Long-term impact of COVID-19 hospitalisation among individuals with pre-existing airway diseases in the UK: a multicentre, longitudinal cohort study – PHOSP-COVID

Background The long-term outcomes of COVID-19 hospitalisation in individuals with pre-existing airway diseases are unknown. Methods Adult participants hospitalised for confirmed or clinically suspected COVID-19 and discharged between 5 March 2020 and 31 March 2021 were recruited to the Post-hospitalisation COVID-19 (PHOSP-COVID) study. Participants attended research visits at 5 months and 1 year post discharge. Clinical characteristics, perceived recovery, burden of symptoms and health-related quality of life (HRQoL) of individuals with pre-existing airway disease (i.e., asthma, COPD or bronchiectasis) were compared to the non-airways group. Results A total of 615 out of 2697 (22.8%) participants had a history of pre-existing airway diseases (72.0% diagnosed with asthma, 22.9% COPD and 5.1% bronchiectasis). At 1 year, the airways group participants were less likely to feel fully recovered (20.4% versus 33.2%, p<0.001), had higher burden of anxiety (29.1% versus 22.0%, p=0.002), depression (31.2% versus 24.7%, p=0.006), higher percentage of impaired mobility using short physical performance battery ≤10 (57.4% versus 45.2%, p<0.001) and 27% had a new disability (assessed by the Washington Group Short Set on Functioning) versus 16.6%, p=0.014. HRQoL assessed using EQ-5D-5L Utility Index was lower in the airways group (mean±SD 0.64±0.27 versus 0.73±0.25, p<0.001). Burden of breathlessness, fatigue and cough measured using a study-specific tool was higher in the airways group. Conclusion Individuals with pre-existing airway diseases hospitalised due to COVID-19 were less likely to feel fully recovered, had lower physiological performance measurements, more burden of symptoms and reduced HRQoL up to 1 year post-hospital discharge.


Introduction
Early in the COVID-19 pandemic, the prevalence of asthma and COPD in hospitalised patients with COVID-19 was low, raising the possibility that pre-existing airway diseases or inhaled corticosteroid (ICS) therapy might play a protective role against contracting SARS-CoV-2 infection or severe outcomes [1,2].However, later reports found no evidence to support these theories [3,4], and the number of hospitalised patients with pre-existing airway diseases increased, likely due to relaxation in social distancing [5].Patients with COPD who were hospitalised were at increased risk of severe COVID-19 illness or death, likely due to factors such as older age, increased number of comorbidities and reduced physiological reserve to survive critical illness [6][7][8][9].Worse clinical outcomes in hospitalised patients with pre-existing asthma were mainly observed in those with severe asthma or those who required multiple courses of oral corticosteroids in the preceding year [10,11].Results from a large UK hospitalised cohort found that patients with asthma were more likely to receive critical care than those without asthma [12].Little is known about the impact of SARS-CoV-2 in patients with pre-existing bronchiectasis largely due to the scarce literature, but there is a suggestion that individuals with pre-existing bronchiectasis had increased risk of worse clinical outcomes after COVID-19 infection [13,14].
The long-term sequelae following COVID-19 hospitalisation in individuals with pre-existing airway diseases are unknown.An international consensus exercise to determine research priorities in patients with pre-existing airway diseases following COVID-19 hospitalisation identified the need to determine the short-and medium-term effects of COVID-19 infection in this group [15].Here we report on the results from a large UK-based multicentre cohort study of hospitalised COVID-19 survivors (Post-hospitalisation COVID-19 study (PHOSP-COVID) study).

Study design and participants
The PHOSP-COVID is a UK national multicentre prospective longitudinal cohort study.The PHOSP-COVID study methods have been described in detail elsewhere [16].Participants were invited to attend two research visits: the first visit between 2 and 7 months; and the second visit between 10 and 14 months post-hospital discharge.Participants were included in the airways group if they self-reported a history of asthma, COPD or bronchiectasis prior to their initial hospitalisation with COVID-19.A small number of the participants indicated a history of combined pre-existing asthma and COPD or bronchiectasis and COPD.Clinical characteristics of this group revealed significant smoking history and older mean age; therefore, these individuals were assigned to the COPD group.Written informed consent was obtained from all participants.The study was approved by the Leeds West Research Ethics Committee (20/YH/0225) and registered on the ISRCTN Registry (ISRCTN10980107).

