Infection https://doi.org/10.1007/s15010-023-02001-2 RESEARCH Diabetes mellitus is associated with 90‑day mortality in old critically ill COVID‑19 patients: a multicenter prospective observational cohort study Timo Mayerhöfer1 · Sebastian Klein1 · Bernhard Wernly2,3 · Hans Flaatten4 · Bertrand Guidet5 · Dylan W. De Lange6 · Jesper Fjølner7 · Susannah Leaver8 · Michael Beil9 · Sigal Sviri9 · Raphael Romano Bruno10 · Antonio Artigas11 · Peter Vernon van Heerden12 · Bernardo Bollen Pinto13 · Joerg C. Schefold14 · Rui Moreno15 · Maurizio Cecconi16,17 · Wojciech Szczeklik18 · Christian Jung10 · Michael Joannidis1 on behalf of COVIP study group Received: 2 October 2022 / Accepted: 7 February 2023 © The Author(s) 2023 Abstract Background Several studies have found an association between diabetes mellitus, disease severity and outcome in COVID- 19 patients. Old critically ill patients are particularly at risk. This study aimed to investigate the impact of diabetes mellitus on 90-day mortality in a high-risk cohort of critically ill patients over 70 years of age. Methods This multicentre international prospective cohort study was performed in 151 ICUs across 26 countries. We included patients ≥ 70 years of age with a confirmed SARS-CoV-2 infection admitted to the intensive care unit from 19th March 2020 through 15th July 2021. Patients were categorized into two groups according to the presence of diabetes mellitus. Primary outcome was 90-day mortality. Kaplan–Meier overall survival curves until day 90 were analysed and compared using the log-rank test. Mixed-effect Weibull regression models were computed to investigate the influence of diabetes mel- litus on 90-day mortality. Results This study included 3420 patients with a median age of 76 years were included. Among these, 37.3% (n = 1277) had a history of diabetes mellitus. Patients with diabetes showed higher rates of frailty (32% vs. 18%) and several comorbidities including chronic heart failure (20% vs. 11%), hypertension (79% vs. 59%) and chronic kidney disease (25% vs. 11%), but not of pulmonary comorbidities (22% vs. 22%). The 90-day mortality was significantly higher in patients with diabetes than those without diabetes (64% vs. 56%, p < 0.001). The association of diabetes and 90-day mortality remained significant (HR 1.18 [1.06–1.31], p = 0.003) after adjustment for age, sex, SOFA-score and other comorbidities in a Weibull regression analysis. Conclusion Diabetes mellitus was a relevant risk factor for 90-day mortality in old critically ill patients with COVID-19. Study registration NCT04321265, registered March 19th, 2020. Keywords SARS-CoV-2 · Intensive care unit · Elderly · Old · Critical care · Ventilation · Risk factors Abbreviations Background CHF Chronic heart failure CKD Chronic kidney disease SARS-CoV-2 leads to various manifestations, ranging from DM Diabetes mellitus asymptomatic infections, to mild symptoms or severe illness ICU Intensive care unit and even death [1]. During the COVID-19 pandemic inten- MV Mechanical ventilation sive care units were heavily affected worldwide, and sev- eral studies have investigated risk factors for a severity and a negative outcome [2, 3]. In addition to well-established vaccinations against COVID-19 [4, 5], antiviral drugs and The members of the COVIP study group are mentioned in the monoclonal antibodies are now available to prevent a severe Acknowledgements section. course of COVID-19 [6–8]. However, it is still important to * Michael Joannidis identify risk factors associated with a worse outcome. michael.joannidis@i-med.ac.at Age and frailty appear to be important predictors of Extended author information available on the last page of the article critical illness and mortality in patients infected with 13 Vol.:(0123456789) T. Mayerhöfer et al. SARS-CoV-2 [9, 10]. Furthermore, various comorbidities patient began on admission to the ICU. The day of admission including obesity, chronic kidney disease (CKD) and dia- was defined as the first day, and all subsequent days were betes mellitus (DM) have also been identified as significant numbered consecutively. For this analysis, all critically ill risk factors [10]. patients ≥ 70 years with known DM status (yes or no) were Multiple studies have shown that patients with type 2 DM included (definitions of comorbidities and treatment limita- have an increased risk for hospitalisation, severe disease, tions in additional file 2). We excluded 507 patients with intensive care unit (ICU) admission, longer length of stay, unknown DM status. and mortality in COVID-19 [10–14]. Meta-analyses con- firmed an independent association between DM with a more Data collection than two-fold increased risk for severe disease and mortality [15, 16]. For DM, the first ICU cohort studies from Wuhan All participating centres used a standardised online elec- suggested a rate of around 20% of patients with known DM tronic case report form, that recorded baseline characteristics in ICUs [17]. Higher