Treatment concepts and mortality
Covid-19 - A comparison between the first and second waves of the corona pandemic from secondary data from inpatient care

>> The basis of the analysis was data in accordance with Section 21 of the Hospital Fees Act (KHEntgG) from the member institutions of QKK e. V. Inpatient treatment cases with discharge in 2020 were included. Treatment cases admitted between January 1st, 2020 and June 30th, 2020 were assigned to the first wave, and those admitted between July 1st, 2020 and December 31st, 2020 to the second wave. However, the obligation of the association 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 at 63 locations. The treatment cases from both waves were descriptively compared with regard to general characteristics and the ventilation situation.
Since February 17, 2020, 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 code “Covid-19, virus detected”. As of November 11, 2020, code U07.5 “Multisystem inflammatory syndrome associated with Covid-19, unspecified” was added. Cases with Covid-19 were identified using both codes. Mechanical ventilation was defined by the number of hours of ventilation. To differentiate between invasive and non-invasive ventilation, codes 8-701 “Simple endotracheal intubation” and 8-706 “Applying a mask for mechanical ventilation” of the operations and procedures code (OPS) were used. In all cases with at least one of the two codes, a non-invasive ventilation attempt was assumed with either the application of a mask alone or prior to intubation. Influenza and pneumonia (ICD-10-GM group J09-J18) as well as other acute lower respiratory tract infections (ICD-10-GM group J20-J22) were grouped together as severe acute respiratory diseases (SARI) [3]. The death was determined based on the reason for discharge and, in this study, related to the inpatient treatment case.
To confirm the descriptive results, a matched pairs analysis was carried out. Cases from the second wave were assigned to cases from the first wave. Matching criteria were the chapter of the main diagnosis, gender, length of stay, age at admission, patient-related overall severity score (PCCL), whether ventilation was performed and the presence of SARI. A deviation of up to three days was tolerated for the length of stay, up to five years for the age and a value of 1 for the PCCL. An exact match was required for all other criteria. Each case from the first wave was assigned exactly one case from the second wave; A case from the second wave was only taken into account once (1:1 matching without replacement). Both groups were compared two-sidedly using a t-test or Fisher's exact test.
The terms treatment case, case and patient are used synonymously.
Results
778,020 treatment cases were included. A total of 7,589 cases suffered from Covid-19. A multisystem inflammatory syndrome associated with Covid-19 was present in 13 cases; Of these, only code U07.5 was recorded in four cases. 1,759 cases were assigned to the first wave and 5,830 cases to the second wave (see Table 1).
The second wave showed a shorter length of stay and a shorter stay in the intensive care unit. Patients with Covid-19 were ventilated less frequently in the second wave (10.7% versus 16.6%) and for a shorter period of time (182 versus 380 hours). In the second wave, a non-invasive ventilation attempt with 67.0% was carried out significantly more frequently than in the first wave with 22.8%. The patients in the first wave were more severely ill (PCCL, comorbidity index [4]) and suffered more often from SARI. In-hospital mortality was identical in both waves at 17.6%; There was also no difference in mortality between the first and second waves in the age groups (see Table 2). At 31.2% (first wave) and 31.5% (second wave), mortality was highest in the age group aged 76 and over.
In ventilated patients, mortality was higher in the second wave at 51.8% than in the first wave at 42.1%. In contrast, the comorbidity index in mechanically ventilated patients in the second wave was 36.6 ± 17.06, lower than in the first wave with 41.3 ± 17.53.
Figure 1 shows the progression of mortality and number of patients with Covid-19 by month of admission. Mortality then fell to a very low level as the first wave subsided, then gradually increased in the second wave and reached the first wave peak of 23% again in December 2020.