Statistical analysis
Descriptive statistics were used to describe participant characteristics.Continuous variables are presented as mean±SD, or medians and interquartile ranges, as appropriate.Binary and categorical variables are presented as counts and percentages of available data.No imputation was performed for the missing data.Results were not adjusted for multiple testing.t-test, analysis of variance (ANOVA F-test) and Kruskal-Wallis H-test were used to compare parametric and non-parametric continuous data as appropriate.Chi-squared test was used to compare categorical data.We did not adjust for cofounders.To examine the predictors of recovery at the second research visit, the participants with pre-existing airway diseases were dichotomised into: "recovered" group for those who answered "Yes" to the perceived full recovery question or "not recovered" group including those who answered "No" or "Not sure" using the PSQ tool.Univariable and multivariable logistic regression were reported to identify predictors of recovery.Only explanatory variables available at hospital discharge were used in the multivariable logistic model comprising: age as a factor, sex at birth, ethnicity, Index of Multiple Deprivation, body mass index (BMI), number of comorbidities, admission duration, severity of acute illness using World Health Organization (WHO) Clinical Progression Scale, history of pre-existing neuropsychiatric disease and the use of systemic steroids during acute admission.R (version 3.6.3)and Stata (version 16.0) were used for all data analysis.
Comparison of the participants' characteristics showed the airways group to have: more females (48.5% versus 35.6%, p<0.001), more from a White ethnic background (82.1% versus 72.9%, p<0.001), a higher prevalence of pre-existing neuropsychiatric comorbidity (29.9% versus 18.2%, p<0.001), higher BMI (median 31.9 versus 30.9 kg•m −2 , p=0.001) and more likely to have received systemic steroids during hospital admission (63.3% versus 54.6%, p=0.001) compared to the non-airways group.There was no difference in age, length of hospital admission or treatment with antibiotics or anticoagulants between the two groups.The level of organ support during acute admission was comparable between the two groups with the exception that receiving invasive mechanical ventilation and other organ support (WHO class 7-9) was lower in the airways group (10.0%versus 14.2%, p=0.041).At hospital discharge, 63.9% of the airways group were prescribed a form of ICS therapy and 27.2% were on antidepressant medications compared to 1.9% and 16.9% in the non-airways group, respectively.The breakdown of the different classes of prescribed medications upon discharge and additional reported changes at 5-month and 1-year visits are listed in supplementary table S1.

