Study design and participants
The study was a prospective, randomized controlled trial. Patients were enrolled from Union Hospital, Tongji medical college, Huazhong university of science and technology (Wuhan, Hubei Province of China) between February 7th, and March 31st, 2020. Pulmonary rehabilitation program was taken place in department of respirotary and critical care medicine, in a medical college affiliated teaching hospital. This study was approved by the Institutional Review Board of Wuhan Union Hospital (project approval number 2020-0127), and participants provided written informed consent.
Randomisation was conducted at the trial centre and the participants were randomly allocated to either the pulmonary rehabilitation exercise group or the control group at a ratio of 1:1. Pulmonary rehabilitation exercise group assignments were generated using Statistical Product and Service Solutions (SPSS) 25.0 software (IBM Corp., Armonk, NY, USA) and communicated using sealed opaque envelopes.
Eligible patients were those who were diagnosed as COVID-19 and were admitted to the hospital during the study. Inclusion criteria including:(1) patients aged from 18 to 70 years; (2) patients whose percutaneous oxygen saturation (SpO2) > 90% while the patients breathed ambient air or had oxygen support. (3) patients who were consciousness and could follow instructions. (4) patients whose disease onset > 2 weeks and had been hospitalized > 72 hours.
Exclusion criteria inculding: (1) pregnant women; (2) unstable coronary heart disease; (3) mean blood pressure < 60mmHg, or blood pressure > 140/90 mmHg; (4) peak temperature > 38℃; (5) unstable hemodynamics; (6) deep venous thrombosis; (7) severe hepato-renal dysfunction; (8) mental disorders or learning disability; (9) patients who couldn’t take at least seven days of pulmonary rehabilitation program.
Criteria of discharge: (1) Normal temperature > 3 days; (2) Respiratory symptoms improved significantly; (3) Pulmonary imaging showed that acute exudative lesions were absorbed obviously; (4) The 2019 nCov nucleic acid test of sputum, nasopharynx swab or other respiratory samples was negative twice(sampling time should have an interval of at least 24hours).
Data collection
Baseline data of all patients were collected, including demographic (age and sex, height, weight, body mass index), general condition, symptoms, comorbidities, vital signs (body temperature, heart rate, respiratory rate, and blood pressure), laboratory tests (blood routine test, arterial blood gas analysis (ABG)), chest computed tomography scan, Barthel activities of daily living (ADL) scores7 and manual muscle test (MMT). Medical and respiratory support, complications, and outcomes were also analysesed. Vital signs, rating of perceived exertion scale (RPE)8 and SpO2 were recorded prior to and after physical rehabilitation exercise. Data of symptoms, MMT and arterial blood gas analysis were recorded before discharge. Data of ADL was recorded after 7days of after exercise and before discharge and SpO2 was recorded after 3days and 7days of after exercise and before discharge.
Assessment Measures
Barthel ADL was modified in 1988 to a 20 point scale that measures in increments of 1 point, which was from 0 (fully dependent)to 20(fully independent). The final score could be multiplied by 5 to obtain a 100 point score, and it was proposed that scores of 0–20 indicating “total” dependency, 21-60 indicating “severe” dependency, 61–90 indicating “moderate” dependency, and 91–99 indicating “slight” dependency. The Barthel ADL assessment comprised: feeding, grooming, bathing, dressing, toilet use, presence of fecal incontinence, presence of urinary incontinence, transfers (e.g. moving from wheelchair to bed), walking on an even surface (or propelling a wheelchair if unable to walk), and ascending and descending stairs. It was already widely used as the measurement of daily living activities and had become a standard measure of physical disability in clinical practice7.
RPE scale was a 15-point scale, ranging from 6 to 20 (corresponds to an approximate heart rate between 60 and 200), designed to estimate how heavy and strenuous the work is perceived8.
Scoring of the manual muscle testing(MMT): Grade 0: no contraction of the tibialis anterior. Grade 1: visible contraction of the tibialis anterior,Grade 2: active movement with gravity eliminated. Grade 3: active movement against gravity. Grade 4: active movement against gravity and moderate resistance. Grade 5: normal response (against gravity with full resistance).
