Eighty-four ARDS survivors of the first three COVID-19 pandemic waves were included in the study (Fig. 1). The ICU mortality rate was not different between w1 and w2 (29% and 36%, p = 0.29) but was significantly lower during w3 vs w1 (18% vs 29%, p = 0.05) and vs w2 (18% vs 36%, p = 0.0002). Demographics and clinical data for the 84 included patients are summarized in Table 1.
Table 1
Baseline characteristics of patients, according to wave of the pandemic
Variables (mean, SD or n (%) | Wave 1 n = 36 | Wave 2 n = 27 | Wave 3 n = 21 | P value (Anova or Kruskal Wallis tests) | p value (wave 1 vs. wave 2) | p value (wave 1 vs. wave 3) | p value (wave 2 vs. wave 3) |
Age | 56.64 ± 10.33 | 59.48 ± 10.92 | 53.1 ± 13.06 | 0.1555 | 0.8868 | 0.7876 | 0.2146 |
Sex, female | 8 (22.2%) | 11 (14.7%) | 12 (57.1%) | 0.0274 | 0.2872 | 0.0246 | 0.5205 |
non obese | 14 (38.9%) | 13 (48.15%) | 11 (52.38%) | 0.5739 | 0.7630 | 0.6111 | 0.9585 |
active smokers | 3 (8.33%) | 20 (74.07%) | 1 (4.76%) | 0.6047 | 0.6814 | 0.9889 | 0.6679 |
Medical history cancer diabetes hypertension HIV OSA | 1 (2.78%) 15 (41.67%) 20 (55.56%) 2 (5.56%) 6 (16.67%) | 1 (3.70%) 15 (55.56%) 17 (62.96%) 0 1 (3.70%) | 1 (2.78%) 8 (38.10%) 10 (47.62%) 0 2 (9.52%) | 0.9261 0.4111 0.5677 N/A 0.2526 | 0.9795 0.5451 0.8377 0.1974 | 0.9346 0.9652 0.8453 0.7258 | 0.9840 0.4732 0.5634 0.7318 |
Length of stay (days) Hospital ICU | 34.72 ± 20.76 20.5 ± 15.97 | 27.81 ± 16.93 15.04 ± 14.03 | 23.43 ± 14.25 16.95 ± 11.72 | 0.6097 0.7744 | 0.1573 0.0546 | 0.0304 0.5661 | 0.4205 0.2543 |
ICU severity scores APACHE SOFA | 10.06 ± 4.05 3.67 ± 2.29 | 9.04 ± 4.82 3.11 ± 1.25 | 7.57 ± 5.02 3.62 ± 1.69 | 0.3045 0.1185 | 0.1587 0.6651 | 0.0038 0.5879 | 0.1347 0.3734 |
respiratory support mech. ventilation mean duration curare prone high flow oxygen CPAP BIPAP ECMO | 23 (63.89%) 12.22 ± 13.44 21 (58.33%) 16 (44.44%) 9 (25.00%) 0 2 (5.56%) 3 (8.33%) | 9 (33.33%) 4.67 ± 7.58 7 (25.93%) 10 (37.04%) 8 (29.63%) 0 9 (33.33%) 1 (3.70%) | 10 (47.62%) 7.86 ± 9.82 9 (42.86%) 8 (38.09%) 1 (4.76%) 4 (19.05%) 6 (28.57%) 3 (14.29%) | 0.0543 0.1099 0.0371 0.8117 0.0900 N/A 0.0137 0.4207 | 0.0412 0.0088 0.0227 0.8377 0.9205 0.0186 0.7325 | 0.4849 0.2146 0.5218 0.8946 0.0621 0.0935 0.8003 | 0.6020 0.2621 0.4661 0.9972 0.0438 0.9388 0.4571 |
Laboratory data D-dimer, mg/dL CRP, mg/dL | 9137.57 ± 244.52 ± 106.17 | 4436.48 ± 174.03 ± 95.62 | 6993.19 ± 9353.9 165.0 ± 94.07 | 0.1408 0.5058 | 0.1439 0.0108 | 1.0000 0.0088 | 0.8471 0.8085 |
Treatments glucocorticoids antiviral remdesevir monoclonal antibody tocilizumab broad spectrum antibiotics | 21 (58.33%) 3 (8.33%) 3 (8.33%) 1 (2.78%) 7 (19.44%) 31 (86.11%) | 27 (100%) 0 2 (7.41%) 0 1 (3.70%) 23 (85.19%) | 21 (100%) 0 0 0 15 (71.43%) 11 (52.38%) | N/A N/A N/A N/A < 0.0001 0.0067 | 0.9908 0.1126 0.9947 | < 0.0001 0.0237 | < 0.0001 0.0387 |
Complications thromboembolic event | 12 (33.33%) | 5 (18.52%) | 3 (14.29%) | 0.1953 | 0.3934 | 0.2206 | 0.9246 |
ICU: intensive care unit, ECMO: extracorporal membranous oxygenation ; OSA, obstructive sleep apnea |
Mean hospital length of stay was shorter during w3 vs w1 (23.4 ± 14.2 days vs 34.7 ± 20.8 days, p = 0.03). Fewer patients required mechanical ventilation (MV) during w2 vs w1 (33.3% vs 63.9%, p = 0.0038), such that the use of curare was also reduced (p = 0.0227). The use of tocilizumab increased significantly during w3 (71.4% vs 19.4% for w1, p < 0.0001 and vs 3.7% for w2, p < 0.0001) while broad-spectrum antibiotic use decreased.