Results from the 5-month visit
The first research visit was attended by 2570 participants at a median of 5.5 months (IQR 4.1-6.4)from hospital discharge, labelled here as "5-month" visit.A total of 595 out of 2570 (23.2%) participants reported a history of pre-existing airway disease prior to hospital admission (table 2).Assessments at the 5-month visit revealed that the airways group participants were more likely to have symptoms consistent with anxiety (34.4% versus 22.8%, p<0.001), depression (44.2% versus 26.5%, p<0.001), post-traumatic stress disorder (PTSD) (19.4% versus 11.6%, p<0.001) and greater breathlessness measured using the Dyspnoea-12 questionnaire (mean±SD 10.2±9.7 versus 5.3±7.4,p<0.001).The airways group participants had a higher percentage of impaired mobility measured using short physical performance battery ⩽10 (59.6% versus 48.5%, p<0.001) and a lower percentage of predicted incremental shuttle walk test distance (52.4% versus 58.7%, p=0.001).They were more likely to be frail using the Rockwood Clinical Frailty score without features of cognitive impairment.Pulmonary function tests revealed lower spirometry measurements in the airways group but no difference in gas transfer measurements between the two groups.The airways group had higher levels of blood neutrophils, eosinophils and higher numbers with systemic inflammation measured by C-reactive protein (CRP) of more than 5 mg•L −1 (table 2).
The perceived recovery question demonstrated a lower proportion of participants in the airways group reporting "full recovery" (19.7% versus 27.6%, p=0.005) (table 3).The participants who attended the 5-month visit were assigned one of the previously identified four cluster memberships of recovery phenotypes [17]: very severe, severe, moderate with cognitive impairment, and mild mental and physical impairment (see supplementary material SM1).Recovery cluster assignment was different between the airways and non-airways groups with a higher proportion of those with pre-existing airway diseases assigned to the very severe mental and physical impairment cluster (32.5% versus 17.5%) and a smaller proportion assigned to the mild cluster (21.1% versus 32.7%), p<0.001 (figure 2, supplementary table S2).
HRQoL assessed using EQ-5D-5L UI and EQ-5D-5L VAS showed lower estimated pre-hospitalisation levels in the airways group (0.74±0.27 versus 0.84±0.21and 73.7±18.5 versus 81.3±16.8,all p<0.001), respectively.The participants in the airways group reported a drop in the EQ-5D-5L UI of 0.12±0.26units similar to the non-airways group (0.11±0.21 units, p=0.451).A higher proportion of the airways group reached the threshold for a new disability using the Washington Group Short Set on Functioning (WG-SS), 24.5% versus 17.5%, p=0.002.In the airways group, the burden of breathlessness, cough, fatigue, sleep disturbance and pain measured using the PSQ scale was higher both at pre-COVID estimate and at the 5-month visit (table 3, figure 3).However, delta difference between pre-COVID level and the 5-month visit was smaller in the case of breathlessness (2.3 versus 2.9, p<0.001) in the airways group compared to the non-airways group but similar in cough, fatigue, sleep disturbance and pain.
The differences in clinical characteristics of the participants attending the 5-month visit stratified by the underlying class of airway disease are included in supplementary tables S3 and S4.The COPD group were older, had more male participants, were mainly from a White background, were less likely to have received invasive ventilation, had more comorbidities and were more likely to be assigned to the moderate mental and physical impairment with cognitive impairment cluster.The COPD group at the 5-month visit were frailer, had a higher percentage of impaired mobility, more evidence of cognitive impairment and showed features of anxiety and depression in over 30% of the group.They also had the lowest spirometry measurements but comparable gas transfer readings.Blood tests revealed higher levels of neutrophils, eosinophils, CRP and higher proportion of participants with heart failure or renal impairment.The COPD group had the smallest drop in EQ-5D-5L UI and VAS measurements despite having the lowest estimates pre-COVID (figure 4).They also had minimal increase in the burden of breathlessness, cough and fatigue measured using PSQ despite having higher levels of baseline burden (figure 3).
The participants with underlying asthma were characterised by: younger age, more females, higher BMI, more likely to report "not recovered" and more than a third were assigned to the very severe cluster.The asthma group had the largest drop in EQ-5D-5L UI and VAS and the highest increase in burden of fatigue (supplementary table S4, figure 3).The change in breathlessness and cough in the asthma group was similar to those without pre-existing airway diseases (figure 3).
Results from the 1-year visit A total of 2100 participants attended a second research visit at "1 year" at a median of 12.6 months (IQR 11.8-13.4)from hospital discharge.A total of 479 (22.8%) participants had a history of pre-existing airway diseases (figure 1).At 1-year visit, the airways group participants remained less likely to report full recovery compared to the non-airways group (20.4% versus 33.2%, p<0.001) and were more likely to have features consistent with anxiety, depression, PTSD, increased frailty, reduced physical performance and  higher CRP levels (table 2).HRQoL measurements at 1 year in the airways group were lower than in the non-airways group and showed no improvement from the 5-month levels with no recovery to pre-COVID estimates (figure 4).A higher proportion of the airways group reached the threshold for a new disability using the WG-SS (27% versus 16.6%, p=0.014).
In the airways group there was some improvement in the burden of symptoms between the 5-month and 1-year visit: anxiety (34.4% to 29.1%), depression (44.2% to 31.1%), cognitive impairment (15.4% to 10.3%) and breathlessness measured using Dyspnoea-12 (mean 10.2 to 8.5), table 2. Despite the participants in the airways group having a higher burden of the symptoms measured using PSQ compared to the non-airways group at 1 year, there was a trend towards improvement from the 5-month levels (figure 3).Clinical characteristics of the participants who attended the 1-year visit stratified by the underlying class of airway diseases are available in supplementary tables S5 and S6.

Factors predicting recovery at 1 year
At the 1-year visit, data about perceived recovery were available in 403 out of 479 (84.1%) of the airways group and 1384 out of 1621 (85.3%) of the non-airways group.The characteristics of the recovered participants in the airways group are listed in supplementary table S7.The multivariable logistic regression model did not identify any statistically significant features to predict recovery at 1 year post discharge in the airways group; however, non-White ethnicity, age ⩾70 years and receiving noninvasive respiratory support during initial admission were associated with increased likelihood of recovery (supplementary table S8, figure SF1).In contrast, female sex, history of neuropsychiatric comorbidity, increased level of deprivation, having one or more comorbidity, and receiving invasive ventilation/organ support were suggestive of reduced likelihood of full recovery.These predicting factors were similar in the non-airways group with the features of female, non-White ethnicity and pre-existing neuropsychiatric comorbidity reaching statistical significance (supplementary table S9 and figure SF2).