rates of about 30% have been reported (Table 1), the sequential organ failure assessment (SOFA) in Europe and the USA [18–20]. Type 2 DM is more com- score at admission, the need for non-invasive or invasive mon in advanced age [21]. Concomitantly, the prevalence of ventilation, prone positioning, tracheostomy, use of vaso- other comorbidities also increases with age, and frailty also pressors and renal replacement therapy and any restriction of becomes more predominant. The specific impact of each risk life-sustaining treatment during the ICU stay. The database factor is therefore difficult to assess. This is especially true ran on a secure server set up at Aarhus University, Denmark. for old critically ill patients, where other comorbidities and frailty are more prevalent than in younger individuals. The Outcomes COVID-19 disease in Very Elderly Intensive care Patients (COVIP) study focuses on prognostic factors and outcome The primary outcome was 90-day mortality, the secondary of old ICU patients (≥ 70 years) with COVID-19. Whether outcomes were overall survival to ICU discharge, survival at DM is also a relevant risk factor in this vulnerable cohort 30 days after ICU admission and ICU length of stay. remains unclear. Therefore, we aimed to investigate the influence of DM on outcomes in a cohort of already vulnerable old critically Statistical analysis ill COVID-19 patients. Kolmogorov–Smirnov test was used to test for normal distri- bution. Differences between the two groups were calculated Methods using the Mann–Whitney U test and Chi-squared test for not normally distributed data and Student’s t-test for nor- Study design, setting and participants mally distributed data. The study population was divided into patients with and without DM. This study is a secondary analysis of the ongoing COVIP- Mixed-effects Weibull regression analysis was applied trial (NCT04321265), a multicentre international observa- with the center as a random effect and DM as fixed effects to tional prospective study that includes critically ill patients evaluate DM as a possible independent predictor for 90-day with COVID-19 over or equal to 70 years of age and is part mortality. For a better estimation of the impact of DM on of the Very old Intensive care Patients (VIP) project (www.​ outcome three different models were created (Model A, B vipst​udy.​org). A list of collaborators is shown in the Addi- and C). Model A included age, sex and the SOFA score, to tional file 1. This study included participants from 151 ICUs account for disease severity. In Model B frailty was added across 26 countries in Europe, Asia, Africa and America. In as a recently established risk factor for mortality in COVID- all centres, ethical approval was required for participation 19 patients [22], and in Model C, all other comorbidities and in most countries informed consent was obtained for obtained in the COVIP study according to the protocol were enrollment in the study, in accordance with local regulations. added as additional variables. Adjusted hazard ratios with This study was conducted in accordance with the Helsinki corresponding 95% confidence intervals (95% CI) were cal- Declaration. The detailed methods of this study were pub- culated. Furthermore, 90-day mortality was analysed using lished previously [22]. Kaplan–Meier curves for patients with and without DM. Dif- In brief, patients with a positive polymerase-chain-reac- ferences were assessed using the log-rank test. tion for SARS-CoV-2, who were 70 years of age or older To further investigate the influence of DM on 90-day and admitted to an ICU were eligible for the study. For this mortality in certain subgroups, we created a forest plot analysis, we included all records in the database from 19th depicting univariable hazard ratios for patients with DM in March 2020 through 15th July 2021. Data collection for each each subgroup. 13 Diabetes mellitus is associated with 90‑day mortality in old critically ill COVID‑19 patients:… Table 1  Patient characteristics No diabetes Diabetes p-value of the study population with and without diabetes mellitus n = 2143 n = 1277 Characteristics Sex 0.12 Male, % (n) 70% (1505) 68% (864) Female, % (n) 30% (638) 32% (413) Age at admission 76 (5) 76 (5) 0.91 Age at admission 0.70 Age < 80 years, % (n) 78% (1662) 78% (998) Age > 79 years, % (n) 22% (480) 22% (279) BMI 28 (5) 30 (6) < 0.001 SOFA score on admission 5 (3) 6 (3) < 0.001 Comorbidities Ischemic heart disease, % (n) 19% (405) 31% (393) < 0.001 Chronic kidney disease, % (n) 11% (238) 25% (319) < 0.001 Arterial hypertension, % (n) 59% (1263) 79% (1005) < 0.001 Pulmonary comorbidity, % (n) 22% (459) 22% (274) 0.92 Chronic heart failure, % (n) 11% (241) 20% (249) < 0.001 Clinical Frailty Scale Score 3 (2) 4 (2) < 0.001 Presence of Frailty (CFS > 4) < 0.001 No frailty, % (n) 82% (1604) 68% (767) Frailty, % (n) 18% (345) 32% (363) Therapeutic measures Intubation and mechanical ventilation, % (n) 71% (1512) 