Characteristics and outcomes of COVID-19 patients admitted to hospital with and without respiratory symptoms

Background COVID-19 is primarily known as a respiratory illness; however, many patients present to hospital without respiratory symptoms. The association between non-respiratory presentations of COVID-19 and outcomes remains unclear. We investigated risk factors and clinical outcomes in patients with no respiratory symptoms (NRS) and respiratory symptoms (RS) at hospital admission. Methods This study describes clinical features, physiological parameters, and outcomes of hospitalised COVID-19 patients, stratified by the presence or absence of respiratory symptoms at hospital admission. RS patients had one or more of: cough, shortness of breath, sore throat, runny nose or wheezing; while NRS patients did not. Results Of 178,640 patients in the study, 86.4 % presented with RS, while 13.6 % had NRS. NRS patients were older (median age: NRS: 74 vs RS: 65) and less likely to be admitted to the ICU (NRS: 36.7 % vs RS: 37.5 %). NRS patients had a higher crude in-hospital case-fatality ratio (NRS 41.1 % vs. RS 32.0 %), but a lower risk of death after adjusting for confounders (HR 0.88 [0.83–0.93]). Conclusion Approximately one in seven COVID-19 patients presented at hospital admission without respiratory symptoms. These patients were older, had lower ICU admission rates, and had a lower risk of in-hospital mortality after adjusting for confounders.

hospital admission.RS patients had one or more of: cough, shortness of breath, sore throat, runny nose or wheezing; while NRS patients did not.Results: Of 178,640 patients in the study, 86.4 % presented with RS, while 13.6 % had NRS.NRS patients were older (median age: NRS: 74 vs RS: 65) and less likely to be admitted to the ICU (NRS: 36.7 % vs RS: 37.5 %).NRS patients had a higher crude in-hospital case-fatality ratio (NRS 41.1 % vs. RS 32.0 %), but a lower risk of death after adjusting for confounders (HR 0.88 [0.83-0.93]).Conclusion: Approximately one in seven COVID-19 patients presented at hospital admission without respiratory symptoms.These patients were older, had lower ICU admission rates, and had a lower risk of in-hospital mortality after adjusting for confounders.
One of the first public health measures to contain transmission of SARS-CoV-2 was identifying febrile patients with respiratory symptoms (RS) and isolating them until laboratory diagnosis was confirmed [11].However, a proportion of patients with COVID-19 present with no respiratory symptoms (NRS) [12].A large proportion of COVID-19 patients require in-hospital treatment and have at least one extrapulmonary manifestation during their acute infection [13][14][15][16].However, the clinical outcomes and factors associated with non-respiratory presentations have not been explored systematically [14].
This study attempted to bridge this knowledge gap by characterising the risk factors and clinical outcomes of patients admitted to the hospital with NRS and RS using the ISARIC-WHO database.We hypothesise that the presumed multisystem involvement in patients with NRS is associated with poor prognosis.This information can be relevant to optimise case management and provide helpful information to clinicians treating patients with COVID-19.

Methods
We used the International Severe Acute Respiratory and Emerging Infection Consortium (ISARIC) -World Health Organization (WHO) Clinical Characterisation Protocol (CCP) for Severe Emerging Infections prospective observational data collection platform for hospitalised patients [17].Participating sites collected the data prospectively using the ISARIC case report forms (CRFs) built on Research Electronic Data Capture (REDCap, version 8.11.11;Vanderbilt University, Nashville, TN, USA), hosted by the University of Oxford (Oxford, UK).Data were also collected on local databases in other settings and submitted for harmonisation and storage at the University of Oxford.Data were converted to Study Data Tabulation Model standards (version 1.7; Clinical Data Interchange Standards Consortium, Austin, TX, USA) to integrate data collected on locally hosted databases with data collected on the ISARIC database.All investigators retain full rights to their data.The protocol, CRFs, and study information are available on the ISARIC website (https:// isaric.org/).
The ISARIC-WHO CCP was approved by the WHO ethics review committee (RPC571 and RPC572).Local ethics approval was obtained for each participating country and site according to local requirements.

