Burden of disease in patients hospitalised with COVID-19 during the first and second pandemic wave in Switzerland: a nationwide cohort study

DOI: https://doi.org/https://doi.org/10.57187/smw.2023.40068

Introduction

In 2020, Switzerland faced two waves of the severe acute respiratory syndrome coronavirus type 2 (SARS-CoV-2) pandemic in spring and winter, respectively. At the beginning of the first wave, many European countries reached their capacity limits of acute and/or intensive care beds for COVID-19 affected patients requiring inpatient care [1]. Therefore, hospitals throughout Europe were forced to restore their capacity by postponing elective treatments and non-emergency surgeries [2, 3]. To address the threat, Swiss authorities introduced public health demands and restrictions to limit the spread of the new virus. While during the first wave, policy interventions included non-pharmacological mitigation measures such as national closures of borders, schools, non-essential stores and businesses with a nationwide lockdown mid of March 2020, [4] during the second wave, the main non-pharmacological measures included social distancing, increased testing, and restricted mobility [5, 6].

With cumulating evidence during the second wave, the antiviral remdesivir was prescribed more frequently [7–9]. In addition, dexamethasone was prescribed to most hospitalised patients needing oxygen therapy during the second wave [10]. Moreover, different strategies were applied regarding the hospitalisation of infected people. While, during the first wave, most infected people were hospitalised, even with only minor symptoms, this was no longer the case during the second wave, where criteria for hospitalisation were far more restrictive and based on clinical parameters [11]. Regardless, given the many infected people, the absolute number of hospitalisations during the second wave were higher.

As both the number of hospitalisations with COVID-19 and treatment options varied substantially between the first and second wave [12], we sought to evaluate epidemiological trends of COVID-19 related hospitalisations and corresponding in-hospital outcomes across different age groups using clinical routine data from Switzerland.

Methods

Study design

This analysis was conducted using a nationwide cohort of patients hospitalised with COVID-19 in Switzerland between January 1^st and December 31^st 2020. Hospitalisation data was provided by the Swiss Federal Statistical Office (FSO, Neuchâtel, Switzerland), based on a nationwide compulsory full census of Swiss hospitals. The dataset includes all Swiss inpatient discharge records from general hospitals for acute somatic care. Individual-level data on patient demographics, healthcare utilisation, hospital typology, medical diagnoses, clinical procedures, and in-hospital patient outcomes were provided. A multi-step anonymisation procedure ensured patient confidentiality, and an unique patient identifier was used to ascertain rehospitalisations. Medical diagnoses were coded using the International Classification of Disease version 10, German Modification (ICD-10-GM) codes. Open-source census data from the FSO on the Swiss population size, stratified by age and year, and data from the Swiss Federal Office of Public Health (FOPH) on the number of positive tests for SARS-CoV-2, stratified by age and period were obtained to calculate population-based incidence rates (IR) of hospitalisation with COVID-19. The institutional review board of Northwestern and Central Switzerland (EKNZ) waived the need for an ethical authorisation due to the use of exclusively anonymised data (EKNZ Project-ID: Req-2021-01397). This study adheres to the “Strengthening The Reporting of Observational Studies in Epidemiology (STROBE)” statement [13].

Case ascertainment and study variables

For this analysis, we included all hospitalisations that were treated with or for COVID-19. Hospitalisations in special clinics (psychiatric clinics, rehabilitation clinics, and other special clinics) were excluded. To identify hospitalisations with COVID-19, we used the following International Statistical Classification of Diseases and Related Health Problems, German Modification (ICD-10-GM) discharge codes at any position: U07.1 and U07.2. Details on all ICD-10-GM codes and Swiss operation classification (CHOP) codes used for the analysis are summarised in Table S1-S3 in the appendix. Comorbidities were measured using the Elixhauser Comorbidity Index [14], and frailty was measured using the Hospital Frailty Score [15].

