The Covid-19 pandemic presents healthcare systems with new challenges [8-11]. There was great concern about temporary overloading of inpatient capacities. Control at regional and national levels is only possible if valid data is available in a timely manner. Registers of intensive care capacities used and reports to health authorities can be helpful, but only provide limited information about what is happening in hospitals. This gap can be filled by routine data that is available on a legal basis in a structured and standardized manner without additional data collection effort [4]. The Quality Indicators for Church Hospitals Association (QKK eV) has been carrying out cross-provider quality benchmarking using routine data for more than a decade [1]. For this purpose, quality indicators published by third parties and developed within the association are calculated quarterly using the data in accordance with Section 21 of the Hospital Fees Act (KHEntG) and the results are reported back to the member institutions in an open comparison. The QKK indicator set includes 120 key figures in the current version 4.21. Of these, 70 were taken from the German Inpatient Quality Indicators (G-IQI), 17 from the Patient Safety Indicators (PSI) of the Agency for Healthcare Research and Quality (AHRQ) and 11 from the external comparative quality assurance in accordance with Section 137 of the Social Security Code V. 22 quality indicators were developed within the club itself. Particular attention was paid to vulnerable patient groups. Members of the QKK e. V. are individual hospitals as well as church sponsors of hospitals and associations. The set of indicators was briefly expanded to include key figures on Covid-19 in the first half of 2020. On the basis of these key figures, the care of Covid-19 patients in small and medium-sized hospitals will be reported.

>> The selection, development and evaluation of quality indicators is carried out by QKK e. V. about a systematic approach [12]. However, in view of the short-term need to define key figures for Covid-19, their selection was carried out in a shortened procedure based on a description of the frequency of corresponding patients, the adaptation of existing indicators such as long-term ventilation and the consideration of disease-specific complications from the literature [14] . No further risk adjustment was carried out in favor of quick and comprehensible results feedback. The Covid-19 key figures were therefore to be viewed as management information, not as quality indicators in the narrower sense. In April 2020, the QKK indicator set was expanded to include a Covid-19 module with 18 key figures.
The member institutions of QKK eV evaluated the data in accordance with Section 21 KHEntG for patients discharged between January 1st, 2020 and June 30th, 2020 by commissioned service providers or by using appropriate software (3M Deutschland GmbH, Neuss; Saatmann GmbH & Co. KG , Worms). The results of the 18 key figures of the Covid-19 module were combined for layoffs in the 1st quarter and 1st half of 2020 to carry out a comparison of the member institutions of QKK eV. However, the obligation of the club members to provide their data was lifted in view of the tense situation caused by the corona pandemic. The comparison therefore only included a portion of the membership, 67 hospitals.
Since February 17, 20202, the International Statistical Classification of Diseases and Related Health Problems 10th Revision, German Modification (ICD-10-GM) has had the possibility of identifying Covid-19 with the code U07.1! to encode. With the introduction of the code U07.2! On March 24, 2020, a distinction was made between Covid-19 with detection of the virus (U07.1!) and Covid-19 without detection of the virus (U07.2!). The code U07.2! was only taken into account for two key figures in the Covid-19 module (QKK C-08 and QKK C-13). All other key figures and information refer to patients with virus detection. The code U99.0 introduced on May 25, 2020! “Special procedures for testing for SARS-CoV-2” were not used.
For the 18 key figures of the Covid-19 module, the results for each hospital were available as values of the numerator and denominator for rates and as total hours with the number of cases included for durations. The overall results for the member institutions of QKK eV were calculated from the results for each hospital. The median was used to describe the distribution across facilities. The connection with hospital size, implemented via the number of inpatient treatment cases in the first half of 2020, was examined using the non-parametric Spearman correlation coefficient. The variability of care was estimated using the coefficient of variation (CV) of the results per hospital. The 95% correlation interval for the CV was calculated according to [3].
Results
The number of inpatient treatment cases per hospital in the first half of 2020 was between 617 and 21,326 (median 6,217 inpatient treatment cases). The proportion of patients with Covid-19 among all inpatient treatment cases was 0.47% (see Table 1). Six of the 67 hospitals had not cared for any patient with Covid-19. The median was 18 cases. The connection between the proportion of patients with Covid-19 and the number of inpatient treatment cases was not statistically significant with a correlation coefficient of 0.138 (p=0.266, N=67). All six hospitals without patients with Covid-19 were half of the smaller facilities in terms of the number of inpatient treatment cases. Of the Covid-19 cases treated in hospital, 17,77% died in hospital. The mortality of Covid-19 cases ranged between 0.94% in the age group 20 to 44 years and 35.20% in the age group over 84 years (Table 1).
Of the 61 hospitals with at least one patient with Covid-19, 53 cared for a median of 6 patients in intensive care (see Table 2). There was also no statistically significant connection with hospital size in the proportion of patients in intensive care with a correlation coefficient of 0.125 (p=0.339, N=61). Overall, slightly less than a third of patients with Covid-19 were treated in intensive care. The average length of stay for patients in the intensive care unit was 11 days. A fifth of patients with Covid-19 were ventilated for an average of 12 days. Of the 53 hospitals with ICU treatment, 50 provided ventilation. With 41.14%, ventilated patients had the highest mortality in all collectives considered (Table 2).