67% (854) 0.024 Non-invasive ventilation, % (n) 25% (535) 30% (375) 0.003 Tracheostomy, % (n) 19% (399) 15% (185) 0.002 Vasoactive drugs, % (n) 67% (1431) 63% (789) 0.004 Corticosteroids, % (n) 68% (1460) 74% (944) 0.003 Renal Replacement Therapy, % (n) 13% (286) 18% (224) < 0.001 Treatment limitations Any treatment limitation, % (n) 36% (772) 32% (398) 0.006 Life sustaining care withheld, % (n) 29% (625) 28% (347) 0.25 Life sustaining care withdrawn, % (n) 20% (426) 17% (210) 0.015 CFS clinical frailty scale All tests were two-sided, and a p value of < 0.05 was con- diagnosis of DM was present in 1277 (37.3%) patients. The sidered statistically significant. Some patients were excluded median age of the total cohort was 76 years and similar in for subgroup analyses due to missing values. For this reason, both groups (p = 0.91). The same applies to sex distribution: not all patient numbers add up to 100%. among patients with and without DM 70% were men and Stata 17 was used for all statistical calculations (Stata- 30% women. Corp LLC, 4905 Lakeway Drive, College Station, Browns- ville, Texas, USA). Comorbidities Patients with DM had a higher median BMI (30 kg/m2 Results vs. 28 kg/m2, p < 0.001) and a higher SOFA score at ICU admission (6 vs 5, p < 0.001), compared to patients with- Patients out the diagnosis. The most common comorbidities such as ischemic heart disease, CKD, hypertension and chronic heart In total, 3420 critically ill patients aged ≥ 70 years from the failure (CHF) were significantly more frequent in patients COVIP study were included in this analysis. Baseline char- with DM. Furthermore, these patients had a higher median acteristics are presented in detail in Table 1. A pre-existing score on the clinical frailty scale and a higher proportion was 13 T. Mayerhöfer et al. Table 2  Outcome of patients with and without diabetes mellitus No diabetes diabetes p-value n = 2143 n = 1277 Outcome Mortality at 30 days, % (n) 51% (1092) 59% (751) < 0.001 Mortality at 90 days, % (n) 56% (1194) 64% (811) < 0.001 Mortality at ICU discharge, 49% (1037) 55% (699) < 0.001 % (n) ICU length of stay (h) 512 (1188) 455 (1297) 0.19 ICU Intensive Care Unit known to be frail (32% vs. 18%, p < 0.001). No difference was observed in the frequency of pulmonary comorbidities. Fig. 1  Kaplan–Meier survival analysis displaying the survival prob- ability in relation to the diabetes mellitus status at admission. CI con- Therapeutic measures during ICU stay fidence interval In 71% (n = 1512) of patients without DM and in 67% (n = 854) of patients with DM MV was used (p = 0.024). with CHF or over 80 years of age, no association was found Tracheostomy was performed more frequently in patients between 90d mortality and the presence of DM (Fig. 2). without DM (19% vs. 15%, p = 0.002), as was the use of vas- oactive drugs (67% vs. 63%, p = 0.004). By contrast, renal replacement therapy was used significantly more often in Discussion patients with DM (18% vs. 13%, p < 0.001). Treatment limitations were less common in patients with In this large multicentre cohort study of old critically ill DM (32% vs 36%, p = 0.006, Table 2) and the ICU length COVID-19 patients, we observed a high rate of DM, with of stay was not significantly different in patients with or an independent association with 90-day mortality. The effect without DM. was more pronounced in younger individuals (70–79 years) and in patients without frailty, CKD and CHF. However, patients with DM had a significantly higher rate of comor- Outcome bidities. These comorbidities are usually considered to be a consequence of diabetes-related end-organ damage. The 90-day mortality was significantly higher in patients Understanding the most important factors for an unfa- with DM than in patients without DM (Table 2). The same vorable outcome in critically ill COVID-19 patients is cru- applies to ICU and 30-day mortality. The ICU length of stay cial. Older individuals are more commonly affected by a was not significantly different in patients with or without severe course of disease [23–26]. However, age is not suit- DM. As illustrated by the Kaplan–Meier analysis, DM was able as the sole factor for predicting outcome in critically ill significantly associated with an impaired survival probabil- COVID-19 patients. Therefore, knowledge about the impact ity (Fig. 1). In the univariable analysis, DM mellitus was a of other comorbidities is of great importance, especially in significant predictor of 90-day mortality with a hazard ratio an old population. of 1.29 (95% CI 1.18–1.41, p < 0.001). Several studies have shown that COVID-19 patients with To evaluate the relevance of DM in the prediction of DM have a worse outcome than patients without DM. In an 90-day mortality in very old critically ill patients, we cre- early meta-analysis of 33 studies, Kumar et al. found a more ated three different models (see Table 3). The association than two-fold increased risk of mortality and an