Outcomes

The primary outcome was the IR of hospitalisation with COVID-19 per 10,000 person-years (PY) with corresponding 95% confidence intervals (CIs) during the first and second waves across the age spectrum. Secondary outcomes comprised of the occurrence of the following in-hospital outcomes during the first and second wave: all-cause in-hospital mortality, acute respiratory distress syndrome (ARDS), length of hospital stay (LOS), intensive care unit (ICU) admission, length of ICU stay, mechanical ventilation, duration of mechanical ventilation, extracorporeal membrane oxygenation (ECMO), and 30-day all-cause hospital readmission. Hospitalisation with COVID-19 and ARDS was defined using ICD-10-GM codes, and the need for ECMO was defined using CHOP codes (table S1-S3 in the appendix). The remaining outcomes were applicable in the dataset provided by the FSO. Analyses of hospital outcomes were stratified by different age groups (0–9, 10–19, 20–49, 50–69, and ≥70 years).Out-of-hospital mortality data was not available in the dataset and could not be linked to the national death registry.

Statistical analysis

Descriptive statistic was calculated for patient demographics, including age, sex, nationality, and insurance status. All baseline characteristics are expressed as mean (standard deviation [SD]), median (interquartile range [IQR]), or frequency (%). Graphical illustration of hospitalisation IR over age spectrum was performed using locally estimated scatterplot smoothing (LOESS). Stratified by waves (first wave: January to May, second wave: June to December), we estimated IR per 10,000 PY with 95% CI, rate differences (RD), and incidence rate ratios (IRR). IR was calculated, allowing multiple hospitalisations for a single person (more than 95% of the patients were hospitalised only once) [16]. These estimates were calculated as the number of individuals with a COVID-19 hospitalisation divided by the sum of “person-time” population at risk in Switzerland, represented by the population size multiplied by the 5-month follow-up during the first wave and by the 7-month follow-up, according to age. To assess the IR of hospitalisation among the overall Swiss population, the denominator was the standard population in 2020 during the first or second waves, respectively. We used the number of people living in Switzerland at the end of 2020 as a surrogate. This information is publicly available and published by the FSO [17]. We aggregated COVID-19 hospitalisations per year of patient age. In detail, we counted the number of COVID-19 hospitalisations for each patient age stratum paired with the published number of persons during the respective wave in Switzerland. Thus, we assumed 5/12 of one person-year for every resident during the first wave and 7/12 of one person-year for every resident during the second wave, ignoring people dying or moving in and out of the country. IR of hospitalisation among those who had a proven SARS-CoV-2 infection were calculated using the overall positively tested population in the denominator. These analyses were performed within the above-mentioned age groups, and the first wave was denoted as a reference.

To explore differences in binary hospital outcomes between the waves and by age groups, we estimated adjusted odds ratios (aOR) and corresponding 95% CIs using a multivariable logistic regression model. For continuous right-skewed outcomes, we assessed changes in percentage and corresponding 95% CIs using a multiple linear generalised log-gamma regression model. All models were adjusted for age, sex and Elixhauser comorbidity index. The selection of covariates was based on the research question at hand and on knowledge such as what was important to guide hospitalisation criteria during the waves in Switzerland.

We evaluated for heterogeneity in OR estimates across age groups using the Wald test for homogeneity. We performed a risk factor analysis for mortality using univariable and multivariable analyses. All p-values are two-sided and have not been adjusted for multiple testing. Results were considered statistically significant at p <0.05. Statistical analyses were performed with STATA 15.1 (STATA Corp., College Station, TX, USA).

Results

Characteristics of the cohort

From January 1^st to December 31^st 2020, we identified 36,517 hospitalisations with COVID-19, 8862 (24.3%) during the first and 27,655 (75.7%) during the second wave. Baseline characteristics of the eligible study population stratified by age groups are shown in table 1. Baseline characteristics stratified by waves are summarised in table S4 in the appendix. Overall, the median age was 72 years (IQR, 58 to 82), 57.3% were male, 74.8% were Swiss residents, and 13.8% had supplementary health care insurance. The majority of hospitalisations was observed in patients aged older than 45 years. Among those, we observed an overall high burden of comorbidities, with similar distribution during the first and second waves.