Respiratory distress syndrome occurred in patients with Covid-19 in 44 hospitals (72% of 61 hospitals), sepsis in 43 hospitals (70%). A median of 3 patients each had to be cared for. Only in 12 hospitals did no patient with Covid-19 develop respiratory distress syndrome or sepsis (20%).
The number of patients with respiratory distress syndrome or sepsis was around 10% (see Table 3).
The hospitals were affected by the pandemic to very different extents. The proportion of Covid-19 patients reached a maximum of 2.25%, corresponding to around 2 out of 100 patients; the median was 0.31%. In the most commonly affected age group of 65 to 84 years, mortality was up to 100% with a median of 20%. With a CV between 0.59 (mean duration of ventilation, 95%-CI 0.439-0.744) and 0.83 (mortality when treated in intensive care, 95%-CI 0.585-1.079), the results for intensive care care were similar between hospitals. However, serious complications developed to varying degrees with a CV of 1.04 (sepsis, 95%-CI 0.709-1.370) and 1.16 (respiratory distress syndrome, 95%-CI 0.761-1.557). With a CV of 2.22 (95% CI 0.972-3.464), the proportion of patients without virus detection was very inhomogeneous.
discussion
The present results from the hospitals affiliated to QKK eV show the feasibility and high information content of Covid-19-related key figures, which are calculated via an analysis of routine data. The results reflect both the medical severity of the pandemic and the associated resource expenditure. Routine data offers the advantage of a complete and timely addition of information beyond the established reports to registries. Since the data is available without
If additional recording effort arises as a by-product of hospital billing, its provision is reasonable even in difficult times.
With the key figures presented, information is available for the first time from a larger group of small and medium-sized hospitals on the care of patients with Covid-19. In contrast to [5], there was no restriction on the group regarding the type of health insurance. The sample included both legally and privately insured people. However, these were exclusively religious institutions. A distortion due to systematic differences in the care of patients with Covid-19 between denominational providers on the one hand and municipal and private providers on the other cannot be ruled out.
Even outside of maximum care, almost every hospital was affected by the corona pandemic. This included intensive care with artificial ventilation. This meant that appropriate care options could also be established in smaller and medium-sized hospitals. The higher homogeneity compared to other Covid-19 key figures could speak for a clearly regulated implementation of intensive care care. For example, the frequency of treatment in the intensive care unit as an indication of the indication and the duration of artificial ventilation as an indication of an evidence base were comparable between the hospitals.
Patients with respiratory distress syndrome or sepsis in Covid-19 are each rare, with a median of 3, but they occur in over 70%
of hospitals with Covid-19 cases. It is then important to provide structural options and human resources for these rare but serious treatment situations. Compared to the literature, the proportion of Covid-19 patients who developed respiratory distress syndrome was low, at around 10%. In a systematic literature review, this proportion was only lower in one of 17 included studies; but the weighted average was 33% [13]. Among the first 50 patients with Covid-19 at the Aachen University Hospital, this proportion was almost 50% (24 of 50 patients) [2].
Significantly higher rates were also reported for sepsis and septic shock. Among the first 191 patients hospitalized in Wuhan, 59% developed sepsis and 20% developed septic shock [15]. It should be investigated whether the challenges in caring for patients with Covid-19 differ between smaller and medium-sized hospitals on the one hand and university hospitals on the other, from which studies are typically reported.
A proportion of 0.47% Covid-19 patients with virus detection at the member institutions of QKK eV in the first half of 2020 almost corresponded to the proportion of 0.45% that was reported by the Quality Medicine Initiative (IQM) for inpatient admissions up to April 19, 2020 [ 6]. This is probably a lower limit. The ability to code Covid-19 was only introduced in February. The high proportion of patients with Covid-19 without virus detection is likely due to a possible lack of availability of tests at the beginning of the pandemic and also to coding errors. Among the cases with a code U07.2! However, there will also have been real cases of Covid-19 that were not taken into account in the QKK eV's key figures, assuming predominantly incorrect coding. This assumption is supported by the high quality of virus detection achieved [7]. Compared to the Covid-19 cases reported at the Robert Koch Institute, the patients treated as inpatients in hospitals were older (see https://corona.rki.de/).
The severity of the disease was reflected in the frequent need for intensive care, the high number of severe complications such as respiratory distress syndrome and sepsis, and a mortality rate of almost 18%. The proportion of ventilated patients was slightly higher at 20% than among AOK insured patients at 17% [5]. AOK insured people, on the other hand, were ventilated for an average of 13.5 days, about 2 days longer than in the collective presented. Mortality was higher among AOK insured people at 52.5% and 16.0% in both ventilated and non-ventilated patients with Covid-19. From a quality perspective, these differences offer a starting point for further comparisons between care structures in Germany.
For federal and state health policy decisions, a monthly evaluation of routine data could provide the appropriate basis for close monitoring of pandemic events. Emerging bottlenecks or the effects of individual measures could be monitored in order to enable timely reactions. Individual hospitals can use the figures to find out about the situation of comparable facilities. The clarification of whether and which key figures are suitable as quality indicators remains subject to further work. When it comes to caring for patients during the coronavirus pandemic, it is very important to prepare small and medium-sized hospitals for the treatment of severe cases requiring intensive care. <<