increased between DM and 90-day mortality was also significant after severity of disease in COVID-19 patients with DM when adjustment for age, sex and the SOFA score in Model A. compared to those without DM [15]. Since DM is often When frailty or other comorbidities were added to the model associated with multiple other comorbidities, it remained (Model B and Model C), DM remained a significant but unclear whether this association was independent of other weaker predictor for mortality. factors. A recently published national cohort study from The univariate hazard ratio for DM was higher in the sub- England showed that DM is indeed an independent predictor group of patients without frailty, without CHF and between the age of 70 and 79. However, for patients with frailty, 13 Diabetes mellitus is associated with 90‑day mortality in old critically ill COVID‑19 patients:… Table 3  Mixed-effects Weibull proportional hazard regression analysis for prediction of 90-day mortality Model A Model B Model C HR (95% CI) p value HR (95% CI) p value HR (95% CI) p value Diabetes 1.24 (1.14–1.38) < 0.001 1.17 (1.06–1.30) 0.002 1.18 (1.06–1.31) 0.003 Age 1.06 (1.05–1.73) < 0.001 1.06 (1.04–1.07) < 0.001 1.06 (1.04–1.06) < 0.001 Sex 0.97 (0.88–1.08) 0.582 0.92 (0.83–1.03) 0.155 0.94 (0.84–1.05) 0.268 SOFA Score 1.14 (1.13–1.16) < 0.001 1.13 (1.11–1.15) < 0.001 1.13 (1.11–1.15) < 0.001 Frailty 1.82 (1.60–2.06) < 0.001 1.69 (1.48–1.93) < 0.001 Ischemic heart disease 1.06 (0.93–1.20) 0.387 Chronic kidney disease 1.33 (1.16–1.52) < 0.001 Arterial hypertension 0.85 (0.76–0.95) 0.005 Pulmonary comorbidity 1.11 (0.99–1.27) 0.071 Chronic heart failure 1.09 (0.93–1.27) 0.296 HR hazard ratio, CI confidence interval, SOFA Sequential Organ Failure Assessment, Center as random effect and Diabetes as fixed effects for mortality [27]. However, the transferability of results due to DM they cannot be seen as factors independent of from a national study to other countries may be limited. DM. The same is true for frailty which was roughly twice DM was one of the most frequent comorbidities in our as frequent in our critically ill patients with DM. Frailty has population at 37%. Compared to other cohort studies of been proven to be a relevant prognostic marker for outcomes critically ill patients, this number is slightly higher [18, in critically ill patients with COVID-19 [22] as well as with- 23, 28, 29]. Our study refers only to old individuals over out COVID-19 [33]. 70 years of age, in whom DM generally has a greater prev- Thus, we may infer, that although frailty and other comor- alence than in the overall population [21]. In support of bidities seem to mask the influence of DM in older patients, our data, a cohort study from Germany with a compara- they might also be a consequence of DM itself and therefore tively old study population yielded similar numbers [25]. be interrelated with this underlying disease. Our multicentre study now provides evidence that even As recently reported treatment limitations were more fre- in a highly vulnerable cohort of old critically ill patients quent in old COVID-19 patients compared to patients without the presence of DM is associated with higher 90-day COVID-19 disease [34]. Varying rates of treatment limitations mortality. Even when adjusting for age, frailty and other in patients with and without DM might have the potential to comorbidities, DM remained independently associated affect the outcome. However, in our study, treatment limita- with mortality. However, since age and frailty are such tions were even less common in patients with DM. Therefore, important characteristics [23, 30], the influence of DM they cannot be responsible for the higher mortality observed was reduced. in this group. The impact of DM was more prominent in younger In addition, DM, especially type 2 DM, is further associated patients in our cohort (70–79 years). A retrospective study with obesity [35]. In our study patients with DM had a higher from Mexico and another study from France made similar body mass index than patients without DM. Obesity and DM observations, describing a decrease in the association of DM negatively affect the outcome during infectious diseases by and death with increasing age, although their cohorts started altering the immune response, which explains the poor out- at a much younger age [31, 32]. For patients age 80 years and come in these patients [36]. above the influence of DM seems to be reduced in relation The main strengths of this study are its multicentre study to other comorbidities. When comparing the prevalence of design and a large number of patients included. However, comorbidities in our cohort, several diseases such as CKD, some limitations such as the observational design of this study cardiovascular disease and CHF showed rates which were must be considered when interpreting the results. First, some between 1.5 and 2 times higher for patients with DM com- important values are missing such as