Study population
We included patients admitted to the hospital between 30 th January 2020 and 30 th December 2022 with clinically diagnosed (i.e., symptoms and findings of SARS-CoV-2 pneumonia seen in thoracic diagnostic images) or laboratory-confirmed (i.e., positive reverse transcription polymerase chain reaction) SARS-CoV2 infection according to American Thoracic Society and Infectious Disease Society of America (ATS/IDSA) COVID-19 guidelines [18].Patients with data on the type of oxygen supplementation status received at any time during their hospitalisation and data on the presence or absence of respiratory symptoms during the first 24 h of admission were included in the study.We excluded patients with missing age or sex, those with missing or unknown respiratory symptoms, and those with missing or negative SARS-CoV-2 status.Sex was defined as the sex assigned at birth and was categorised into male or female.

Variables and measurement
The following variables were included in the analysis: age, sex, comorbidities, complications, country of recruitment and its region according to the World Bank criteria (https://data.worldbank.org/country),vital signs during the first 24 h of admission, treatments, and clinical outcome, that is, in-patient death, and loss to follow up.The key outcome of interest was in-hospital mortality.Patients presenting with one or more symptoms of cough, shortness of breath, sore throat, runny nose or wheezing at the time of hospital admission, irrespective of other symptoms, were classified in the RS group.Regardless of other symptoms, patients not presenting with these respiratory symptoms were classified in the NRS group.Patients who were lost follow-up (i.e., transferred to another hospital or receiving ongoing care) were not considered for fatal outcomes analyses.

Statistical methods
We used descriptive statistics to summarise patient demographics and baseline characteristics.For continuous variables, characteristics were reported as medians and interquartile ranges (IQRs).For categorical variables, counts and percentages were reported.Patient characteristics were compared between the NRS and RS patient groups.
The administration of oxygen therapy at any time during hospitalisation by oxygen delivery methodsbasic oxygen therapy, a high-flow nasal cannula (HFNC), non-invasive ventilation (NIV), invasive mechanical ventilation (IMV), and extracorporeal membrane oxygenation (ECMO) -was compared between the NRS and RS patient groups.The overall baseline median (IQR) oxygen saturation (SpO 2 ) levels, stratified by age groups, were also compared between the two groups.
We used the Cox proportional hazards model after testing for proportional hazards in the survival analysis to assess the associations of non-respiratory symptoms with the hazard of death.We assessed the proportional hazards assumption using scaled Schoenfeld residuals.Hazard ratios (HRs) and 95 % CIs were estimated for the entire hospitalisation duration and restricted to a shorter hospitalisation duration of 7 and 14 days.Models were adjusted for age (in ten-year age bands), sex, all comorbidities and risk factors, and stratified by country.We grouped countries with less than 50 individuals into a single category.
Comorbidities and risk factors included HIV/AIDS, asthma, cardiac disease, chronic kidney disease, chronic neurological disorder, chronic pulmonary disease, dementia, diabetes, hypertension, liver disease, malignant neoplasm, malnutrition, obesity, smoking, transplantation, rheumatologic disorder and immunosuppression.Immunosuppression was defined according to specific criteria outlined in the case record form for patients who had (i) Pre-admission medication including immunosuppressants such as oral corticosteroids (excluding low-dose hydrocortisone); (ii) People identified as part of clinically extremely vulnerable groups; (iii) People who underwent bone marrow or stem cell transplants within the previous 6 months or were currently under immunosuppression medication; and (iv) People receiving immunosuppressive therapies sufficient to significantly increase risk of infection.
Patients were censored if they were lost to follow-up, which in our dataset could mean they were transferred to another facility or were receiving ongoing care at the time of most recent data collection.Time from symptom onset to time of death or censoring (time to last known to be alive), whichever occurred earlier, was used as the timescale.Patients were considered at risk from symptom onset or admission, whichever occurred later.For all outcomes, censoring times of discharged patients were modified and set to be equal to the maximum time to censoring/event (to account for informative censoring).All statistical analyses were performed using the R statistical programming language, version 4.0.2, and packages survival, ggplot2, and finalfit.

Results
We included a total of 178,640 patients (Fig. 1) from 66 countries.