Intervention
Respiratory support and medication treatment was applied to each patient, which was consistent with the guide for the prevention and treatment of coronavirus disease 20199 .Each patient received supportive treatment. The pulmonary rehabilitation program based on muscle training exercise, including bedside exercises for patients who were able to stand, and in-bed exercises for those patients who were bedridden. Procedures of excise training are detailed further in e-Appendix 1.Patients who were able to stand, took step 1 to step 6 program, which respiratory muscle, torso and limb movements were designed in different positions. For those patients who were bedridden, took step 1 to step 3 program, movement of upper and lower limbs and the change of body position were implemented to all.
Breathing rehabilitation would take 15 minutes each time, twice a day. Patients who took inbed or bedside exercise depended on the decision of doctor. Patients who had a RR<28, HR<100, and lower lim MMT>4 would preferred bedside exercise. Whether patients performed muscle training exercise and pulmonary rehabilitation under O2 therapy decided by SpO2 during exercise. In the course of rehabilitation, patients followed original respiratory support, and adjust the level of respiratory support according to the change of SpO2, making ensure that SpO2 > 88% could be maintained in the course of rehabilitation. It should be terminated if SpO2 could not be maintained above 88%. The program was performed by two respiratory therapists in intervention group, and supervised by a physician. Each patients took at least seven days of pulmonary rehabilitation program based on muscle training exercise. In control group, no special exercise training were guided by docters.
In the process of muscle training exercises, once one of the following situations occured, the patient should immediately stop exercising:10 (1) Borg scale > 3 (10 points in total); (2) one of the symptoms: chest tightness, dyspnea, dizziness, headache, unclear vision, palpitation, sweating, inability to maintain balance, pain (chest pain, strong headache or strong abdominal pain) or fatigue; (3) other situations which clinicians judged that were unsuitable continuing pulmonary rehabilitation: heart rate < 40 or > 130 beats/min; mean blood pressure < 65mmHg or > 110mmhg, or more than 20% change from baseline; respiratory rate: > 40 times/min or < 5 times / min; SpO2 ≤ 88% or decrease > 4%; new arrhythmia; new myocardial ischemia or unstable angina.
Outcomes evaluation
Primary outcome was improvement of ADL following pulmonary rehabilitation before enrollment and before discharge. Intra-group and inter-group comparisons were conducted. Secondary outcome were changes of MMT, ABG and length of hospital staying (LOS). Adverse effects of pulmonary rehabilitation were also analysed during pulmonary rehabilitation program.
Sample Size
Since the COVID-19 was considered as a new disease, there was no reference data for pulmonary rehabilitation. Sample size was calculated based on our pilot study. In the previous work, we carried out pulmonary rehabilitation exercise on 5 patients, and observed their symptom changes and ADL score improvement. Meanwhile, We also referred to data based on community-acquired pneumonia patients, which previously reported by Anderson J and Simone DC.11 We estimate a 35% improvement in ADL score between the pulmonary rehabilitation group and control group with a two-sided test size of 5% and power of 80%. 27 eligible participants were required per study arm. Considering that about 10% of the enrolled cases might dropped out of the study, we finally set the sample size as 60 cases, with 30 cases in the pulmonary rehabilitation group and 30 cases in the control group. The sample size was determined using SPSS 25.0 software.
Statistical analysis
SPSS 25.0 software was used for the statistical analysis. Results are expressed as mean±SD forvariables normally distributed and abnormally distributed data were expressed as median (IQR). For primary and secondary outcome, intra-group comparisons of quantitative data were done using the paired Student’t t test or Wilkinson rank test for variables which were not normally distributed. Inter-group comparisons were done using unpaired Student’s t tests or with Mann-Whitney U test for variables non-normally distributed. Categorical variables were compared using X2 tests or Fisher test. Correlation analysesbetween two ADL and variables were done by Spearman correlation. P value <0.05 indicated a statistically significant difference.