From December 2013 to November 2020, 211 patients received ECMO support at our ICU. Twenty-four patients were excluded from this study because they received V-A ECMO or ECMO support for surgery or internal interventional procedures. Among the 187 ARDS patients who received V-V ECMO support, 30 (16%) underwent tracheostomy. Eighteen of these tracheostomies (60%) were performed while the patient was receiving ECMO support, and the other 12 were performed after ECMO decannulation. The flow chart is shown in Fig. 1.
Among the 18 patients who underwent tracheostomy while receiving ECMO support, the average age was 50.6 years; 15 patients (83.3%) were male, and 5 patients had immunodeficiency (27.8%). The median PaO2/FiO2 was 62.7mmHg, and the median RESP, PRESERVE, APACHE Ⅱ, SOFA, and Murray scores were 2.2, 3.9, 14.7, 7.9, and 3.3, respectively. The risk factors for ARDS were severe pneumonia (13 patients) and interstitial lung disease (5 patients). The pathogens of the 13 patients with severe pneumonia were influenza virus in 8, Pneumocystis jirovecii in 2, adenovirus in 1, aspergillus in 1 and Mycobacterium tuberculosis in 1. The 5 patients with interstitial lung disease included 1 with acute interstitial pneumonitis, 1 with acute exacerbation of idiopathic pulmonary fibrosis, 1 with secondary organizing pneumonia and 2 with interstitial lung disease associated with anti-synthase syndrome and dermatomyositis with MDA5 antibody positivity.
Tracheostomy was performed between 0 and 31 days after the establishment of ECMO. The median interval from ECMO cannulation to tracheostomy was 5 days, with 25th and 75th percentiles of 0 days and 9.5 days, respectively. Eleven patients (61.1%) received ET, and 7 received DT. In the ET group, the median interval from ECMO cannulation to tracheostomy was 1 day, and the 25th and 75th percentiles were 0 days and 5 days, respectively. The median interval from ECMO cannulation to tracheostomy in the DT group was 11 days, and the 25th and 75th percentiles were 9 days and 16 days, respectively. Figure 2 shows the distribution of the intervals between ECMO cannulation and tracheostomy. No significant difference was found between the ET and DT groups in terms of demographic data, medical history, disease severity (estimated using the RESP, PRESERVE, APACHE Ⅱ, SOFA and Murray scores), ARDS risk factors or duration of mechanical ventilation before ECMO. Table 1 shows the baseline characteristics of the two groups.
The overall incidence rate of VAP during ECMO support was 66.7% in this retrospective cohort. In the ET group, 5 patients acquired VAP; 7 patients in the DT group acquired VAP. The ET group showed a more than 50% decrease in the incidence of VAP during ECMO support (45.5% vs. 100%; P = 0.038). Figure 3 shows the accumulated 28-day VAP rates of the two groups; there was a departure of the two curves, but the difference did not reach significance (chi square = 1.572, P = 0.21; HR 0.484, 95% CI: 0.145–1.676).The median interval from ECMO cannulation to the occurance of VAP was 2 days (from 1 day to 21 days) and 8 days (from 4 to 32 days) in the ET and DT groups, respectively, with no signifiant difference (P = 0.149). Shorter durations of ECMO (9.0 vs. 27.0 days; P = 0.011) and mechanical ventilation (16.0 vs. 56.0 days; P = 0.027) were observed in the ET group.However, ET did not significantly alter the all-cause mortality rate in the ICU (54.5% vs. 28.6%; P = 0.367), the all-cause mortality rate in the hospital (which was the same as the ICU mortality rate), the length of ICU stay (336 vs. 627 hours; P = 0.085) or the length of hospital stay (26 vs. 37 days; P = 0.285). The incidence of tracheostomy complications did not differ between the two groups (local bleeding: 27.3% vs. 42.9%, P = 0.627). No procedure-related tracheal rupture, pneumothorax, subcutaneous emphysema or periprocedural clotting events were reported. No major bleeding occured. Table 2 shows the outcomes of the two groups.
We also compared the outcomes of patients with or without tracheostomy-related local bleeding (Table 3). There was no significant difference between the two groups in the all-cause mortality rates in the ICU or in the hospital, VAP incidence, duration of mechanical ventilation, duration of ECMO support, length of ICU stay or length of hospital stay (P༞0.05).