Clinical characteristics of patients with NRS and RS
Compared to RS patients, NRS patients were older, with a median (IQR) age of 74 (60-84) vs 65 (53-77) for RS patients.There were more male than female patients in both NRS and RS groups, and more male patients in the RS than the NRS group (NRS: 55.9 % [13,559/24,248] and RS: 60.6 % [93,585/154,392]) (Table 1).
The frequency of some comorbidities and risk factors varied between patients with or without respiratory symptoms: hypertension (NRS: 50.0 % [10, 1).
In the Cox proportional hazards survival analysis, adjusted for age, sex, country, all comorbidities and risk factors, patients with NRS had a lower in-patient mortality risk than patients with RS during their entire hospitalisation (HR [95 % CI] 0.88 (0.83-0.93, p < 0.001) (Table 2; Fig. 3).The in-patient mortality risk remained similar after performing a sensitivity analysis restricted to a shorter hospitalisation duration of 7 and 14 days; however, this was not statistically significant after when restricted to 7 days (   2).

A.3).
We compared the administration of oxygen therapy at any time during hospitalisation by oxygen delivery methods (Table 4).During hospitalisation, basic oxygen therapy was the most frequent form of oxygen therapy used in NRS patients (52.4).

Discussion
In this large multicentre and prospective cohort, we found that around one in seven hospitalised patients diagnosed with SARS-CoV-2 had no respiratory symptoms of cough, shortness of breath, sore throat, runny nose or wheezing at hospital admission.Compared to those who presented with RS, patients with NRS were older and more likely to suffer from comorbidities other than asthma and chronic pulmonary disease.During hospitalisation, those with NRS were less likely to receive treatment with vasopressors, corticosteroids, and admission to the ICU; however, they developed respiratory failure comparable to those with RS.Notably, the risk for in-hospital mortality was lower in patients with NRS after adjusting for confounders.
COVID-19 has a broad clinical spectrum [10], though its principal manifestation is respiratory [19,20].Hence, respiratory symptoms have been a critical criterion for identifying SARS-CoV-2 infection [21].Thus, patients with lung comorbidities have been prioritised during vaccination campaigns for patient care since they are at a higher risk of developing more severe respiratory symptoms [22][23][24].This can be attributable to the already dysregulated pulmonary physiology [25,26].In contrast, at least in the initial phases of COVID-19, patients without apparent respiratory symptoms may be overlooked [8,9,27].Observational studies have found that almost 30 % of patients manifest atypical symptoms, increasing the risk of misdiagnosis and leading to delays in healthcare, the development of multiorgan failure, and worse clinical outcomes [28][29][30][31].Our results show that most patients with NRS admitted to the hospital required supplementary oxygen at some point during their hospital stay, and almost a third were admitted to ICU, which aligns with prior data [28][29][30][31].
One of the main results of our study is that patients with NRS had higher crude in-hospital mortality risk but lower risk than RS patients after adjusting for confounders.Some small prior studies have shown that atypical (most frequently patients with NRS) COVID-19 symptoms are frequent in older patients and are associated with higher mortality [29,32].Hariyanto et al. and Raymond Pranata et al., in a systematic review and meta-regression, found a significant association of extrapulmonary symptoms, such as delirium with death (OR 1.90 [1.55-2.33],p < 0.00001 and 1.50 [1.16, 1.94], p = 0.002, respectively).This relationship was not significantly influenced by age, sex, hypertension, diabetes, and dementia [33,34].Additionally, patients with NRS could develop profound hypoxemia without dyspnoea, called "silent or happy hypoxemia", which may deteriorate rapidly without warning and has been associated with increased mortality [35].However, this association remains controversial [36][37][38].
Early during the pandemic, respiratory symptoms and fever were used to detect patients with possible SARS-CoV-2 infection.However, we found that both patients presenting with and without respiratory symptoms early into the course of COVID-19 could subsequently develop respiratory failure and systemic complications, require oxygen support and die.Targeting patients with respiratory symptoms and/or reduced oxygen saturation will overlook those cases.Jiayi Tan et al., in a systematic review and metaanalysis, found that some public health interventions, such as stroke education campaigns on stroke symptom recognition and intention to call emergency medical services increased the estimated pool risk ratio (RR) for symptoms recognition (RR 1.20) and intention to reach emergency services (RR 1.19) [39].
Our study has strengths and limitations that should be recognised.Firstly, our study population was composed mainly of patients in HICs, which limits the generalisability of these results.Secondly, we do not have complete data on respiratory symptoms, nor extrapulmonary symptoms (i.e., gastrointestinal, cardiac, neurological, among others), during hospitalisation.Therefore, we cannot investigate the association of the progression and impact of respiratory symptoms, nor extrapulmonary symptoms, with outcomes in patients who present with RS or NRS.Moreover, our study had limited data on SARS-CoV-2 variants which restricted our ability to analyse their impact on COVID-19 disease progression.Future studies that incorporate detailed variant data are essential to provide a more in-depth understanding of their impact on COVID-19 patients admitted to hospital with and without respiratory symptoms.Finally, throughout the COVID-19 pandemic, hospitalised patients were treated with a wide range of medications and supportive care protocols, which may bias the factors associated with fatality using observational study methodologies in a fluctuating setting.However, including large numbers of patients over a long period adds to the robustness of our data.To our knowledge, this is one of the largest cohorts comparing patients with NRS and RS globally.
In conclusion, while many COVID-19 patients are hospitalised with respiratory symptoms, about one in seven do not have obvious respiratory symptoms on admission.These NRS patients are usually older and have multiple chronic conditions often unrelated to pulmonary comorbidities.While in the hospital, these patients are less likely to be admitted to the ICU and less likely to receive vasopressors and corticosteroids.About two in five patients may die, but their risk for in-hospital mortality is lower than those presenting with respiratory symptoms after adjusting for confounders.Therefore, more strategies should be implemented to identify patients with COVID-19 and to prevent fatal outcomes in this at-risk population.

