Using nationwide inpatient billing data from the InEK and information on mechanical ventilation from the DIVI, we first aimed to determine the COVID-19 intensive care costs in Germany over the course of the pandemic and, second, to assess the economic impact of COVID-19 intensive care from the perspective of the German health care system. In doing so, we calculated the intensive care costs incurred for the nonventilated and ventilated ICU bed days of COVID-19 patients from January 1, 2020, to March 31, 2022, and related these to SHI expenditures and the cost of illness for the German hospital sector. In this respect, we emphasize that our cost calculation served as an approximation for the respective SHI expenses, as their actual reimbursement of COVID-19 intensive care may have been different. Furthermore, COVID-19 intensive care patients may well have had a main diagnosis other than a respiratory disease, which we generally assumed in our study, also in view of the increasing number of viral pneumonia cases (ICD-10 code J12) (18, 23).
The COVID-19 intensive care costs we determined for Germany were €776 million in 2020 and €1.7 billion in 2021, representing a relative cost increase of 220%. Compared to the total annual SHI expenditures on hospital treatment of €81.5 billion and €85.9 billion, our findings were less than 2.0% (9). On a quarterly basis, our cost figures ranged slightly above this percentage, particularly at the turn of the year, which appeared to be in line with the large increase in COVID-19 intensive care patients during the two winter seasons captured (24–27). The rather low SHI expenditures for COVID-19 intensive care, which we have derived (one-to-one) from our cost figures, seemed quite plausible given a total of 17.3 million hospitalizations in 2020 and 17.2 million in 2021, of which less than 10% were COVID-19 intensive care patients in each case according to our results (28). For further comparison, inpatient treatments for neoplasms, injuries and diseases of the circulatory system were nearly or more than twice as common as for respiratory diseases, to which we assigned all COVID-19 intensive care patients (29, 30).
Therefore, the COVID-19 intensive care costs we determined for 2020 (€776 million) amounted to 11% of the cost of respiratory disease illness for hospitals. When further narrowed down to pneumonia, whose cost of illness amounted to approximately €2.8 billion, the share was as high as 30% (10). However, respiratory diseases only accounted for around 6.0% of total cost of illness for German hospitals and pneumonia for just under 2.5%. Given that all COVID-19 intensive care patients had a main diagnosis of respiratory disease, this could also suggest that the corresponding SHI expenditure was rather low compared to that of other ICD-10 diagnosis groups. We would like to point out once again that although this restrictive allocation made it possible to calculate COVID-19 intensive care costs, it certainly did not apply to every patient. Although more recent illness figures for 2021 are not yet available for comparison, the immense cost increase to €1.7 billion (+ 220% compared to 2020) revealed the growing health economic impact of COVID-19 intensive care, which was due to more than doubling the number of patients and the still high proportion of patients receiving mechanical ventilation (5). The sharp increase in the number of diagnosed viral pneumonia cases, which more than doubled to over 200,000 in 2021, also appeared to substantiate this (23).
