*4.6. Methodological Characteristics and Limitations of NGS*

The relevance of the microorganisms detected by NGS often remains unclear. Clinical experience and treatment recommendations are limited or lacking. Similar to other molecular genetic detection methods, it is uncertain whether the detection of cfDNA corresponds to clinically relevant infection. The genomic material obtained could originate from non-viable or commensal microorganisms and, therefore, might lead to false positive results and mimic an active infection. Furthermore, molecular techniques, currently do not allow for antimicrobial susceptibility testing [16]. In our institution, NGS currently offers no advantage over culture-based diagnostics in terms of turn-around time for logistic reasons. Improved workflows that might provide faster results in the future are currently

under investigation [23]. Results of prospective studies showing a positive effect of NGS on clinical outcomes and, thus, justifying the additional costs are still pending [22].

Identifying microorganisms of uncertain clinical relevance and lack of antimicrobial susceptibility testing can result in antimicrobial overtreatment and prevent de-escalation and rational use of antibiotics. These technical limitations demonstrate that NGS-based analysis cannot be used as a substitute for cultural methods, but rather may be considered as a complementary test in patients with severe COVID-19.

### *4.7. Limitations*

This study suffers from numerous limitations, which are mainly attributable to the retrospective study design and the limited cohort. As such, the study design was not suitable to investigate the impact of additional NGS-based diagnostics on morbidity and mortality. A very specific subgroup of patients with severe COVID-19 treated in the ICU was examined, consequently, our results cannot be generalized to other patient groups. The lack of routine screening for common viral infections impedes any direct comparison of the two methods, while sample collection from different sites could lead to divergent results between the methods. Significant differences in demographic variables, such as age or pre-existing diseases, could constitute confounding factors and the small number of individual pathogens precluded an analysis of the association between read count, outcome, and clinical relevance. Despite a comprehensive review of clinical data and documentation, important information may have been unavailable, potentially biasing the evaluation of treatment decisions.

#### **5. Conclusions**

The results of our study suggest that NGS-based diagnostics might offer a higher positivity rate than conventional culture-based methods and, therefore, may enable new therapeutic approaches in critically ill COVID-19 patients. However, further experience regarding the interpretation of the results is required and treatment decisions should be carefully considered to avoid overtreatment. Larger, prospective studies will be necessary to determine whether the identification of additional pathogens by NGS can improve the outcome of critically ill ICU patients with severe COVID-19.

**Author Contributions:** Conceptualization, C.J.L. and F.D.; Investigation, C.J.L., S.E.S., T.K., H.S. and F.D.; Supervision, F.D.; Writing—original draft, C.J.L.; Writing—review & editing, S.E.S., W.A.W., A.M., B.W.B., H.S. and F.D. All authors have read and agreed to the published version of the manuscript.

**Funding:** We acknowledge support for the Article Processing Charge from the DFG (German Research Foundation, 491454339).

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Ethics Committee of the Medical Faculty of the University of Cologne (Reference No. 21-1444).

**Informed Consent Statement:** Patient consent was waived due to its retrospective design.

**Data Availability Statement:** The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.

**Acknowledgments:** The authors thank Noscendo GmbH for performing the NGS analysis and for providing expertise on general technical inquiries. We also thank Susanna Dreißig and Samuel Steinbach for their help with data collection.

**Conflicts of Interest:** The authors declare no conflict of interest.
