came from the German Hospital Federation [13] and quite correctly introduced a (long overdue) age stratification. In the aforementioned report and in Position Paper 8 [6], this approach was expanded in the 14-day approach [11] to include further indicators on host and environmental factors as well as testing frequency (see Table 1).
The gold standard intervention:
Complex multiple interventions
In addition, there are fundamental considerations that affect the understanding of an epidemic and the structuring of preventive measures. There can be no doubt: it is not enough to simply adjust a single screw. Any single-point strategy is doomed to failure from the start, as decades of experiments and studies, for example, in the areas of nosocomial infections or patient safety, have clearly shown. It is irrelevant whether a single-point intervention is based on regulations ("commandments"), sanctions, technical systems, training, knowledge-based persuasion, or financial incentives – the result is always the same: if any effect is observed at all, it is small, short-term, and in the medium term worsens the mood ("it was nothing again").
The reason for this predictable failure of single-point interventions lies in the ability of complex systems to absorb punctual changes and stimuli from the environment and then return to their previous state. Lasting changes in complex systems can only be initiated through multiple interventions that coordinate different intervention levels simultaneously or in a well-timed manner. In addition to the aforementioned intervention levels, group training, social appreciation, etc. should also be considered. These are also referred to as complex multicomponent interventions (CMIs), a concept that was developed and extensively evaluated in the control and management of hospital-acquired infections. The preventive measures for controlling and managing an epidemic comparable to SARS-CoV-2/COVID-19 are a typical case in which such complex multiple interventions must be used.
The CMIs originate from scientific landmark studies in the field of infection control in the USA, where more than 20 years ago, attempts were made to develop urgently needed measures to manage infectious epidemiological emergencies that could surpass the limited effectiveness of traditional approaches. These so-called Michigan studies by Peter Pronovost et al. [14] initially originated from supra-regional hospital associations (i.e., they were of organizational theory origin), but were extensively replicated internationally and can now be considered the standard for all efforts to achieve sustainable progress in the field of infection control. These studies were based on the insight that although epidemics are caused by a biological agent (pathogen), they are essentially social events that can only be controlled by attacking several points simultaneously. Against this backdrop, the one-dimensional, linear notion that in the case of SARS-CoV-2/COVID, one only needs to prevent transmission (the well-known "R factor below 1" requirement) can only be classified as reductive and ultimately outdated. The repeatedly asked question as to why this view is still being put forward and attracting such resounding attention in the media and in politics can only be answered by the fact that virtually all experts who have had personal experience with epidemics in developed countries (e.g., HIV infection, nosocomial infections) and not only have a molecular genetic or big data-driven background report such complex interventions as the key to success. These experts also address practically relevant issues (such as the question of transmission via inanimate objects in March 2020).
Work programme: Quality and
Security post-Corona
The pre-COVID state of the discussion outlined above must be reviewed in a concerted effort by expert institutions and scientific bodies to determine the extent to which adjustments or expansions have become necessary due to the COVID-19 epidemic. The following topics must be considered first and foremost:
• Societal, regional, and population perspectives must be given greater relevance when formulating requirements for the quality of healthcare. This relevance is materialized not only in terms of differentiation from the institutional perspective, but also in terms of differentiation from the individual patient perspective—the strongly utilitarian viewpoints that have emerged in recent months make this seem unavoidable. Strict attention must be paid to the accurate specification of the relevant requirements.
• The understanding of patient safety must be examined to determine whether it can be maintained in its current form and in what form social factors can be reliably demanded in the realization of patient safety compared to individual protection against adverse events under the conditions of an epidemic.
• With regard to the measurement of quality and safety, it is necessary to examine, while maintaining the differentiated application of the different methods, how a relapse into the use of singular, largely neither reliable nor valid measurements could have occurred, and how more valid measurement instruments (e.g. scores or quantitative measurements from cohort studies) can be developed, especially with regard to possible future epidemics.
• Patient-centeredness and patient-reported outcomes require fundamental reflection as to whether this is really a robust concept or whether it is a scientific and professional fad that has become completely irrelevant when it is first put to the test.
• Evidence generation and synthesis through EBM represents one of the most important achievements in the health sciences of recent decades and has always been closely linked to the issue of quality and safety. It is worth discussing how it was possible for eminence-generated approaches to once again come to the fore, and how individuals and institutions that advocated EBM methodology, including with regard to the coronavirus, were even attacked and defamed. This aspect is of paramount and overarching importance for the scientific system.
• When planning preventive measures and other interventions, complex multicomponent interventions should be considered as a standard. Insistence on a single measure (e.g., lockdowns) is long outdated and should be replaced by intelligent, complex interventions. Numerous, scientifically well-evaluated instruments and approaches exist in practical clinical infection epidemiology.