In summary, the cost share of COVID-19 intensive care in SHI expenditures for hospital treatment in 2020 and 2021 appeared to be rather low. If the respective expenditures of the private health insurance funds of €8.4 billion and €8.3 billion had also been considered, this cost share would have been even lower (31). A similar picture emerged when COVID-19 intensive care costs were compared to the cost of illness in the German hospital sector. However, the sheer number of critically ill COVID-19 patients has become a rapidly growing cost factor, with the frequent use of mechanical ventilation, in particular, increasing treatment costs (6, 11–13). According to our results, a total of 2.2 million ICU bed days, including 1.2 million (55%) ventilated ICU bed days, were spent by German hospitals on COVID-19 patients until the end of March 2022, for which the required healthcare resources, such as ICU beds, ventilators, medical and nursing staff, etc., were (to some extent) used at the expense of care for other patients (14). In this context, the costs incurred by the German health care system due to the postponement of elective procedures to provide free ICU capacity for COVID-19 patients must also be considered, as hospitals, for example, were remunerated with flat rates for providing free COVID-19 ICU capacity (15, 16). Hence, our cost calculation did not take into account that COVID-19 intensive care patients may have displaced treatment for non-COVID-19 patients. From the hospital's perspective, such a displacement scenario would mean that overall costs increased only by the additional variable costs of treating COVID-19, assuming these exceed the variable costs compared to a "traditional" patient. However, from a societal perspective, the overall cost of this displacement scenario would be much greater, as displaced patients may have suffered a substantial health loss (for example, through the postponement of elective procedures), which also entailed costs. Unfortunately, in this study, we were unable to quantify the degree of displacement or its financial consequences. Instead, our cost calculation implicitly assumed that COVID-19 intensive care patients were 100% additional to “traditional” patients. This assumption may seem unrealistic in its entirety, but it was necessary to obtain the COVID-19-related intensive care costs in Germany using the chosen approach. Ultimately, it seems adequate to assume that the societal costs of a displaced non-COVID-19 patient exceeded those of the refused treatment. Therefore, we believe that the direct treatment costs presented here should be interpreted as the lower bound because the displacement scenario was not captured.
Nevertheless, we would like to substantiate our approach to determining COVID-19 intensive care costs in Germany, which was essentially based on the study by Kaier et al., whereby the economic valuation of an ICU day is based on the cost accounting method following the InEK handbook of calculation (13). According to this method, standardized unit costs per inpatient case are defined on the basis of the average expenditure (within the cost categories of staff, material and infrastructure) for an ICU day. In doing so, Kaier et al. included more than 10,500 adult intensive care patients to explicitly address the cost impact of mechanical ventilation per ICU day. Using the same ‘bottom-up’ method for cost accounting, a study by Schallner et al. reported a mean cost of €3,720 per ICU day for COVID-19 patients, which is more than twice as high as the cost figures per ventilated ICU day that we referred to (see Supplementary Table S1) (7). However, the number of included COVID-19 intensive care patients was relatively low at 49, almost half of whom also received ECMO therapy, which may further have increased the cost of treatment. Furthermore, no explicit distinction was made according to the use of mechanical ventilation, which was of central importance to our cost calculation, as it was frequently used in COVID-19 patients (mostly above 50% of patients nationwide during our study period) (5). A competing method of cost accounting is the willingness-to-pay (WTP) approach, which uses questionnaires to assess the WTP to liberate ICU beds given (or contingent on) a set of hypothetical conditions (32). The resulting amount of costs per ICU day then reflects the mean amount respondents would pay to release an additional ICU bed. Typically, WTP-based estimates are substantially lower than the accounting-based costs we used. On the other hand, those estimates usually represent the average WTP under low-, medium-, and high-demand scenarios. WTP-based estimates under pandemic conditions, where the demand for ICU beds has increased dramatically, are not currently available. However, as the per-case flat rates of the G-DRG system are based on a standardized cost accounting scheme for defined medically and economically homogeneous groups, the WTP approach appears to be less reliable for determining the COVID-19 intensive care costs actually billed by German hospitals (8).
A key strength of the present study is the representativeness of the InEK data, which includes inpatient billing cases from all hospitals in Germany that are subject to data transmission obligations according to the Hospital Remuneration Act (17). In addition, we relied on the study by Kaier et al., which explicitly distinguished between nonventilated and ventilated ICU bed days for cost estimation (13). This allowed us to account for the cost impact of mechanical ventilation while incorporating the clinical severity of COVID-19 in intensive care patients. One limitation, however, is that we calculated our cost figures based on ICU treatments in 2013, while the G-DRG-based cost accounting standard changed from 2020 onward, with nursing staff costs being separated from the per-case flat rates and financed through a hospital-specific nursing budget based on cost coverage (33). Furthermore, our cost figures only reflect direct COVID-19-related treatment, without considering the potential costs of displaced treatment for non-COVID-19 patients. Overall, more accurate cost figures would have required case-specific billing data for the entire course of the COVID-19 pandemic.