Hospitalisation rates of patients with COVID-19

While IR of hospitalisations during the first and second waves were generally low in children and adolescents, they increased at around the age of 40 years, with a peak in older patients during both waves (figure 1a).

Figure 1a Age-dependent incidence rates for hospitalisations for COVID-19 among the overall Swiss population per 10,000 person-years during the first wave (dark blue) and second wave (light blue), using locally estimated scatterplot smoothing (LOESS).

Figure 1b Age-dependent incidence rates for hospitalisations for COVID-19 among the overall SARS-CoV-2 positive tested population per 10,000 person-years during the first wave (dark blue) and second wave (light blue), using locally estimated scatterplot smoothing (LOESS).

Considering the overall standard population in the denominator, IR were higher during the second wave compared with the first wave. COVID-19 hospitalisation rates were highest in patients above 50 years (50–69 years: first wave: 31.49 per 10,000 person-years; second wave: 62.81 per 10,000 person-years; RD 31.32 [95% CI: 29.56 to 33.08] per 10,000 person-years; ≥70 years: first wave: 88.59 per 10,000 person-years; second wave: 228.41 per 10,000 person-years; RD 139.83 [95% CI: 135.42 to 144.23] per 10,000 person-years) with a strong predominance during the second wave (table 2).

Illustration of hospitalisation IR across the age spectrum among those who were tested positive for SARS-CoV-2 showed different patterns for the first and second wave, with a first peak of hospitalisations among children aged 0 to 9 year and a second peak with a continuous increase of hospitalisations among adults beyond age 50 (figure 1b). While rates of hospitalisation among the overall Swiss population were higher during the second wave (figure 1a), the proportion of hospitalised people among positively tested individuals was higher during the first wave (figure 1b).

In-hospital outcomes between the first and second wave

Among the overall population, 1028 (11.6%) died in hospital during the first wave and 3264 (11.8%) during the second wave, corresponding to lower odds of in-hospital mortality during the second wave (aOR 0.88 [95% CI: 0.81 to 0.95]). Similarly, compared with the first wave, we observed lower numbers of patients with ARDS (1056 [11.9%] vs. 1945 [7.0%], aOR 0.56 [95% CI: 0.51 to 0.61]), ICU admission (1463 [16.5%] vs. 3160 [11.4%], aOR 0.66 [95% CI: 0.61 to 0.70]), in need for mechanical ventilation (1101 [12.4%] vs. 2120 [7.7%], aOR 0.67 [95% CI: 0.58 to 0.78]), and in need for ECMO (38 [0.4%] vs. 46 [0.2%], aOR 0.60 [95% CI: 0.38 to 0.92]) during the second wave. The odds for 30-day all-cause hospital readmission remained similar during the first and second wave (427 [4.8%] vs. 1340 [4.9%], aOR of 0.99 [95% CI: 0.88 to 1.10]) (figure 2). Compared with the first wave, LOS (7.4 [SD 2.7] vs. 6.7 [SD 2.5]) and ICU LOS (6.2 [SD 4.0] vs. 4.1 [SD 3.9]) were shorter during the second wave with a reduction of –16.1% (95% CI: –17.8 to –14.2) and –32.7% (95% CI: –37.2 to –27.9), respectively (figure 3). There was no evidence for effect modification by age for all hospital outcomes of interest (p of interaction >0.05).

Figure 2 Odds ratios and 95% confidence intervals for adverse outcomes in prespecified age subgroups. First wave was chosen as a reference. There was no evidence for effect modification by age for all hospital outcomes of interest (p of interaction >0.05).