For the matched pairs analysis, 1,438 patients from the first wave could be assigned a control from the second wave (81.8% of 1,759 cases from the first wave). There were no differences between the two groups in the matching criteria with unclear assignment, length of stay (p=0.709), age at admission (p=0.999) and PCCL (p=0.824). In-hospital mortality was almost identical (p=0.961) for all patients with Covid-19 at 17.4% (first wave) and 17.2% (second wave), with 51.4% (first wave) and 50.8% (second wave). wave) as well as the mortality of ventilated patients with Covid-19 (p=1.0). The duration of ventilation in the second wave was also shorter in this analysis (201 versus 268 hours, p=0.002) and the proportion of non-invasive ventilation attempts was higher (69.1% versus 25.3%, p<0.001). The disease severity determined using the comorbidity score according to [4] tended to be lower in the second wave in the ventilated cases at 36.0 ± 15.61 than in the first wave at 39.3 ± 17.67 (p = 0.56 ). The difference in the descriptive analysis of all cases with Covid-19 was not confirmed after matching with a comorbidity index of 18.1 ± 17.80 (first wave) and 18.3 ± 17.01 (second wave) (p = 0.734).
discussion
The differences between the first and second waves of the corona pandemic indicate a change in the ventilation strategy. In accordance with the recommendations in the national S3 guideline [2], the indication for ventilation was made more cautious and non-invasive ventilation with a mask was attempted more frequently when mechanical ventilation was indicated. The duration of mechanical ventilation was also significantly shorter. At the same time, the severity of the disease was lower, although this may be due to fewer side effects from less intensive ventilation. It is therefore surprising that mortality in the second wave remained high at 17.6%. Of the ventilated patients, more died in the second wave with over 50% than in the first wave, although ventilation was more often non-invasive. However, the comorbidity of the ventilated patients was lower in the second wave than in the first wave. Data from the Institute for the Hospital Remuneration System (InEK) also reports an unchanged high mortality rate for ventilation in the intensive care unit in the second wave in Germany with over 50% [5].
The decline in mortality in German university hospitals from January to April 2020 to May to September 2020 [6], which is described as the success of optimizing treatment, is therefore probably due to the choice of summer 2020 as the comparison period. Figure 1 shows that mortality from Covid-19 was significantly associated with the frequency of Covid-19. Seasonal aggressiveness of the virus [7] and lockdown together may have affected both infectivity and disease severity. It therefore seems sensible to reconsider the recommendations regarding the indication and implementation of mechanical ventilation. According to the results presented, a more conservative ventilation strategy may not be associated with survival benefits. However, the lower proportion of cases with SARI and the lower comorbidity in the second wave may indicate that undesirable consequences of mechanical ventilation can be avoided by changing the indication.
A network of four Bavarian hospitals also reported that the therapy strategy had been changed to favor non-invasive ventilation [8]. There, however, mortality in the first wave with 526 patients included was around 26%, in the second wave with 618 patients included around 18%. Mortality in the first wave was thus significantly higher than the level of 21% reported from data from the general local health insurance fund (AOK) [9]. The decrease found in [8] is therefore possibly due to an unusually high mortality rate in the first wave, which may be due to the specific care circumstances of the four clinics. The success of therapy optimization cannot be derived from the reduction in mortality.
The sample of the presented comparison between the first and second waves of the corona pandemic consisted of 63 small and medium-sized hospitals run by the church. The increase in the number of treatments from the first to the second wave illustrates the importance of caring for patients with Covid-19 beyond maximum care providers. The need to build up and maintain intensive care resources with the ability to provide artificial ventilation at all levels of care was already pointed out after the first wave [10].
The results are limited, among other things, by the lack of information on other treatments for Covid-19, in particular on drug therapy. The lack of a decrease in mortality in hospital with Covid-19 can therefore also be due to other unsuccessful changes in therapy. Given the significantly higher occupancy of hospitals due to Covid-19, neither overload, e.g. B. due to staff shortages in intensive care units, nor triage, e.g. B. can be excluded by a more frequent transition to palliative care [11]. This would undermine potential success in optimizing treatment concepts. According to the matched-pairs analysis, the shorter observation period for inpatient admissions of the second wave does not explain the results.
With the comparison made between the first and second wave, the mortality rate in the hospital in all cases of treatment with Covid-19 could be related to a changed ventilation strategy. In view of the results, the current challenge for the treating facilities is to offer patients with Covid-19 individual care when the evidence is weak. <<