Table 1
Baseline characteristics according to tracheostomy timing
| All patients (n = 18) | Early tracheostomy (n = 11) | Delayed tracheostomy (n = 7) | P |
Age, years | 50.6 ± 20.3 | 53.5 ± 19.9 | 46.1 ± 21.8 | 0.474 |
Male sex | 15 (83.3) | 9 (81.8) | 6 (85.7) | 1.000 |
Body mass index, kg/m2 | 23.5 ± 3.9 | 23.7 ± 4.5 | 23.2 ± 3.1 | 0.799 |
Comorbidity | | | | |
Idiopathic pulmonary fibrosis | 1 (5.6) | 1 (9.1) | 0 | 1.000 |
Hypertension | 5 (27.8) | 4 (36.4) | 1 (14.3) | 0.596 |
Diabetes mellitus | 2 (11.1) | 1 (9.1) | 1 (14.3) | 1.000 |
Coronary heart disease | 3 (16.7) | 3 (27.3) | 0 | 0.245 |
Chronic heart dysfunction | 1 (5.6) | 1 (9.1) | 0 | 1.000 |
Chronic kidney disease | 2 (11.1) | 2 (18.2) | 0 | 0.497 |
Pregnancy | 1 (5.6) | 0 | 1 (14.3) | 0.389 |
Active malignancy | 1 (5.6) | 1 (9.1) | 0 | 1.000 |
Immunodeficiency | 5 (27.8) | 4 (36.4) | 1 (14.3) | 0.596 |
Smoking | 10 (55.6) | 6 (54.5) | 4 (57.1) | 1.000 |
RESP score | 2.2 ± 2.9 | 2.2 ± 3.1 | 2.1 ± 2.9 | 0.979 |
PRESERVE score | 3.9 ± 2.0 | 4.4 ± 1.6 | 3.1 ± 2.3 | 0.204 |
Murray score | 3.3 ± 0.6 | 3.4 ± 0.5 | 3.1 ± 0.6 | 0.287 |
SOFA score | 7.9 ± 3.5 | 8.4 ± 4.1 | 7.0 ± 1.9 | 0.491 |
APACHE Ⅱ score | 14.7 ± 6.4 | 13.8 ± 6.1 | 16.2 ± 7.3 | 0.519 |
ARDS risk factor | | | | |
Pneumonia | 13 (72.2) | 7 (63.6) | 6 (85.7) | 0.596 |
Idiopathic or secondary interstitial lung disease | 5 (27.8) | 4 (36.4) | 1 (14.3) | 0.596 |
PaO2/FiO2 on ECMO initiation, mmHg | 62.7 (56.5, 75.2) | 64.0 (57.0, 82.3) | 61.5 (54.0, 70.2) | 0.350 |
MV time before ECMO, hours | 44.5 (5.0, 149.8) | 96.0 (2.0, 224.0) | 8.0 (6.0, 96.0) | 0.659 |
APACHE, Acute Physiology and Chronic Health Evaluation; ARDS, acute respiratory distress syndrome; ECMO, extracorporeal membrane oxygenation; MV, mechanical ventilation; PaO2/FiO2, ratio of partial pressure of arterial oxygen to fraction of inspired oxygen; PRESERVE, PRedicting dEath for SEvere ARDS on VV-ECMO; RESP, Respiratory ECMO Survival Prediction; SOFA, Sequential Organ Failure Assessment. |
Data are shown as n (%), mean ± standard deviation, or median (25th, 75th percentiles).
Table 2
Outcomes for different tracheostomy times
| All patients (n = 18) | Early tracheostomy (n = 11) | Delayed tracheostomy (n = 7) | P |
Primary outcomes | | | | |
ICU death | 8 (44.4) | 6 (54.5) | 2 (28.6) | 0.367 |
Hospital death | 8 (44.4) | 6 (54.5) | 2 (28.6) | 0.367 |
Secondary outcomes | | | | |
VAP during ECMO | 12 (66.7) | 5 (45.5) | 7 (100.0) | 0.038 |
Tracheostomy complications | | | | |
Bleeding | 6 (33.3) | 3 (27.3) | 3 (42.9) | 0.627 |
Wound bleeding | 4 (22.2) | 2 (18.2) | 2 (28.6) | 1.000 |
Intratracheal bleeding | 3 (16.7) | 2 (18.2) | 1 (14.3) | 1.000 |
Mechanical ventilation duration, days | 20.0 (12.0, 57.0) | 16.0 (10.0, 33.0) | 56.0 (19.0, 73.0) | 0.027 |
ECMO duration, days | 14.0 (7.5, 29.0) | 9.0 (6.0, 19.0) | 27.0 (17.0, 38.0) | 0.011 |
ICU LOS, hours | 613.0 (330.0, 1008.0) | 336.0 (260.0, 850.0) | 672.0 (459.0, 2208.0) | 0.085 |
Hospital LOS, days | 27.0 (16.25, 76.25) | 26.0 (11.0, 37.0) | 37.0 (19.0, 92.0) | 0.285 |
ECMO, extracorporeal membrane oxygenation; ICU, intensive care unit; LOS, length of stay; VAP, ventilator-associated pneumonia. |
Data are shown as n (%), mean ± standard deviation, or median (25th, 75th percentiles).
Table 3
Outcomes of patients with or without tracheostomy-related local bleeding
| All patients (n = 18) | With bleeding (n = 6) | Without bleeding (n = 12) | P |
ICU death | 8 (44.4) | 3 (50.0) | 5 (41.7) | 1.000 |
Hospital death | 8 (44.4) | 3 (50.0) | 5 (41.7) | 1.000 |
VAP during ECMO | 12 (66.7) | 4 (66.7) | 8 (66.7) | 1.000 |
Mechanical ventilation duration, days | 20.0 (12.0, 57.0) | 23.5 (15.8, 72.3) | 19.0 (12.0, 53.0) | 0.616 |
ECMO duration, days | 14.0 (7.5, 29.0) | 14.5 (8.8, 23.5) | 14.0 (6.0, 29.0) | 0.892 |
ICU LOS, hours | 613.0 (330.0, 1008.0) | 624.0 (426.0, 2010.0) | 554.0 (273.0, 860.0) | 0.437 |
Hospital LOS, days | 27.0 (16.3, 76.3) | 26.0 (17.8, 83.0) | 27.5 (12.5, 65.5) | 0.892 |
ECMO, extracorporeal membrane oxygenation; ICU, intensive care unit; LOS, length of stay; VAP, ventilator-associated pneumonia. |
Data are shown as n (%) or median (25th, 75th percentiles).