Respiratory and Psychophysical Sequelae Among Patients With COVID-19 Four Months After Hospital Discharge

Key Points Question What respiratory, functional, and psychological sequalae are associated with recovery from coronavirus disease 2019 (COVID-19)? Findings In this cohort study of 238 patients with COVID-19 hospitalized in an academic hospital in Northern Italy, more than half of participants had a significant reduction of diffusing lung capacity for carbon monoxide or measurable functional impairment and approximately one-fifth of patients had symptoms of posttraumatic stress 4 months after discharge. Meaning These findings suggest that despite virological recovery, a sizable proportion of patients with COVID-19 experienced respiratory, functional, or psychological sequelae months after hospital discharge.


Introduction
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection can be completely asymptomatic or, conversely, can cause coronavirus disease 2019 (COVID- 19), for which the clinical outcomes range from mild upper airways symptoms to a severe disease with respiratory failure and a high fatality rate. 1 As of November 2020, more than 60 million people have been infected with SARS-CoV-2 worldwide, and more than 1.4 million people have died. 2 Since the beginning of the COVID-19 pandemic, many researchers focused attention on clinical features and prognosis of the acute phase of SARS-CoV-2 infection. 3,4 As a consequence, we are now far more able to estimate prognoses and optimize clinical treatment of patients with COVID-19 compared with the beginning of the pandemic. 5 In contrast, the type and severity of respiratory or functional sequelae COVID-19 are unknown.
While COVID-19 is a systemic disease, 6 the lungs are most commonly affected, with histopathological findings that may include diffuse alveolar epithelium destruction, capillary damage or bleeding, hyaline membrane formation, alveolar septal fibrous proliferation, and pulmonary consolidation. 7 As a consequence, the diffusion capacity of the lung for carbon monoxide (D LCO ) is commonly altered in patients who recover from COVID-19, 8 similarly to SARS and Middle East respiratory syndrome (MERS), illnesses that are associated with an impairment of lung function lasting months to years. 9,10 Impairment in exercise capacity often parallels D LCO reduction: patients who recover from SARS pneumonia have been found to have 6-minute walking test and 36-item Short Form General Health Survey scores persistently lower than the general population. 9 Moreover, the functional impairment associated with COVID-19 may also be associated with adverse psychological outcomes. A multidisciplinary approach investigating the functional and psychological aspects associated with COVID-19 may be more effective in disclosing potential sequelae associated with COVID-19. In this prospective cohort study, we aimed to investigate prevalence and clinical associations of functional and psychological impairment 4 months after recovery from COVID-19.

Methods
This cohort study was approved by the Comitato Etico Interaziendale Novara ethical committee. All participants provided written informed consent. This study was conducted in strict accordance with the principles of the Declaration of Helsinki 11 and reported following the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies.

Study Population
We contacted 767 consecutive patients (or their caregivers) aged 18 years or older who were declined participation (Figure). Among 238 patients who agreed to participate, 232 participants (97.5%) had diagnosis of COVID-19 confirmed during their hospital stay by reverse-transcriptionpolymerase chain reaction (RT-PCR) of a nasopharyngeal swab; in 1 participants whose RT-PCR test was negative for SARS-CoV-2, the diagnosis was confirmed by bronchoalveolar lavage. The remaining 5 participants were diagnosed according to a combination of serological tests positive for SARS-CoV-2 antibodies and suggestive computed tomography.
An electronic case report form was generated using the Research Electronic Data Capture software (Vanderbilt University) to collect clinical data following pseudonymization. Data entry was performed by clinicians involved in the treatment of patients with COVID-19; inpatient clinical data were retrospectively collected from clinical charts.
We collected data on patients' demographic characteristics and regular medication use, symptoms at COVID-19 diagnosis and complications during the hospital stay (retrospectively documented), and type and number of comorbidities, including hypertension, type 2 diabetes, dyslipidemia, chronic obstructive pulmonary disease (COPD), obesity, inflammatory bowel disease, chronic liver disease, autoimmune disease, hematological diseases, coronary artery disease (CAD), atrial fibrillation and other structural or arrythmogenic heart disease, endocrine diseases, chronic kidney disease (CKD), previous stroke or venous thromboembolism, anxiety and depression, or active malignant neoplasm. We additionally collected data on patients' symptoms at follow-up, including fever, cough, dyspnea, ageusia, anosmia, diarrhea, arthralgia, myalgia, chest pain, sore throat, headache, and perception of reduced tolerance to physical activity compared with before they contracted COVID-19.

