Study characteristics.
Within the follow-up period of 3 months, none of the patients was lost to follow-up. But 2 patients died from trauma at 1.8 and 2.7 months respectively. Related data were considered as censored data.
In severe COVID-19 with COPD group, 35 (24.8%) patients died of COVID-19 and related complications with a median survival time of 14 days (range 3–89 days). In severe COVID-19 group, 87 (40.3%) patients died of COVID-19 and related complications with a median survival time of 12 days (range 3–83 days), 2 patients died of trauma.
None of our subjects received tocilizumab, nintedanib and pirfenidone treatment. All subjects received the third vaccine in one year. Subjects with COPD were under inhaler treatment. General demographics and basal characteristics of subjects were shown in Table 1.
Table 1
Basal characteristics and general demographics and of subjects
Category
|
Severe COVID-19 with COPD
|
Severe COVID-19
|
Subjects (n)
|
141
|
216
|
Age(years, mean ± SD)*
|
80.6 ± 9.9
|
81.6 ± 11.5
|
Gender*
|
|
|
Male (n, %)
|
107(75.9)
|
150(69.4)
|
Female (n, %)
|
34(24.1)
|
66(30.6)
|
Smoking (n, %)*
|
|
|
Ex
|
74(52.5)
|
75(34.7)
|
Current
|
15(10.6)
|
9(4.2)
|
Never
|
52(36.9)
|
132(61.1)
|
Pack years of smokers ( mean ± SD)*
|
28.8.±10.5
|
20.2 ± 8.6
|
Receiving antiviral therapy in 5 days (n, %)
|
92(65.2)
|
133(61.6)
|
OI at admission( mean ± SD)*
|
174.5 ± 54.7
|
198.3 ± 67.1
|
* P < 0.05 for comparison between severe COVID-19 with COPD group and severe COVID-19 group.
SD, standard deviation; OI: oxygenation index
The influence of the time to initiate systemic corticosteroid therapy on lung fibrosis and pulmonary function.
After three months, survivors were received the examination of chest HRCT. Lung fibrosis score on HRCT was evaluated. For fibrosis score of COPD patients, basal fibrosis score before COVID-19 was reduced. All survivors of severe COVID-19 with COPD group received systemic corticosteroid therapy in 14 days, they were divided into 2 subgroups, the 1–7 day subgroup and the 8–14 day subgroup according to the time to initiate systemic corticosteroid therapy after diagnosed with COVID-19. Survivors of severe COVID-19 group were divided into 3 subgroups, the 1–7 day subgroup, the 8–14 day subgroup and 15–21 day subgroup. If systemic corticosteroid was never used, survival time was recorded as the time to initiate systemic corticosteroid therapy.
The 1–7 day subgroups had a lower lung fibrosis score, both in severe COVID-19 with COPD group and severe COVID-19 group (both P = 0.0000). The 1–7 day subgroup of severe COVID-19 with COPD group had a lower lung fibrosis score compared with corresponding subgroup of non-COPD group (P = 0.0497). No statistical difference was observed between the 8–14 day subgroups (P = 0.3924). Data were shown in Table 2.
Patients received pulmonary function measurement three months later after diagnosed with COVID-19 (Table 2). 8 patients couldn’t tolerate pulmonary function measurement. Because COPD patients had baseline pulmonary function impairment, we didn’t make comparisons between two groups. In Severe COVID-19 with COPD group, the 1–7 day subgroup had higher FVC% and DLCO% value (P = 0.0010 and 0.0001, respectively), while no significant difference of FEV1% value was observed (P = 0.1950). Compared with other two subgroups, the 1–7 day subgroup had higher FVC% value (P = 0.0130 and P = 0.0054) and higher DLCO% value ( P = 0.0491 and P = 0.0440) in severe COVID-19 patients, while no significant difference of FEV1% value was observed among three subgroups (P = 0.1920 ,0.6482, and 0.9813).
Table 2
3-month lung fibrosis score and pulmonary function of subgroups with different initiating time of systemic corticosteroid.