* adjusted for age, sex and Elixhauser comorbidity Index

ARDS: acute respiratory distress syndrome; CI: confidence interval; ECMO: extracorporeal membrane oxygenation; ICU: intensive care unit; N/A: not applicable; no: number; OR: odds ratio

Figure 3 Changes in frequency for adverse outcomes in prespecified age subgroups.

First wave was chosen as a reference. There was no evidence for effect modification by age for all hospital outcomes of interest (p of interaction >0.05).

* adjusted for age, sex and Elixhauser comorbidity Index

CI: confidence interval; ICU: intensive care unit; SD: standard deviation

Predictors for mortality

Table 3 and 4 show baseline risk factors for in-hospital mortality during the first and second waves. Among the adjusted regression analysis, the main risk factors were the prevalence of chronic kidney disease stage 5 and/or hemodialysis (first wave: aOR 3.26 [95% CI: 2.47 to 4.31), p <0.001]; second wave: aOR 4.02 [95% CI: 3.31 to 4.89], p <0.001) and haematological malignancy (first wave: aOR 3.03 [95% CI: 1.97 to 4.66)], p <0.001; second wave: aOR 2.39 [95% CI: 1.92 to 2.98], p <0.001).

Discussion

This nationwide cohort study of hospitalised COVID-19 patients in Switzerland revealed several key findings: First, hospitalisation rates in patients with COVID-19 were highest among adults older than 50 years and more pronounced during the second wave. Second, while the hospitalisation rate was higher among the overall Swiss population during the second wave, it was lower among positively tested individuals during the same period. Third, except for readmission, the risk for hospital-associated adverse outcomes was lower during the second wave, irrespective of patient age.

The highest hospitalisation rates among the overall Swiss population were observed in patients aged 50 years and older, both during the first and second wave. However, the peak of hospitalisations during the second wave was much higher than the first wave. These results are plausible since we observed significantly more infections during the second wave, resulting in more hospitalisations. Although data must be interpreted carefully, as hospitalisation and test criteria differed between the two waves, we observed a lower proportion of hospitalised people among the overall positively tested population during the second wave.

Improvement of patient management was a further important key component which was achieved during the second wave. Overall, the odds of adverse outcomes decreased during the second wave, except for hospital readmission. The expected decrease of in-hospital adverse events during the second wave is probably explained by the improved understanding of the treating physicians and nursing staff about the management of COVID-19 patients [18]. Importantly, the widespread administration of dexamethasone during the second wave may have contributed to a stronger reduction in in-hospital mortality, as already shown by previous studies [19]. Consistent with other studies [20–24], we observed a lower risk of in-hospital mortality during the second wave (aOR 0.88 [95% CI: 0.81 to 0.95]). While, as compared with findings from clinical trials, a relative risk reduction of 12% may seem small, hospitalised patients during the second wave tended to be older and more comorbid, both characteristics known to independently increase the risk of COVID-19-related mortality [25–28].

Similarly, corticosteroids may also have reduced the rate of progression to severe COVID-19 due to an attenuation of the inflammation, resulting in lower rates of ARDS, admission to ICU, need for mechanical ventilation and ECMO support as well as a 16% shorter LOS of 7.4 vs. 6.7 days [29]. In addition, the reduction in LOS during the second wave could also be due to improved discharge processes and earlier transfer to rehabilitation facilities. This was possible since many rehabilitation facilities were obligated to unburden acute care hospitals by accepting still-infectious COVID-19 patients [30].