blood glucose levels or pared to those without. CKD and CHF are both risk factors HbA1c. These values might provide additional information for a severe course of COVID-19 [10]. Since these diseases especially on glycemic control, which was associated with out- are considered typical manifestations of end-organ damage come in COVID-19 patients in another study [37]. Second, we 13 T. Mayerhöfer et al. Fig. 2  Forest plot for univari- able hazard ratios for patients with DM in different subgroups. CI confidence interval, aHR adjusted hazard ratio, aHT arterial hypertension, ASCVD atherosclerotic cardiovascular disease, BMI body mass index, CHF chronic heart failure, CKD chronic kidney disease, MV mechanical ventilation did not differentiate between type 1 and type 2 DM. Since type Conclusion 2 DM is much more prevalent in older individuals [21] and patients with type 1 DM have a reduced life expectancy [38], DM is independently associated with an impaired outcome it is likely that nearly all patients with DM in our study cohort in old critically ill COVID-19 patients. The influence of had type 2 DM. However, it cannot be completely excluded, DM was strongest in “younger” patients without frailty. that some patients with type 1 DM influenced the results. DM might be of additional value for risk prediction even Moreover, there is a small risk of selection bias, since 507 in old critically ill COVID-19 patients who are already at patients were excluded due to missing DM status. high risk of a poor outcome. 13 Diabetes mellitus is associated with 90‑day mortality in old critically ill COVID‑19 patients:… Supplementary Information The online version contains supplemen- BW. All authors participated in the interpretation of the results. TM, tary material available at https://d​ oi.o​ rg/1​ 0.1​ 007/s​ 15010-0​ 23-0​ 2001-2. MJ, SK, CJ and BG wrote the initial draft of the manuscript. All other authors revised the manuscript and had access to the final version. Acknowledgements The authors want to thank all investigators and study personal for their great support of the study. The COVIP study Funding Open access funding provided by University of Innsbruck and group consists of the authors and the following persons: Philipp Eller, Medical University of Innsbruck. The support of the study in France by Michael Joannidis, Dieter Mesotten, Pascal Reper, Sandra Oeyen, Wal- a grant from Fondation Assistance Publique-Hôpitaux de Paris pour la ter Swinnen, Helene Brix, Jens Brushoej, Maja Villefrance, Helene recherche is greatly appreciated. In Norway, the study was supported Korvenius Nedergaard, Anders Thais Bjerregaard, Ida Riise Balleby, by a grant from the Health Region West. In addition, the study was Kasper Andersen, Maria Aagaard Hansen, Stine Uhrenholt, Helle Bun- funded by a grant from the European Open Science Cloud (EOSC) by dgaard, Jesper Fjølner, Aliae AR Mohamed Hussein, Rehab Salah, the European Commission. No further specific funding was received. Yasmin Khairy NasrEldin Mohamed Ali, Kyrillos Wassim, Yumna A. Elgazzar, Samar Tharwat, Ahmed Y. Azzam, Ayman abdelmawgoad Availability of data and materials Individual participant data that Habib, Hazem Maarouf Abosheaishaa, Mohammed A Azab, Susannah underlie the results reported in this article are available to investiga- Leaver, Arnaud Galbois, Bertrand Guidet, Cyril Charron, Emmanuel tors whose proposed use of the data has been approved by the COVIP Guerot, Guillaume Besch, Jean-Philippe Rigaud, Julien Maizel, Michel steering committee. Djibré, Philippe Burtin, Pierre Garcon, Saad Nseir, Xavier Valette, Nica Alexandru, Nathalie Marin, Marie Vaissiere, Gaëtan Plantefeve, Declarations Thierry Vanderlinden, Igor Jurcisin, Buno Megarbane, Anais Caillard, Arnaud Valent, Marc Garnier, Sebastien Besset, Johanna Oziel, Jean- Conflict of interest JCS (full departmental disclosure:) report grants herlé Raphaelen, Stéphane Dauger, Guillaume Dumas, Bruno Gon- from Orion Pharma, Abbott Nutrition International, B. Braun Medical calves, Gaël Piton, Christian Jung, Raphael Romano Bruno, Malte AG, CSEM AG, Edwards Lifesciences Services GmbH, Kenta Biotech Kelm, Georg Wolff, Eberhard Barth, Ulrich Goebel, Eberhard Barth, Ltd, Maquet Critical Care AB, Omnicare Clinical Research AG, Nes- Anselm Kunstein, Michael Schuster, Martin Welte, Matthias Lutz, Pat- tle, Pierre Fabre Pharma AG, Pfizer, Bard Medica S.A., Abbott AG, rick Meybohm, Stephan Steiner, Tudor Poerner, Hendrik Haake, Stefan Anandic Medical Systems, Pan Gas AG Healthcare, Bracco, Hamil- Schaller, Detlef Kindgen-Milles, Christian Meyer, Muhammed Kurt, ton Medical AG, Fresenius Kabi, Getinge Group Maquet AG, Dräger Karl Friedrich Kuhn, Winfried Randerath, Jakob Wollborn, Zouhir AG, Teleflex Medical GmbH, Glaxo Smith Kline, Merck Sharp and Dindane, Hans-Joachim Kabitz, Ingo Voigt, Gonxhe Shala, Andreas Dohme AG, Eli Lilly and Company, Baxter, Astellas, Astra Zeneca, Faltlhauser, Nikoletta Rovina, Zoi Aidoni, Evangelia Chrisantho- CSL