Discussion
To our knowledge, this is the first report to focus on the long-term impact of COVID-19 hospitalisation on individuals with pre-existing airway diseases using results from a large multicentre prospective longitudinal UK cohort study.Around a quarter of the PHOSP-COVID cohort had a history of pre-existing airway diseases with the majority of those reporting asthma.Individuals with pre-existing airway diseases were less likely to feel fully recovered at 5-month and at 1-year post-hospital discharge and they had a significant burden of anxiety, depression, PTSD, breathlessness, cough and fatigue compared to the non-airways group.There was evidence of reduced physiological performance, lower spirometry measurements, reduced level of HRQoL both pre-COVID and at the follow-up visits with raised neutrophils, eosinophils and systemic inflammation measured using CRP in the airways group.
Although the prevalence of airway diseases in our cohort might seem high, it is comparable to reports from the large ISARIC study and to the prevalence of asthma in the age group of 55-64 years in the UK [12].Furthermore, around two-thirds of the airways group were prescribed inhaled bronchodilators or ICS therapy.
The results from this cohort study support previous findings from smaller studies including a UK-wide survey where patients with pre-existing lung diseases were more likely to report "breathing complications" after contracting .Another UK-based online survey among patients with underlying asthma revealed that more than half of the participants experienced features of "long COVID" that were not related to personal characteristics such as age, sex, ethnicity or household income [19].A cohort study of 2649 participants from Russia revealed that patients with chronic pulmonary diseases were more likely to report respiratory symptoms and chronic fatigue [20].Numerous reports identified breathlessness as one of the commonest persistent symptoms post-COVID in the general population [21][22][23].In our study, people with pre-existing airway disease had a higher burden of breathlessness measured using the Dyspnoea-12 questionnaire and PSQ breathlessness scale at both visits.However, the non-airways group had the largest increase in breathlessness from pre-COVID levels measured using the PSQ scale.As expected, spirometry results were lower in the airways group, but we observed no difference in gas transfer measurements between the two groups.Despite the smaller number of participants who completed these procedures (due to the restrictions around aerosol-generating procedures), the latter finding suggests that a pre-existing airway disease is not necessarily a major risk factor for further lung function impairment post-COVID-19 hospitalisation and the pathophysiology of persistent breathlessness is likely to be multifactorial [24].
Other symptoms of cough, sleep disturbance and pain were higher in the airways group at pre-COVID baseline but increased in similar proportion to the non-airways group, resulting in an overall higher burden of these symptoms at both visits in those with pre-existing airway disease.Fatigue is highly prevalent in patients with airway diseases [25,26], and the COPD group in our cohort had double the level of fatigue burden pre-COVID compared to the non-airways group.Interestingly, the asthma group, which was dominated by female participants, demonstrated the highest increase in fatigue burden compared to the COPD and the non-airways groups.Multiple reports have identified female sex as an independent risk factor for developing chronic fatigue post- 27,28].
Patients with pre-existing airway diseases are known to have reduced HRQoL compared to controls [29][30][31].In our study, HRQoL in the airways group was reduced both at pre-COVID estimates and at both follow-up visits, with minimal improvement between 5 months and 1 year post discharge.This was similar to the results of 1-year follow-up of 1276 COVID-19 survivors from Wuhan, China [21] and a German study of COVID-19 survivors who required intensive care unit admission [32], where HRQoL remained reduced at 1 year after hospital discharge.Interestingly, in our cohort the magnitude of decline in HRQoL from pre-COVID estimates to the research visits was similar in both groups, suggesting that patients with pre-existing airway diseases are not at increased risk of significant deterioration of HRQoL compared to those without airway diseases.There was a general trend of improvement in the burden of symptoms between the 5-month and 1-year visit in the overall cohort and more specifically in the airways group.This was similar to the findings from the Wuhan study [21].
Owing to the small number of the bronchiectasis cases in this cohort, reaching robust conclusions about the long-term sequelae of COVID-19 in this disease is challenging, but our results support earlier indications that pre-existing bronchiectasis is associated with increased morbidity after COVID-19 [13,14].