We did not observe a change in readmission rates between the two waves. This can be explained by the fact that COVID-19 is an acute disease and may not lead to readmissions per se, whereas the burden of comorbidities as a potential driver for readmission tended to be higher during the second wave and may have diluted any beneficial effect on hospital readmission. These data must be interpreted in the context of the study design. As COVID-19 hospitalisations were identified using ICD-10-GM codes used for billing purposes, thus, misclassification and underreporting are possible, especially during the beginning of the pandemic when no specific ICD-10 codes were available. In line, the database provided by the FSO revealed larger numbers of hospitalisations as compared with data from the CH-SUR database. Since algorithms based on the U07.1 code had high sensitivity among hospitalised patients but at the expense of low specificity [31], it is likely that the number of hospitalisation with COVID-19 as provided by the FSO might be overestimated. However, it can also not be excluded that some hospitalisations with COVID-19 in the FSO-database were not lab-confirmed during the hospital stay but diagnosed based on an out-of-hospital test or clinical features. Second, unmeasurable confounding like smoking status, genetic susceptibility or home medication must be considered, as the used dataset does not include any information in this regard. Third, due to the study’s retrospective design, no causal inference is possible. Fourth, we did not account for potential within-patient/hospital clustering. However, we do not think that accounting for clustering relevantly changes the conclusion of our manuscript, as the number of clusters (hospitals, hospital admissions per patient) were comparable between the first and second waves of the pandemic. Fifth, medical treatments received during the hospitalisation for COVID-19 could not be analysed, as this information is missing in the analysed dataset. However, there are several strengths of note. This analysis was based on nationwide hospital care data with high external validity, strong statistical power and high generalisability across all regions in Switzerland. Moreover, this study provides insights into different age groups that were not comprehensively addressed in earlier studies in Switzerland.

Conclusion

In this nationwide cohort study, hospitalisation rates of COVID-19 patients were highest among adults older than 50 years and during the second wave. Except for readmission, the risk of hospital adverse outcomes was decreased during the second wave, regardless of the patient`s age.

Data sharing statement

The data supporting this study’s findings are available upon request from the Swiss Federal Statistical Office (Neuchâtel, Switzerland). Restrictions apply to the availability of these data, which were used under license for this study. Data are available as part of the data on “Medizinische Statistik der Krankenhäuser” with the permission of the Swiss Federal Statistical Office, Section Health Services and Population Health.

Acknowledgement

We thank the Swiss Federal Statistics Office (Neuchâtel, Switzerland) for the acquisition and provision of data.

References

1. Verelst F , Kuylen E , Beutels P . Indications for healthcare surge capacity in European countries facing an exponential increase in coronavirus disease (COVID-19) cases, March 2020. Euro Surveill. 2020 Apr;25(13):2000323. https://doi.org/10.2807/1560-7917.ES.2020.25.13.2000323  

2. Winkelmann J , Webb E , Williams GA , Hernández-Quevedo C , Maier CB , Panteli D . European countries’ responses in ensuring sufficient physical infrastructure and workforce capacity during the first COVID-19 wave. Health Policy. 2022 May;126(5):362–72. https://doi.org/10.1016/j.healthpol.2021.06.015  

3. Webb E , Hernández-Quevedo C , Williams G , Scarpetti G , Reed S , Panteli D . Providing health services effectively during the first wave of COVID-19: A cross-country comparison on planning services, managing cases, and maintaining essential services. Health Policy. 2022 May;126(5):382–90. https://doi.org/10.1016/j.healthpol.2021.04.016  

4. Zimmermann BM , Fiske A , McLennan S , Sierawska A , Hangel N , Buyx A . Motivations and Limits for COVID-19 Policy Compliance in Germany and Switzerland. Int J Health Policy Manag. 2021 Apr;11(8):1342–53. https://doi.org/10.34172/ijhpm.2021.30  

5. Salathé M , Althaus CL , Neher R , Stringhini S , Hodcroft E , Fellay J , et al.  COVID-19 epidemic in Switzerland: on the importance of testing, contact tracing and isolation. Swiss Med Wkly. 2020 Mar;150:w20225. https://doi.org/10.4414/smw.2020.20225  

6. Flaxman S , Mishra S , Gandy A , Unwin HJ , Mellan TA , Coupland H , et al.; Imperial College COVID-19 Response Team . Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe. Nature. 2020 Aug;584(7820):257–61. https://doi.org/10.1038/s41586-020-2405-7  