Behring, Novartis, Covidien, Phagenesis, and Nycomed outside poulou, Antonios Papadogoulas, Mohan Gurjar, Ata Mahmoodpoor, the submitted work. The money was paid into departmental funds. No Abdullah khudhur Ahmed, Brian Marsh, Ahmed Elsaka, Sigal Sviri, personal financial gain applied. The other authors declare that they Vittoria Comellini, Ahmed Rabha, Hazem Ahmed, Silvio a Namendys- have no conflicts of interest. Silva, Abdelilah Ghannam, Martijn Groenendijk, Marieke Zegers, Dylan de Lange, Alex Cornet, Mirjam Evers, Lenneke Haas, Tom Ethics approval and consent to participate The primary competent eth- Dormans, Willem Dieperink, Luis Romundstad, Britt Sjøbø, Finn H. ics committee was the Ethics Committee of the University of Duessel- Andersen, Hans Frank Strietzel, Theresa Olasveengen, Michael Hahn, dorf, Germany. Institutional research ethic board approval was obtained Miroslaw Czuczwar, Ryszard Gawda, Jakub Klimkiewicz, Maria de from each study site. Informed consent was obtained if not waived by Lurdes Campos Santos, André Gordinho, Henrique Santos, Rui Assis, the local ethical approval. This study was carried out in accordance Ana Isabel Pinho Oliveira, Mohamed Raafat Badawy, David Perez-Tor- with the Helsinki Declaration. res, Gemma Gomà, Mercedes Ibarz Villamayor, Angela Prado Mira, Patricia Jimeno Cubero, Susana Arias Rivera, Teresa Tomasa, David Consent for publication Not applicable. Iglesias, Eric Mayor Vázquez, Cesar Aldecoa, Aida Fernández Fer- reira, Begoña Zalba-Etayo, Isabel Canas-Perez, Luis Tamayo-Lomas, Cristina Diaz-Rodriguez, Susana Sancho, Jesús Priego, Enas M.Y. Open Access This article is licensed under a Creative Commons Attri- Abualqumboz, Momin Majed Yousuf Hilles, Mahmoud Saleh, Nawfel bution 4.0 International License, which permits use, sharing, adapta- Ben-HAmouda, Andrea Roberti, Alexander Dullenkopf, Yvan Fleury, tion, distribution and reproduction in any medium or format, as long Bernardo Bollen Pinto, Joerg C. Schefold, Mohammed Al-Sadawi, as you give appropriate credit to the original author(s) and the source, Nicolas Serck, Elisabeth Dewaele, Pritpal Kumar, Camilla Bundesen, provide a link to the Creative Commons licence, and indicate if changes Richard Innes, James Gooch, Lenka Cagova, Elizabeth Potter, Michael were made. The images or other third party material in this article are Reay, Miriam Davey, Sally Humphreys, Caroline Hauw Berlemont, included in the article's Creative Commons licence, unless indicated Benjamin Glenn Chousterman, François Dépret, Alexis Ferre, Lucie otherwise in a credit line to the material. If material is not included in Vettoretti, Didier Thevenin, Andreas Faltlhauser, Milena Milovanovic, the article's Creative Commons licence and your intended use is not Philipp Simon, Marco Lorenz, Sandra Emily Stoll, Simon Dubler, Kris- permitted by statutory regulation or exceeds the permitted use, you will tina Fuest, Francesk Mulita, Eumorifa Kondili, Ioannis Andrianop- need to obtain permission directly from the copyright holder. To view a oulos, Iwan Meynaar, Alexander Daniel Cornet, Britt Sjøbøe, Anna copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. Kluzik, Paweł Zatorski, Tomasz Drygalski, Wojciech Szczeklik, Joanna Solek-pastuszka, Dariusz Onichimowski, Jan Stefaniak, Karina Ste- fanska-Wronka, Ewa Zabul, Filipe Sousa Cardoso, Maria José Arche Banzo, Teresa Maria Tomasa-Irriguible, Ángela Prado Mira, Susana References Arias-Rivera, Fernando Frutos-Vivar, Sonia Lopez-Cuenca, Pablo Ruiz de Gopegui, Nour Abidi, Ivan Chau, Richard Pugh, Sara Smuts, Kle- 1. Wu Z, McGoogan JM. Characteristics of and important les- mens Zotter. sons from the coronavirus disease 2019 (COVID-19) out- break in China: summary of a report of 72 314 cases from the Author contributions All authors participated in the design and con- Chinese center for disease control and prevention. JAMA. ductance of the trial and were involved in patient recruitment and/or 2020;323:1239–42. study organisation. The primary statistical analysis was performed by 13 T. Mayerhöfer et al. 2. Ferrando-Vivas P, Doidge J, Thomas K, Gould DW, Mouncey P, of critically ill adults with COVID-19 in New York City: a pro- Shankar-Hari M, et al. Prognostic factors for 30-day mortality in spective cohort study. The Lancet. 2020;69:343. critically ill patients with coronavirus disease 2019: an observa- 20. Karagiannidis C, Mostert C, Hentschker C, Voshaar T, Malzahn tional cohort study. Crit Care Med. 2020;49:102–11. J, Schillinger G, et al. Case characteristics, resource use, and 3. Taylor EH, Marson EJ, Elhadi M, Macleod KDM, Yu YC, outcomes of 10 021 patients with COVID-19 admitted to 920 Davids R, et al. Factors associated with mortality in patients with German hospitals: an observational study. Lancet Respir Med. COVID-19 admitted to intensive care: a systematic review and 2020;8:853–62. meta-analysis. Anaesthesia. 2021;76:1224–32. 