Three months after ICU discharge, no differences in the number of patients exhibiting impairment in the 6MWDT (< 80%) were observed between waves. The results of PFTs, RMS tests, and 6MWDTs were similar overall, regardless of the wave. Quality of life, assessed by the SF-36, was worse in the vitality and mental health items for patients from w1 vs w3 (total scores 64.7 ± 16.3 vs 49.2 ± 23.2, p = 0.0169). These results are summarized in Table 2.
Table 2
Comparisons of clinical outcomes between waves
Variables (mean, SD or %) | Wave 1 | Wave 2 | Wave 3 | P value (Anova or K.Wallis) | p (w1 vs. w2) | p (w1 vs. w3) | p (w2 vs. w3) |
PFT FEV1 (L) FEV1 (%) FVC (L) FVC (%) FEV1/FCV DLCO (%) | 2.88 ± 0.72 87.92 ± 14.71 3.41 ± 0.83 82.28 ± 15.73 0.83 ± 0.07 81.12 ± 18.32 | 2.47 ± 0.82 79.81 ± 21.06 2.95 ± 0.96 74.52 ± 19.58 0.83 ± 0.08 78.24 ± 22.23 | 2.26 ± 0.74 79.19 ± 17.67 2.8 ± 0.94 78.33 ± 18.25 0.82 ± 0.1 75.35 ± 22.34 | 0.0096 0.1056 0.0322 0.2299 0.2218 0.6118 | 0.1230 0.2315 0.1466 0.2585 0.8973 1.0 | 0.0094 0.1489 0.0443 1.0 0.9969 0.9340 | 1.0 1.0 1.0 1.0 0.9130 1.0 |
6MWDT distance (m) distance (%) distance < 80% oxygen desaturation | 500.18 ± 89.07 73.05 ± 11.36 n = 24 (66.67%) n = 11 (33.33%) | 417.71 ± 132.89 83.67 ± 91.36 n = 19 (70.37%) n = 6 (24.00%) | 434.1 ± 109.0 67.88 ± 16.45 n = 14 (66.67%) n = 6 (30.00%) | 0.0250 0.1831 0.7904 0.7411 | 0.0533 0.2060 0.8134 0.7331 | 0.0596 0.5454 1.0000 0.9683 | 1.0 1.0 0.8564 0.9038 |
RMS IP max (cm H20) EP max (cm H20) | 89.03 ± 28.06 105.95 ± 32.06 | 70.98 ± 34.01 87.71 ± 39.09 | 77.81 ± 35.9 85.57 ± 36.07 | 0.0931 0.0637 | 0.0802 0.1612 | 0.6038 0.1040 | 1.0 1.0 |
mMRC | 0.74 ± 0.96 | 1.26 ± 1.1 | 1.33 ± 1.2 | 0.0722 | 0.1551 | 0.1399 | 1.0000 |
PCFS | 1.31 ± 1.18 | 1.56 ± 1.19 | 1.67 ± 1.43 | 0.7272 | 1.0 | 1.0 | 1.0 |
SF36 | 64.71 ± 16.26 | 57.23 ± 21.46 | 49.19 ± 23.24 | 0.0241 | 0.4004 | 0.0169 | 0.6743 |
PFT: pulmonary function test, FEV: forced expiratory volume; FVC: forced vital capacity; DLCO: diffusing capacity of the lungs for carbon monoxide; 6MWT: 6-minute walking distance test, RMS: respiratory muscle strength, SF36: Short Form 36, PCFS: post-COVID Functional Status, mMRC: modified Medical Research Council dyspnea scale, IP: inspiratory pressure, EP: expiratory pressure |
No differences in affected segments were observed at baseline and 3-month chest CT, but at 3 months, more patients from w2 vs w1 exhibited ground glass opacities (41% vs 3%, p = 0.0006) and fibrosis was much more frequent in w1 vs w2 (94% vs 67%, p = 0.0186).
The multiple linear or logistic regression analysis highlighted that for w1 patients, total lung capacity (TLC), forced expiratory volume 1 (FEV1), diffusing capacity of the lungs for carbon monoxide (DLCO), maximum inspiratory pressure (PI Max), and 6MWDT at 3 months were correlated with mechanical ventilation (MV) and high APACHE II score (all p < 0.05 and p ≤ 0.01, respectively), as well as SF-36 (p = 0.006). The number of affected segments on 3-month CT correlated with the initial number of affected segments on chest CT scan (p = 0.04) and with high APACHE II score (p = 0.04).
For w2 patients, TLC, FEV1, DLCO, and PI Max at 3 months were correlated with MV and high SOFA score (all p ≤ 0.02 and p ≤ 0.03, respectively). SF-36 correlated with SOFA (p = 0.02), and PCFS with extracorporeal membrane oxygenation (ECMO) duration (p = 0.03).
For w3 patients, TLC, FEV1, DLCO, and 6MWDT were correlated with MV (p ≤ 0.01) as was SF-36 (p = 0.03). Low SF-36 and high PCFS scores were correlated with low 6MWDT (p = 0.01 and 0.03).
The use of glucocorticoids was associated with better TLC, FEV1, DLCO, and number of affected segments on 3-month CT (all p < 0.0001), and tocilizumab was associated with higher TLC (p = 0.027) (w1), and with a reduction in affected segments on 3-month CT (p = 0.0065) (w2). Remdesevir improved MV duration in w2 (p = 0.0083).
Patients in w3 suffering from acute thromboembolic events exhibited a decreased 6MWDT (p = 0.035) and a higher number of affected segments on 3-month CT (p = 0.0293).