7. Cao B , Wang Y , Wen D , Liu W , Wang J , Fan G , et al.  A Trial of Lopinavir-Ritonavir in Adults Hospitalized with Severe Covid-19. N Engl J Med. 2020 May;382(19):1787–99. https://doi.org/10.1056/NEJMoa2001282  

8. Boulware DR , Pullen MF , Bangdiwala AS , Pastick KA , Lofgren SM , Okafor EC , et al.  A Randomized Trial of Hydroxychloroquine as Postexposure Prophylaxis for Covid-19. N Engl J Med. 2020 Aug;383(6):517–25. https://doi.org/10.1056/NEJMoa2016638  

9. Wang Y , Zhang D , Du G , Du R , Zhao J , Jin Y , et al.  Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial. Lancet. 2020 May;395(10236):1569–78. https://doi.org/10.1016/S0140-6736(20)31022-9  

10. Horby P , Lim WS , Emberson JR , Mafham M , Bell JL , Linsell L , et al.; RECOVERY Collaborative Group . Dexamethasone in Hospitalized Patients with Covid-19. N Engl J Med. 2021 Feb;384(8):693–704. https://doi.org/10.1056/NEJMoa2021436  

11. Biswas M , Rahaman S , Biswas TK , Haque Z , Ibrahim B . Association of Sex, Age, and Comorbidities with Mortality in COVID-19 Patients: A Systematic Review and Meta-Analysis. Intervirology. 2020 Dec;1–12.  

12.   Bundesamt für Statistik . Auswirkungen der Covid-19-Pandemie auf die Gesundheitsversorgung im Jahr 2020. 2021. 

13. von Elm E , Altman DG , Egger M , Pocock SJ , Gøtzsche PC , Vandenbroucke JP ; STROBE Initiative . Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ. 2007 Oct;335(7624):806–8. https://doi.org/10.1136/bmj.39335.541782.AD  

14. Elixhauser A , Steiner C , Harris DR , Coffey RM . Comorbidity measures for use with administrative data. Med Care. 1998 Jan;36(1):8–27. https://doi.org/10.1097/00005650-199801000-00004  

15. Gilbert T , Neuburger J , Kraindler J , Keeble E , Smith P , Ariti C , et al.  Development and validation of a Hospital Frailty Risk Score focusing on older people in acute care settings using electronic hospital records: an observational study. Lancet. 2018 May;391(10132):1775–82. https://doi.org/10.1016/S0140-6736(18)30668-8  

16. Bundesamt für Gesundheit (BAG) . Coronavirus: Monitoring [Available from: https://www.bag.admin.ch/bag/de/home/krankheiten/ausbrueche-epidemien-pandemien/aktuelle-ausbrueche-epidemien/novel-cov/situation-schweiz-und-international/monitoring.html#19385953 ], last access: 23.12.2022. 

17. Bundesamt für Statistik . Permanent resident population by age, sex and category of citizenship, 2010-2021 2010-2021 [Available from: https://www.bfs.admin.ch/bfs/de/home/statistiken/bevoelkerung/stand-entwicklung/alter-zivilstand-staatsangehoerigkeit.assetdetail.23064709.html ], last access: 23.12.2022. 

18.   World Health Organization (WHO) . Clinical management of COVID-19: Living guideline, 23 June 2022 2022 [Available from: WHO-2019-nCoV-Clinical-2022.1-eng.pdf.], last access, 23.12.2022. 

19. van Paassen J , Vos JS , Hoekstra EM , Neumann KM , Boot PC , Arbous SM . Corticosteroid use in COVID-19 patients: a systematic review and meta-analysis on clinical outcomes. Crit Care. 2020 Dec;24(1):696. https://doi.org/10.1186/s13054-020-03400-9  

20. Oladunjoye O , Gallagher M , Wasser T , Oladunjoye A , Paladugu S , Donato A . Mortality due to COVID-19 infection: A comparison of first and second waves. J Community Hosp Intern Med Perspect. 2021 Nov;11(6):747–52. https://doi.org/10.1080/20009666.2021.1978154  