Category
|
Severe COVID-19 with COPD
Time to initiate corticosteroid therapy (day)
1–7 8–14
|
Severe COVID-19
Time to initiate corticosteroid therapy (day)
1–7 8–14 15–21
|
Subjects (n)
|
69
|
37
|
56
|
56
|
15
|
lung fibrosis score ( mean ± SD)
|
0.86 ± 0.67
|
1.68 ± 0.75
|
1.09 ± 0.09
|
1.82 ± 0.11
|
2.60 ± 0.24
|
FEV1% (mean ± SD)
|
63.18 ± 9.58
|
60.68 ± 8.11
|
91.09 ± 16.60
|
86.69 ± 8.43
|
87.46 ± 5.55
|
FVC% mean ± SD)
|
83.79 ± 14.61
|
73.21 ± 15.20
|
84.47 ± 13.76
|
77.18 ± 8.54
|
73.77 ± 12.43
|
DLco% ( mean ± SD)
|
69.09 ± 13.86
|
58.47 ± 8.26
|
69.87 ± 10.32
|
64.80 ± 11.62
|
61.23 ± 14.56
|
Note: each group had 4 patients couldn’t tolerate pulmonary function measurement. |
The influence of the 1st week dose of systemic corticosteroid on lung fibrosis and pulmonary function.
The cumulative dose of systemic corticosteroid was calculated for the 1st week after initiating systemic corticosteroid therapy. For patients died in the first week, the cumulative dose was calculated as daily dose multiply by 7. If systemic corticosteroid was never used, the cumulative dose was recorded as 0. The doses of all kinds of systemic corticosteroids were transferred to dexamethasone injection. All subgroups were further divided into two subsets, one was with the 1st week dexamethasone injection dose less than 42mg, another was with the dose equal to 42mg or greater.
In 1–7 day subgroups, OI at admission in ≥ 42mg dexamethasone subset were lower (P = 0.0131 and P = 0.0001 in two groups). ≥42mg dexamethasone subset of non-COPD group showed higher FVC% value, higher DLco%value and lower lung fibrosis score (P = 0.0419, P = 0.0110 and 0.0443, Table 3). higher DLco% and lower lung fibrosis score were also observed for the ≥ 42mg dexamethasone subset in severe COVID-19 with COPD group (P = 0.0458 and 0.0013, Table 3).
Table 3
3-month lung fibrosis score and pulmonary function of subsets with different 1st week corticosteroid dose in patients initiating corticosteroid within 7 days.
Category
|
Severe COVID-19 with COPD
Dexamethasone dose (mg) P value
< 42 ≥42
|
Severe COVID-19
Dexamethasone dose (mg) P value
< 42 ≥42
|
Subjects (n)
|
15
|
54
|
|
25
|
31
|
|
OI at admission( mean ± SD)
|
208.27 ± 45.92
|
172.28 ± 49.01
|
0.0131
|
272.76 ± 20.03
|
232.61 ± 44.19
|
0.0001
|
lung fibrosis score ( mean ± SD)
|
1.33 ± 0.82
|
0.72 ± 0.67
|
0.0013
|
1.28 ± 0.54
|
0.94 ± 0.12
|
0.0443
|
FEV1% (mean ± SD)
|
59.73 ± 7.60
|
62.11 ± 9.17
|
0.3616
|
86.67 ± 18.25
|
94.29 ± 13.04
|
0.05316
|
FVC% mean ± SD)
|
74.47 ± 20.71
|
82.35 ± 15.97
|
0.1181
|
80.21 ± 11.50
|
87.77 ± 14.61
|
0.0419
|
DLco% ( mean ± SD)
|
58.20 ± 9.09
|
68.09 ± 11.78
|
0.0458
|
65.92 ± 7.69
|
72.94 ± 11.14
|
0.0110
|
Note: 1 patient couldn’t tolerate pulmonary function measurement in the < 42 subgroup of severe COVID-19 group. |
SD, standard deviation; OI: oxygenation index |
In 8–14 day subgroup, a lower lung fibrosis score and a higher DLco% were observed for the ≥ 42mg dexamethasone subset in severe COVID-19 with COPD group (P = 0.0013 and P = 0.0205, respectively). While only a lower lung fibrosis score was observed for the ≥ 42mg dexamethasone subset in severe COVID-19 group (P = 0.0443, Table 4).
Table 4
3-month lung fibrosis score and pulmonary function of subsets with different 1st week corticosteroid dose in patients initiating corticosteroid in 8–14 days.