21. Caspersen CJ, Thomas GD, Boseman LA, Beckles GLA, Albright 4. Walsh EE, Frenck RW Jr, Falsey AR, Kitchin N, Absalon J, Gurt- AL. Aging, diabetes, and the public health system in the United man A, et al. Safety and immunogenicity of two RNA-based States. Am J Public Health. 2012;102:1482–97. Covid-19 vaccine candidates. N Engl J Med. 2020;383:2439–50. 22. Jung C, Flaatten H, Fjølner J, Bruno RR, Wernly B, Artigas A, 5. Voysey M, Clemens SAC, Madhi SA, Weckx LY, Folegatti PM, et al. The impact of frailty on survival in elderly intensive care Aley PK, et al. Safety and efficacy of the ChAdOx1 nCoV-19 patients with COVID-19: the COVIP study. Critical care (London, vaccine (AZD1222) against SARS-CoV-2: an interim analysis of England). 2021;25:149. four randomised controlled trials in Brazil, South Africa, and the 23. Richardson S, Hirsch JS, Narasimhan M, Crawford JM, McGinn UK. Lancet. 2021;397:99–111. T, Davidson KW, et al. Presenting characteristics, comorbidities, 6. Gottlieb RL, Vaca CE, Paredes R, Mera J, Webb BJ, Perez G, and outcomes among 5700 patients hospitalized with COVID-19 et al. Early remdesivir to prevent progression to severe Covid-19 in the New York City Area. JAMA. 2020;323:2052. in outpatients. N Engl J Med. 2021;386:305–15. 24. Bhatraju PK, Ghassemieh BJ, Nichols M, Kim R, Jerome KR, 7. Gupta A, Gonzalez-Rojas Y, Juarez E, Crespo Casal M, Moya Nalla AK, et al. Covid-19 in critically ill patients in the Seattle J, Falci DR, et al. Early Treatment for Covid-19 with SARS- region—case series. N Engl J Medicine. 2020;382:2012–22. CoV-2 Neutralizing Antibody Sotrovimab. N Engl J Med. 25. Karagiannidis C, Hentschker C, Westhoff M, Weber-Carstens S, 2021;385:1941–50. Janssens U, Kluge S, et al. Changes in utilization and outcomes 8. Weinreich DM, Sivapalasingam S, Norton T, Ali S, Gao H, Bhore of mechanical ventilation of COVID-19 during the course of the R, et al. REGEN-COV antibody combination and outcomes in pandemic in Germany in 2020: an observational study of 7,490 outpatients with Covid-19. N Engl J Med. 2021;385: e81. patients. Medrxiv. 2021;8:853 (2021.03.28.21254477). 9. Huang C, Wang Y, Li X, Ren L, Zhao J, Hu Y, et al. Clinical fea- 26. Armstrong RA, Kane AD, Cook TM. Outcomes from intensive tures of patients infected with 2019 novel coronavirus in Wuhan. care in patients with COVID-19: a systematic review and meta- China Lancet. 2020;395:497–506. analysis of observational studies. Anaesthesia. 2020;75:1340–9. 10. Williamson EJ, Walker AJ, Bhaskaran K, Bacon S, Bates C, Mor- 27. Dennis JM, Mateen BA, Sonabend R, Thomas NJ, Patel KA, Hat- ton CE, et al. Factors associated with COVID-19-related death tersley AT, et al. Type 2 diabetes and COVID-19–related mortality using OpenSAFELY. Nature. 2020;584:430–6. in the critical care setting: a National Cohort Study in England, 11. Petrilli CM, Jones SA, Yang J, Rajagopalan H, O’Donnell L, March–July 2020. Diabetes Care. 2020;44:50–7. Chernyak Y, et al. Factors associated with hospital admission 28. Grasselli G, Zangrillo A, Zanella A, Antonelli M, Cabrini L, and critical illness among 5279 people with coronavirus disease Castelli A, et al. Baseline Characteristics and Outcomes of 1591 2019 in New York City: prospective cohort study. BMJ. 2020;369: Patients Infected With SARS-CoV-2 Admitted to ICUs of the m1966. Lombardy Region, Italy. JAMA. 2020;323:1574. 12. Targher G, Mantovani A, Wang XB, Yan HD, Sun QF, Pan KH, 29. Mayerhöfer T, Klein SJ, Peer A, Perschinka F, Lehner GF, Has- et al. Patients with diabetes are at higher risk for severe illness slacher J, et al. Changes in characteristics and outcomes of criti- from COVID-19. Diabetes Metab. 2020;46:335–7. cally ill COVID-19 patients in Tyrol (Austria) over 1 year. Wien 13. Shang J, Wang Q, Zhang H, Wang X, Wan J, Yan Y, et al. The Klin Wochenschr. 2021;133:1–11. relationship between diabetes mellitus and COVID-19 progno- 30. Wang W, Xin C, Xiong Z, Yan X, Cai Y, Zhou K, et al. Clini- sis: a retrospective cohort study in Wuhan. China Am J Med. cal characteristics and outcomes of 421 patients with corona- 2021;134:e6–14. virus disease 2019 treated in a mobile cabin hospital. Chest. 14. Bode B, Garrett V, Messler J, McFarland R, Crowe J, Booth R, 2020;158:939–46. et al. Glycemic characteristics and clinical outcomes of COVID- 31. Woolcott OO, Castilla-Bancayán JP. The effect of age on the 19 patients hospitalized in the United States. J Diabetes Sci Tech- association between diabetes and mortality in adult patients with nol. 2020;14:813–21. COVID-19 in Mexico. Sci Rep. 2021;11:8386. 15. Kumar A, Arora A, Sharma P, Anikhindi SA, Bansal N, Sin- 32. Diedisheim M, Dancoisne E, Gautier J-F, Larger E, Cosson gla V, et al. Is diabetes mellitus associated with mortality and E, Fève B, et al. Diabetes increases severe COVID-19 out- severity of COVID-19? A meta-analysis. Diabetes Metab Syndr. comes primarily in younger adults. J Clin Endocrinol Metab. 2020;14:535–45. 2021;106:e3364–8. 