Pulmonary Function Tests
All patients underwent standard pulmonary function testing (PFT) with a Quark PFT with X9 pneumotach (COSMED) for forced expiratory volume in 1 second (FEV 1 ), vital capacity, forced vital capacity (FVC), D LCO , D LCO constant, and total lung capacity. D LCO and total lung capacity were determined by the single-breath CO technique. The hemoglobin value was evaluated before PFT to apply the appropriate correction to D LCO .
The spirometer underwent calibration the day the test was performed, and barometric pressure and temperature were simultaneously recorded. A trained technician coached the patient, while a pulmonologist (E. C., E. P., or F. P.) was responsible for test validation and interpretation based on the 2005 the American Thoracic Society and European Respiratory Society statements. 12,13 Briefly, the following safety measures were adopted: in a dedicated room, a dedicated spirometer was used to avoid cross-infection of patients not included in this program. The technician (and the pulmonologist, if needed) used full PPE (ie, face mask, N95 respirator, gown, and gloves). To avoid cross-infections between patients included in the program, a mouthpiece with an antimicrobial filter was used and changed for every patient. At the end of each day, the room underwent disinfection.

Physical Performance Tests
We assessed the patients' physical performance with the Short Physical Performance Battery (SPPB), which includes balance assessment in standing position, walking speed for 4 m, and standing up from  with a good predictivity on the disability level in daily activities. A score greater than 10 is the expected value for healthy individuals. 14,15 However, it should be noted that SPPB may not distinguish performance level in highfunctioning patients. 16 To improve the sensitivity of functional impairment detection, patients with SPPB scores greater than 10 were tested with the 2-minute walk test to evaluate the residual aerobic capacity; the 2-minute walk test score was compared with reference data for an age-and sex-matched population. [17][18][19]

Psychological Symptoms Tests
We assessed the presence of posttraumatic stress (PTS) symptoms by administering the Impact of Event Scale-Revised (IES-R), 20 a 15-item self-rated 4-point scale based on how often an event has occurred in the past 7 days (0 indicates not at all; 1, rarely; 3, sometimes; 5, often). All IES-R items are anchored to a specific stressor. Besides the IES total subjective stress score, 2 subscales were identified. One subscale measured intrusive symptoms, including intrusive thoughts, nightmares, and intrusive feelings and imagery, using 7 items, with scores ranging from 0 to 35; the other subscale measured avoidance symptoms, such as numbing of responsiveness and avoidance of feelings, situations, or ideas, using 8 items, with scores ranging from 0 to 40.

Statistical Analysis
Data were analyzed using the Stata statistical software version 15.1 (StataCorp). Normality was assessed by Shapiro-Wilk test. The measures of centrality and dispersion chosen for continuous variables were medians and interquartile ranges (IQRs); comparisons between groups for these variables were performed using the Mann-Whitney test. Categorical variables, whenever dichotomous or nominal, were reported as frequencies and percentages and analyzed through the Pearson χ 2 , Cochran-Armitage test, or Fisher exact test, as appropriate. The primary end point was the proportion of patients with a D LCO less than 80% of expected. The study was sufficiently powered to detect as statistically significant a 0.12 increase in the proportion of patients with D LCO less than 80% of expected among survivors of COVID-19 compared with that observed in a reference population 21 (0.30 vs 0.18, respectively), with an α = .005. Secondary end points were prevalence of a more severe respiratory impairment (defined as D LCO <60% of expected), potentially associated with a higher risk of pulmonary fibrosis; factors associated with a D LCO less than 80% or less than 60% of expected; prevalence and factors associated with functional impairment (defined as SBBP score <11 or SBBP score Ն11 in presence of a 2-minute walk test score outside of age-and sex-matched reference range); and prevalence and factors associated with moderate to severe PTS symptoms.
To identify the associations with the different end points used, we conducted a univariate analysis, including the comorbidities with a biological plausible correlation with disease sequelae, age, sex, smoking status, intensive care unit (ICU) admission during hospital stay, number of comorbidities, and modality of oxygen delivery during hospital stay. P values were 2-sided, and statistical significance was set at P = .05. All associations with P < .20 were then included in logistic regression models.

Results
Among 238 patients included in analysis, the median (IQR) age was 61 (50-71) years, and 142 (59.7%) were men. The main characteristics of participants are listed in Table 1

JAMA Network Open | Infectious Diseases
Respiratory and Psychophysical Sequelae Among Patients With COVID-19 After Discharge     ( Table 3) found that the perception of reduced tolerance to physical exercise was associated with younger age and ICU admission during hospitalization.