Ethics and consent statement
This observational study required no change to clinical management.The ISARIC-WHO Clinical Characterisation Protocol was approved by the World Health Organization Ethics Review Committee (RPC571 and RPC572 on 25 April 2013).Institutional approval was additionally obtained by participating sites including the South Central Oxford C Research Ethics Committee in England (Ref 13/ SC/0149) and the Scotland A Research Ethics Committee (Ref 20/SS/0028) for the United Kingdom, representing the majority of the data.Requirement for consent was waived by the confidentiality advisory group of UK Health Regulations Authority and approved by the sponsor.
Other institutional and national approvals were obtained by participating sites as per local requirements.Regionally appropriate decisions regarding a waiver or requirement of patient consent and/or assent were made by each committee and implemented at the sites.b The reference group for comorbidities is not having the particular comorbidity/risk factor.

Fig. 1 .
Fig. 1.Flow diagram for the study showing the number of patients included in the analysis.

Fig. A. 1 .
Fig. A.1.Countries included in the analysis.

Fig. A. 2 .
Fig. A.2. Frequency of comorbidities for different age groups, stratified by respiratory symptoms.

Table A . 3
; Table A.4).Other risk factors associated with the highest increased mortality risks were pre-existing transplantation (HR 1.34 [1.14-1.57],p<Fig.2.Frequency of comorbidities for all patients, stratified by respiratory symptoms.B.W. Citarella et al.

Table 2
Hazard ratios (HR) of death by respiratory symptoms group from Cox Proportional Hazards analysis*.

Table 3
Physiological parameters of patients during the first 24 h, stratified by respiratory symptoms at hospital admission.

Table 4
Oxygen supplementation at any time during hospitalisation stratified by respiratory symptoms at hospital admission.

Table A . 1
Demographics of patients, stratified by respiratory symptoms.

Table A . 2
General symptoms at admission, stratified by respiratory symptoms.

Table A .
The reference group for comorbidities is not having the particular comorbidity/risk factor.Hazard ratios (HR) of death by symptoms group from Cox Proportional Hazards analysis a , restricted to 14 days hospitalisation a Cox proportional hazards model adjusted for age, sex, country, all comorbidities and risk factors.b

Table A .
a Cox proportional hazards model adjusted for age, sex, country, all comorbidities and risk factors.