The multivariable logistic regression suggested that in both airways and non-airways groups, being female, more severe acute illness, increased number of comorbidities and history of pre-existing neuropsychiatric diseases were all associated with reduced likelihood of reporting full recovery at 1 year post discharge.These identified risk factors were consistent with previously published systematic reviews and meta-analyses exploring the risk factors of prolonged symptoms post-COVID in hospitalised individuals [21,[33][34][35].
Results from this cohort study are important for policy decisions and clinical practice as they highlight the significant burden of symptoms and morbidity in an already vulnerable group [36].The challenges facing healthcare providers globally will likely worsen the clinical outcomes of individuals with pre-existing airway diseases unless healthcare provision is prioritised in this group in the form of offering pulmonary rehabilitation, reviews of inhaler technique, delivery of vaccinations, clinical monitoring and self-management plan implementation [37,38].The high prevalence of anxiety, depression and PTSD in this group highlights the rising need to improve access to mental health and counselling services [39,40].
The strengths of this cohort analysis include reporting the findings from a large multicentre study with in-depth assessments using validated and novel tools to measure recovery, burden of symptoms and HRQoL in hospitalised COVID-19 individuals with pre-existing airway diseases.Our study had several limitations.First, the high prevalence of symptoms across the whole of the PHOSP-COVID study participants raises the possibility of selection bias where individuals with high burden of symptoms choose to participate in the study.Second, we relied on the recall of certain measurements by the participants including the history of pre-existing illness and the pre-COVID estimates of HRQoL and burden of symptoms.This includes the use of the study-specific PSQ, which is not externally validated but supports results from other validated tools, e.g., Dyspnoea-12, FACIT fatigue subscale scores.Although a large proportion of the airways group individuals were prescribed inhaled therapy on hospital discharge, which supports the self-reported diagnosis of pre-existing airway disease, this observation alone cannot confirm a pre-existing diagnosis due to discrepancy between prescribing inhaled therapy and the prevalence of diagnosed asthma and COPD [41,42].Third, the lack of a control group of participants with pre-existing airway diseases who were not hospitalised for COVID-19 infection is also a limitation; however, the overall design of the PHOSP-COVID study did not include a control group of non-hospitalised individuals.Fourth, it is not clear how much of the reported burden of symptoms up to 1 year post discharge can be attributed to the pre-existing airway diseases versus emergent impaired health status.Fifth, no data were collected regarding the frequency or severity of exacerbations of pre-existing airway diseases nor the use of rescue medications.Sixth, the participants in this cohort were mainly individuals who were hospitalised during the first wave of the pandemic in the UK prior to the widespread use of in-hospital COVID-19 therapeutic interventions and the uptake of vaccination, therefore limiting the generalisability of these findings to the overall population.
In conclusion, individuals with pre-existing airway diseases who were hospitalised due to COVID-19 were less likely to feel fully recovered and had greater burden of symptoms and reduced HRQoL up to 1 year post discharge.Prioritisation of clinical care provision in this group is essential to minimise further decline in health status in an already premorbid population.
ERJ OPEN RESEARCH ORIGINAL RESEARCH ARTICLE | O. ELNEIMA ET AL.

FIGURE 3
FIGURE 3 Symptoms over time.Symptoms change measured using Patient Symptoms Questionnaire (PSQ) from pre-COVID, to 5-month and 1-year visits stratified by presence or absence of pre-existing airway diseases: a) breathlessness, b) cough and c) fatigue.***p<0.001.

TABLE 1
Patient characteristic

TABLE 2
Patient characteristics at the 5-month and 1-year research visits stratified by the presence of pre-existing airway diseases

TABLE 3
Recovery, health-related quality of life and symptoms burden at the 5-month and 1-year research visits stratified by the presence of pre-existing airway diseases Recovery cluster membership assignment at 5-month visit stratified by the underlying class of airway disease.The four clusters are: very severe mental and physical impairment, severe mental and physical impairment, moderate mental and physical impairment with cognitive impairment, and mild.
Change in health-related quality of life (HRQoL) measured using EQ-5D-5L Utility Index from pre-COVID to 5-month and 1-year visit by presence or absence of pre-existing airway diseases.COPD: chronic obstructive pulmonary disease.p-values calculated using t-test: ***p<0.001.