Psychological Symptoms Tests
Finally, we tested patients for PTS symptoms. Results of the IES-R questionnaire were within reference ranges in 136 patients (57.1%), while in 61 patients (25.6%) had mild symptoms, 27 patients (11.3%) had moderate symptoms, and 14 patients (5.9%) had severe symptoms. Male sex was the only factor independently associated with the presence of moderate to severe PTS symptoms (eTable 5 in the Supplement; Table 4).

Discussion
Little is known about the lasting effects of SARS-CoV-2 infection in survivors of COVID-19. In this cohort study, we found that a significant proportion of survivors of COVID-19 experienced respiratory or functional impairment 4 months after hospital discharge, with clinically relevant psychological consequences. Indeed, at the end of follow-up, more than half of the study population still had D LCO less than 80% of expected. When a more stringent threshold of less than 60% of expected was applied, the proportion of patients with severe impairment decreased to 15%. Thus, a significantly less than 80% of expected 3 months after hospital discharge. The populations these studies described were significantly different from ours, being younger (median age, 48 years) and with a very low prevalence of smokers; accordingly, these studies included no patients with underlying pulmonary disease.
As for factors associated with reduced D LCO , female sex was a significant factor, possibly reflecting fitness level. 24 Interestingly, a history of CKD and the modality of oxygen administration during hospitalization were associated with reductions in D LCO , probably owing to a more severe acute illness. 25,26 Conversely, COPD and ICU admission emerged as factors associated with severe lung function impairment. A reduction of D LCO is associated with pulmonary fibrosis in different clinical settings, such as interstitial lung diseases and systemic sclerosis 27,28 ; whether survivors of COVID-19 with impaired D LCO are at increased risk of progressive lung fibrosis will require a longer follow-up. Indeed, fibrotic evolution has been described after SARS pneumonia. 29 Some degree of motor impairment was observed in 53.8% of our study population. Different factors might be invoked to explain this observation, including lung damage, circulatory limitation, muscle weakness, critical illness neuropathy, and myopathy. Our data suggest an association of pulmonary function impairment; indeed, D LCO and a history of COPD were independently associated with impaired physical function.
Interestingly, many people perceived that COVID-19 had a detrimental impact on their physical performances, and ICU admission being associated with this perception may be associated with deconditioning. Remarkably, age was not associated with reduced D LCO or impaired motor function.
In COVID-19, the case fatality rate increases with age 30 ; therefore, it would be reasonable to expect a higher burden of residual impairment in older patients. Our data are in contrast with this hypothesis.
We speculate that older people may have a higher baseline comorbidity burden, which was detrimentally associated with their survival probability during acute illness, but in survivors, the residual damage was not worse than in younger people. Essentially, this finding confirms that older individuals who survive COVID-19 may not be less able than their younger counterparts to revert to How long this loss will persist is unknown. Interestingly, one-third of patients reporting arthralgia and myalgia during the acute phase still experienced these symptoms at 4 months. Our findings are in line with 2 cohort studies. 33,37 However, the proportion of patients who were symptomatic in our study was lower than those reported in the studies by Carfi et al 33 and Carvalho-Schneidfer et al, 37 suggesting a progressive improvement over time. Overall, these findings suggest that many patients experience a slow recovery after the acute phase of COVID-19.

Limitations
This study has several limitations. A key limitation of our study is patient selection: we contacted only patients who had a severe enough COVID-19 to be admitted to a hospital. Moreover, many discharged patients declined study participation for reasons that may have included perceived full recovery in some or the inability or unwillingness to attend extra visits. This might have generated a selection bias, considering that the real proportion of patients still experiencing functional or psychological sequelae might have been lower if all had participated. Our psychological evaluation was limited to PTS symptoms, which does not allow us to draw definitive conclusions about the full psychological impact of COVID-19, which may include many other aspects, such as sleep disturbances, anxiety, and depression. Additionally, we reported an unexpectedly high residual mortality in the first few months after hospital discharge. This is a novel, worrisome finding that warrants further examination.

Conclusions
In this cohort study, a significant proportion of patients hospitalized for COVID-19 still reported a high proportion of symptoms associated with COVID-19 up to 4 months after hospital discharge, with reduced exercise tolerance being the most common. Other midterm sequelae of COVID-19, such as respiratory and physical functional impairment, may impact psychological health. Residual lung injury may be associated with reduced quality of life in survivors of COVID-19. Although age is a major

JAMA Network Open | Infectious Diseases
Respiratory and Psychophysical Sequelae Among Patients With COVID-19 After Discharge