Category
|
Severe COVID-19 with COPD
Dexamethasone dose (mg) P value
< 42 ≥42
|
Severe COVID-19
Dexamethasone dose (mg) P value
< 42 ≥42
|
Subjects (n)(注)
|
17
|
20
|
|
30
|
26
|
|
OI at admission( mean ± SD)
|
199.94 ± 44.73
|
173.85 ± 44.67
|
0.0855
|
251.03 ± 33.33
|
202.19 ± 47.25
|
0.0000
|
lung fibrosis score ( mean ± SD)
|
2.18 ± 0.53
|
1.25 ± 0.64
|
0.0000
|
1.97 ± 0.93
|
1.65 ± 0.69
|
0.1633
|
FEV1% (mean ± SD)
|
62.00 ± 9.59
|
65.50 ± 9.76
|
0.3092
|
87.31 ± 9.34
|
86.00 ± 7.41
|
0.5700
|
FVC% mean ± SD)
|
75.80 ± 6.73
|
82.56 ± 13.71
|
0.0920
|
76.10 ± 10.11
|
78.38 ± 6.34
|
0.3275
|
DLCO% ( mean ± SD)
|
56.00 ± 6.85
|
65.67 ± 13.97
|
0.0205
|
63.14 ± 12.88
|
66.65 ± 7.88
|
0.2338
|
Note: each subgroup had 2 patients couldn’t tolerate pulmonary function measurement in severe COVID-19 with COPD group. For severe COVID-19 controls, 1 patient couldn’t tolerate pulmonary function measurement in the < 42 subgroup. |
For 15–21 day subgroup of severe COVID-19 controls, statistical analysis was not carried out because of the small sample size.
The influence of the time to initiate systemic corticosteroid therapy on 3-month survival
According to the time to initiate systemic corticosteroid therapy, severe COVID-19 with COPD group and severe COVID-19 group were divided into 3 subgroups respectively: the 1–7 day subgroup with an initiating time in 7 days, the 8–14 day subgroup with an initiating time in 8–14 days, and ≥ 15 day subgroup with an initiating time later than 14 days.
In severe COVID-19 with COPD group, the > 15 day subgroup patients all died of COVID-19 and related complications, the 1–7 day subgroup had a higher 3-month survival than the 8–14 day subgroup (Log-rank: P = 0.0001, Fig. 1A). In severe COVID-19 patients, the 1–7 day subgroup had a higher 3-month survival than other subgroups (Log-rank: P = 0.0109, Fig. 1B).
For the 3-month survival of the 1–7 day subgroups of two groups, no significant difference was found (Log-rank: P = 0.1578, Fig. 2A). Nevertheless, 3-month survival of the 8–14 day subgroup of severe COVID-19 with COPD patients was higher than non-COPD patients (Log-rank: P = 0.0481, Fig. 2B). For the 3-month survival of the ≥ 15 day subgroups, no significant difference was observed between two groups (Log-rank: P = 0.2192, Fig. 2C).
Finally, the 3-month survival of severe COVID-19 with COPD group was higher than that of severe COVID-19 group (Log-rank: P = 0.0018, Fig. 3), while the initiating time of systemic corticosteroid therapy of severe COVID-19 with COPD group was earlier than that of severe COVID-19 group (7.17 ± 5.70 day VS 10.12 ± 5.78 day, P = 0.0000).
The influence of the 1st week dose of systemic corticosteroid on 3-month survival
In severe COVID-19 with COPD group initiating corticosteroid in 1–7 days, a relatively larger dose of dexamethasone injection of the 1st week had a higher 3-month survival (Log-rank: P = 0.0004, Fig. 4A). In severe COVID-19 group initiating corticosteroid in 1–7 days, the ≥ 42mg subset also had a higher 3-month survival (Log-rank: P = 0.0175, Fig. 4B).
In severe COVID-19 with COPD group initiating corticosteroid in 8–14 days, the ≥ 42mg subset still had a higher 3-month survival (Log-rank: P = 0.0067, Fig. 5A). In severe COVID-19 group, significant difference of 3-month survival was also observed between ≥ 42mg subset and < 42mg subset (Log-rank: P = 0.0252, Fig. 5B).