21. Saito S , Asai Y , Matsunaga N , Hayakawa K , Terada M , Ohtsu H , et al.  First and second COVID-19 waves in Japan: A comparison of disease severity and characteristics. J Infect. 2021 Apr;82(4):84–123. https://doi.org/10.1016/j.jinf.2020.10.033  

22. Di Fusco M , Shea KM , Lin J , Nguyen JL , Angulo FJ , Benigno M , et al.  Health outcomes and economic burden of hospitalized COVID-19 patients in the United States. J Med Econ. 2021;24(1):308–17. https://doi.org/10.1080/13696998.2021.1886109  

23. Mallow PJ , Belk KW , Topmiller M , Hooker EA . Outcomes of Hospitalized COVID-19 Patients by Risk Factors: Results from a United States Hospital Claims Database. J Health Econ Outcomes Res. 2020 Sep;7(2):165–74. https://doi.org/10.36469/jheor.2020.17331  

24. Iftimie S , López-Azcona AF , Vallverdú I , Hernández-Flix S , de Febrer G , Parra S , et al.  First and second waves of coronavirus disease-19: A comparative study in hospitalized patients in Reus, Spain. PLoS One. 2021 Mar;16(3):e0248029. https://doi.org/10.1371/journal.pone.0248029  

25. Hothorn T , Bopp M , Günthard H , Keiser O , Roelens M , Weibull CE , et al.  Assessing relative COVID-19 mortality: a Swiss population-based study. BMJ Open. 2021 Mar;11(3):e042387. https://doi.org/10.1136/bmjopen-2020-042387  

26. Williamson EJ , Walker AJ , Bhaskaran K , Bacon S , Bates C , Morton CE , et al.  Factors associated with COVID-19-related death using OpenSAFELY. Nature. 2020 Aug;584(7821):430–6. https://doi.org/10.1038/s41586-020-2521-4  

27. Bravi F , Flacco ME , Carradori T , Volta CA , Cosenza G , De Togni A , et al.  Predictors of severe or lethal COVID-19, including Angiotensin Converting Enzyme inhibitors and Angiotensin II Receptor Blockers, in a sample of infected Italian citizens. PLoS One. 2020 Jun;15(6):e0235248. https://doi.org/10.1371/journal.pone.0235248  

28. Maximiano Sousa F , Roelens M , Fricker B , Thiabaud A , Iten A , Cusini A , et al.; Ch-Sur Study Group . Risk factors for severe outcomes for COVID-19 patients hospitalised in Switzerland during the first pandemic wave, February to August 2020: prospective observational cohort study. Swiss Med Wkly. 2021 Jul;151(2930):w20547. https://doi.org/10.4414/smw.2021.20547  

29. Ye Z , Wang Y , Colunga-Lozano LE , Prasad M , Tangamornsuksan W , Rochwerg B , et al.  Efficacy and safety of corticosteroids in COVID-19 based on evidence for COVID-19, other coronavirus infections, influenza, community-acquired pneumonia and acute respiratory distress syndrome: a systematic review and meta-analysis. CMAJ. 2020 Jul;192(27):E756–67. https://doi.org/10.1503/cmaj.200645  

30. Liebl ME , Gutenbrunner C , Glaesener JJ , Schwarzkopf S , Best N , Lichti G , et al.  Early Rehabilitation in COVID-19 – Best Practice Recommendations for the Early Rehabilitation of COVID-19 Patients. Phys Med Rehabilmed Kurortmed. 2020 Jun;30(3):129–34. https://doi.org/10.1055/a-1162-4919  

31. Brown CA , Londhe AA , He F , Cheng A , Ma J , Zhang J , et al.  Development and Validation of Algorithms to Identify COVID-19 Patients Using a US Electronic Health Records Database: A Retrospective Cohort Study. Clin Epidemiol. 2022 May;14:699–709. https://doi.org/10.2147/CLEP.S355086   

Appendix