16. Varikasuvu SR, Dutt N, Thangappazham B, Varshney S. Diabetes 33. Guidet B, Jung C, Flaatten H, Fjølner J, Artigas A, Pinto BB, and COVID-19: A pooled analysis related to disease severity and et al. Increased 30-day mortality in very old ICU patients with mortality. Prim Care Diabetes. 2021;15:24–7. COVID-19 compared to patients with respiratory failure without 17. Yang X, Yu Y, Xu J, Shu H, Ja X, Liu H, et al. Clinical course and COVID-19. Intensive Care Med. 2022;48:1–13. outcomes of critically ill patients with SARS-CoV-2 pneumonia 34. Guidet B, Jung C, Flaatten H, Fjølner J, Artigas A, Pinto BB, in Wuhan, China: a single-centered, retrospective, observational et al. Increased 30-day mortality in very old ICU patients with study. Lancet Respir Med. 2020;8(5):475–81. COVID-19 compared to patients with respiratory failure without 18. Schmidt M, Hajage D, Demoule A, Pham T, Combes A, Dres M, COVID-19. Intensive Care Med. 2022;48:435–47. et al. Clinical characteristics and day-90 outcomes of 4244 criti- 35. Mokdad AH, Ford ES, Bowman BA, Dietz WH, Vinicor F, Bales cally ill adults with COVID-19: a prospective cohort study. Intens VS, et al. Prevalence of obesity, diabetes, and obesity-related Care Med. 2021;47:60–73. health risk factors, 2001. JAMA. 2003;289(1):76–9. 19. Cummings MJ, Baldwin MR, Abrams D, Jacobson SD, Meyer BJ, 36. Frydrych LM, Bian G, O’Lone DE, Ward PA, Delano MJ. Obe- Balough EM, et al. Epidemiology, clinical course, and outcomes sity and type 2 diabetes mellitus drive immune dysfunction, 13 Diabetes mellitus is associated with 90‑day mortality in old critically ill COVID‑19 patients:… infection development, and sepsis mortality. J Leukoc Biol. 38. Livingstone SJ, Levin D, Looker HC, Lindsay RS, Wild SH, Joss 2018;104:525–34. N, et al. Estimated life expectancy in a Scottish cohort with type 37. Zhu L, She ZG, Cheng X, Qin JJ, Zhang XJ, Cai J, et al. Asso- 1 diabetes, 2008–2010. JAMA. 2015;313:37–44. ciation of blood glucose control and outcomes in patients with COVID-19 and pre-existing type 2 diabetes. Cell Metab. 2020;31:1068-77.e3. Authors and Affiliations Timo Mayerhöfer1 · Sebastian Klein1 · Bernhard Wernly2,3 · Hans Flaatten4 · Bertrand Guidet5 · Dylan W. De Lange6 · Jesper Fjølner7 · Susannah Leaver8 · Michael Beil9 · Sigal Sviri9 · Raphael Romano Bruno10 · Antonio Artigas11 · Peter Vernon van Heerden12 · Bernardo Bollen Pinto13 · Joerg C. Schefold14 · Rui Moreno15 · Maurizio Cecconi16,17 · Wojciech Szczeklik18 · Christian Jung10 · Michael Joannidis1 on behalf of COVIP study group 10 COVIP study group Department of Cardiology, Pulmonology and Vascular covip@med.uni-duesseldorf.de Medicine, Medical Faculty, Heinrich-Heine-University Duesseldorf, Moorenstraße 5, 40225 Duesseldorf, Germany 1 Division of Intensive Care and Emergency Medicine, 11 Intensive Intensive Care Medicine Department Corporacion Department of Internal Medicine, Medical University Sanitària Parc Tauli CIBER Enfermedades Respiratorias Innsbruck, Anichstrasse 35, 6020 Innsbruck, Austria Institut de Investigacio e Innovació I3PT, Autonomous 2 Department of Internal Medicine, General Hospital University of Barcelona Sabadell, Sabadell, Spain Oberndorf, Teaching Hospital of the Paracelsus Medical 12 Department of Anesthesia, Intensive Care and Pain Medicine, University Salzburg, Oberndorf, Salzburg, Austria Hadassah Medical Center and Faculty of Medicine, Hebrew 3 Center for Public Health and Healthcare Research, Paracelsus University of Jerusalem, Jerusalem, Israel Medical University of Salzburg, Salzburg, Austria 13 Department of Anaesthesiology, Pharmacology and Intensive 4 Department of Anaesthesia and Intensive Care, Haukeland Care, Geneva University Hospitals, Geneva, Switzerland University Hospital, Bergen, Norway 14 Department of Intensive Care Medicine, Inselspital, Bern 5 Assistance Publique, Hôpitaux de Paris, Sorbonne University Hospital, University of Bern, Bern, Switzerland Universités, UPMC Univ Paris 06, INSERM, UMR_S 1136, 15 Unidade de Cuidados Intensivos Neurocríticos e Trauma, Institut Pierre Louis d’Epidémiologie et de Santé Publique, Hospital de São José, Centro Hospitalar Universitário de Equipe: Epidémiologie Hospitalière Qualité et Organisation Lisboa Central, Faculdade de Ciências Médicas de Lisboa des Soins, 75012 Paris, France (Nova Médical School), Lisbon, Portugal 6 Department of Intensive Care Medicine, University Medical 16 Department of Anesthesia and Intensive Care Medicine, Center, University Utrecht, Utrecht, The Netherlands Humanitas Clinical and Research Center, IRCCS, Rozzano, 7 Department of Anaesthesia and Intensive Care, Viborg MI, Italy Regional Hospital, Viborg, Denmark 17 Department of Biomedical Sciences, Humanitas University, 8 Department of Critical Care, St George’s Hospital, London, Rozzano, MI, Italy UK 18 Center for Intensive Care and Perioperative Medicine, 9 Department of Medical Intensive Care, Hadassah Medical Jagiellonian University Medical College, Kraków, Poland Center and Faculty of Medicine, Hebrew University of Jerusalem, Jerusalem, Israel 13