In severe COVID-19 patients initiating corticosteroid later than 14 days, low 3-month survivals were observed in ≥ 42mg subset and < 42mg subset, and no significant difference was observed between two subsets (Log-rank: P = 0.6299, Fig. 6). For ≥ 15 day subset of severe COVID-19 with COPD group, related statistical analysis was not carried out because of the small sample size.
Between two groups, the subsets of initiating corticosteroid in 1-7days with the 1st week dexamethasone dose < 42mg showed no significant difference of 3-month survival (Log-rank: P = 0.8872, Fig. 7A). No significant difference of 3-month survival was observed in ≥ 42mg subsets, either (Log-rank: P = 0.4150, Fig. 7B). Between the subsets of initiating corticosteroid in 8–14 days with the 1st week dexamethasone dose < 42mg, no significant difference of 3-month survival was observed between two groups (Log-rank: P = 0.4936, Fig. 7C), while a higher 3-month survival was observed in the ≥ 42mg subset of severe COVID-19 with COPD group(Log-rank: P = 0.0472, Fig. 7D).
Finally, the 3-month survival of severe COVID-19 with COPD group was higher than that of severe COVID-19 group (Log-rank: P = 0.0018, Fig. 3), while the 1st week dose of systemic dexamethasone of severe COVID-19 with COPD group was larger than that of severe COVID-19 group (46.41 ± 26.56 mg VS 38.49 ± 31.56 mg, P = 0.0142).
Cox regression analysis of clinical parameters on 3-month cumulative survival
For smoking is one of the risk factors of COPD, it was not included in Cox regression analysis. Univariate Cox regression analysis showed that age, time to initiate systemic corticosteroid therapy and the 1st week dose of systemic corticosteroid were significantly associated with 3-month survival in severe COVID-19 with COPD group (Table 5). Further, multivariate Cox regression analysis was performed for clinical parameters with significance. Finally, age and the 1st week dose of corticosteroid showed significant association with 3-month survival (Table 6)。
Table 5
Univariate Cox regression analysis of clinical parameters for 3-month survival in severe COVID-19 with COPD patients.
Variable
|
HR (95% CI)
|
P Value
|
Age
|
1.060(1.015–1.107)
|
0.008
|
Gender
|
1.021(0.688–1.515)
|
0.918
|
Time to initiate corticosteroid
|
1.074(1.038–1.111)
|
0.000
|
The 1st week dose of corticosteroid
|
0.952(0.935–0.968)
|
0.000
|
HR, hazard ratio; CI, confidence interval.
Table 6
Multivariate Cox regression analysis of clinical parameters for 3-month survival in severe COVID-19 with COPD patients.
Variable
|
HR (95% CI)
|
P Value
|
Age
|
1.045(1.002–1.091)
|
0.041
|
Time to initiate corticosteroid
|
1.006(0.961–1.053)
|
0.800
|
The 1st week dose of corticosteroid
|
0.956(0.936–0.976)
|
0.000
|
HR, hazard ratio; CI, confidence interval.
In severe COVID-19 group, univariate Cox regression analysis showed that gender, time to initiate systemic corticosteroid therapy and the 1st week dose of systemic corticosteroid were significantly associated with 3-month survival (Table 7). Multivariate Cox regression analysis finally indicated that gender was with significant association with 3-month survival (Table 8)。
Table 7
Univariate Cox regression analysis of clinical parameters for 3-month cumulative survival in severe COVID-19 patients.
Variable
|
HR (95% CI)
|
P Value
|
Age
|
1.010(0.989–1.031)
|
0.351
|
Gender
|
5.902(2.849–12.226)
|
0.000
|
Time to initiate corticosteroid
|
1.043(1.017–1.069)
|
0.001
|
The 1st week dose of corticosteroid
|
0.985(0.975–0.995)
|
0.005
|
HR, hazard ratio; CI, confidence interval.
Table 8
Multivariate Cox regression analysis of clinical parameters for 3-month survival in severe COVID-19 patients.
Variable
|
HR (95% CI)
|
P Value
|
Gender
|
5.713(2.756–11.843)
|
0.000
|
Time to initiate corticosteroid
|
1.026(0.999–1.054)
|
0.062
|
The 1st week dose of corticosteroid
|
0.990(0.980-1.000)
|
0.054
|
HR, hazard ratio; CI, confidence interval.